CRYSTALLINITY IN LINEAR POLYAMIDES: A

UNIVERSITY OF SOUTH AUSTRALIA

CRYSTALLINITY IN LINEAR POLYAMIDES: A STUDY USING MELT BLENDING WITH SMALL-MOLECULE DILUENTS by

John Pockett Bachelor of Science

A thesis submitted to the University of South Australia for the Degree of Doctor of Philosophy (Applied Science)

Ian Wark Research Institute University of South Australia Mawsons Lakes, Australia

October 2004

TABLE OF CONTENTS

Table of Contents

i

List of Tables

v

Glossary of Terms

vi

Abstract

viii

Declaration

x

Acknowledgments

xi

Chapter 1

1

Introduction

1

1.1

Preamble

2

1.2

Review of background theory

4

1.3

Relevant papers in the area to be covered in the research

40

1.4

The focus of the research project

45

1.5

Experimental Techniques Used

50

1.6

Structure of the Thesis

74

1.7

Summary

75

Chapter 2

77

Experimental

77

2.1

Introduction

77

2.2

Materials

78

2.3

Materials handling

78

2.4

Preparation of Melt Blends

81

2.5

Thermogravimetric Analysis

86

2.6

Differential Scanning Calorimetry

87

2.7

Simultaneous Differential Thermal Analysis-Thermogravimetric Analysis

92

2.8

Fourier Transform Infra-Red Spectroscopy

93

2.9

Small Angle X-Ray Scattering

95

2.10

Cross Polarised/Magic Angle Spinning Solid State 13C Nuclear Magnetic Resonance Spectroscopy

95

Chapter 3

97

Polyamide-4,6 with carbazole

97

3.1

Introduction

98

3.2

Materials, Handling, Sample Preparation and Techniques Used

98

i

3.3


100

3.4


104

3.5

Fourier Transform Infra Red Spectroscopy

129

3.6

Small Angle X-ray Scattering

135

3.7

Nuclear Magnetic Resonance Spectroscopy

136

3.8

Summary

139

Chapter 4

144

Polyamide-6 with carbazole

144

4.1

Introduction

144

4.2


145

4.3


148

4.4


165

4.5


170

4.6

Summary

171

Chapter 5

174

Polyamide-6,9 with carbazole

174

5.1

Introduction

174

5.2


175

5.3


176

5.4


190

5.5

Summary

190

Chapter 6

192

Polyamide-6,12 with Carbazole

192

6.1

Introduction

192

6.2


193

6.3


194

6.4


208

6.5

Summary

208

Chapter 7

211

Polyamide-4,6 with phenothiazine

211

7.1

Introduction

211

7.2


212

7.3


213

ii

7.4


231

7.5

Summary

231

Chapter 8

234

Polyamide-6 with phenothiazine

234

8.1

Introduction

234

8.2


235

8.3


236

8.4


251

8.5

Summary

251

Chapter 9

253


253

9.1

Introduction

253

9.2


254

9.3


255

9.4


270

9.5

Summary

270

Chapter 10

272


272

10.1

Introduction

272

10.2


273

10.3


274

10.4


289

10.5

Summary

289

Chapter 11

291

General Conclusions

291

11.1

Thesis arguments

291

11.2

Overall common relationships

299

11.3

Hydrogen bonding

300

11.4

Other areas of potential study for future workers

302

11.5

Practical implications of the work

303

11.6

Summary of conclusions

303

iii

Appendix A

304

Further detail from DSC thermograms

304

A.1

Chapter 5 polyamide-6,9 with carbazole

304

A.2

Chapter 6 polyamide-6,12 with carbazole

305

A.3

Chapter 8 polyamide-6 with phenothiazine.

306

A.4

Chapter 9 polyamide-6,9 with phenothiazine

308

A.5

Chapter 10 polyamide-6,12 with phenothiazine

311

Appendix B

313

DSC, TMDSC and Lissajous Figures

313

B.1

Introduction

313

B.2

Experimental

325

B.3

Results and Discussion

327

B.4

Conclusions

333

Appendix C

335

Mid range IR Assignments for polyamide types studied in thesis

335

C.1

Mid range IR and hydrogen bond interactions

Bibliography

335

338

iv

LIST OF TABLES

Table C-1 Assignments for polyamide bands Bold= [99, pp 85-88], Italic [48, p. 504], Normal from ampoule samples and PA46 Gaymans are solution cast FTIR peaks from [122] in cm-1

v

335

GLOSSARY OF TERMS

Aramid

Polyamide with aromatic groups in the main chain

ATR

Attenuated Total Reflection

Dalton

One twelfth of the mass of a carbon-12 atom in its ground state

DRIFT

Diffuse Reflection Infra-red Fourier Transform

DSC

Differential Scanning Calorimeter

DTA

Differential Thermal Analysis

FTIR

Fourier Transform Infra Red

G

Free Energy (Gibbs)

H

Enthalpy

∆Hf

Heat of Fusion

iPP

Isotactic poly(propylene)

l

Length

LCST

Lower Critical Solubility Temperature

µi

Chemical potential equivalent to the partial molar Gibbs Free energy of a mixture constituent

MAS

Magic Angle Spinning

Mn

Number average molecular weight

MTDSC

Modulated Temperature Differential Scanning Calorimetry

Mw

Weight average molecular weight

NIR

Near Infra Red range covering wavenumbers between 11,000 and 4,000 cm-1.

NMR

Nuclear Magnetic Resonance

Nylon

Aliphatic polyamide

PA-n

Polyamide-n where n is the number of carbon atoms per repeat unit

PA-m,n

Polyamide-m,n where m and n are the number of carbon atoms per diamine and diacid repeat unit

PA46

Polyamide-4,6

PA46Car

Polyamide-4,6/carbazole blend

PA46PTh

Polyamide-4,6/phenothiazine blend

PA6

Polyamide-6 vi

PA6Car

Polyamide-6/carbazole blend

PA6PTh

Polyamide-6/phenothiazine blend

PA69

Polyamide-6,9

PA69Car


PA69PTh


PA612

Polyamide-6,12

PA612Car


PA612PTh


PAS

PhotoAcoustic Spectroscopy

PEG

Poly(ethyleneglycol)

PEO

Poly(ethyleneoxide)

PVDF

Poly(vinylidenefluoride)

R

Gas Constant

Reptate S, ∆S

Entropy, Entropy change

SAXS


SDT

Simultaneous Differential Thermal Analysis – Thermogravimetric Analysis

t

Time

T

Absolute Temperature (0K)

TB

Brill temperature

Tc

Peak crystallisation temperature

Tg

Glass Transition Temperature

TLS peak

Temperature Limited Solubility peak (see Section 1.5.2.2)

Tm

Peak melting temperature

TGA


TIPS

Thermally Induced Phase Separation

TMDSC

Temperature Modulated Differential Scanning Calorimetry

UCST

Upper Critical Solubility Temperature

V

Volume

WAXD

Wide Angle X-Ray Diffraction

vii

ABSTRACT

Linear polyamides, commonly known as Nylons, are widely used for their high

melting

temperatures,

heat

stability,

toughness

and

abrasion

resistance, allowing diverse commercial applications such as carpets, nylon stockings and automotive parts. The work here has possible ramifications for membrane production and drug delivery systems and makes a scientific contribution to the area of binary polymer/diluent systems where the polymer is semicrystalline and the diluent crystallises at a quite different temperature to the polymer.

Melt blended crystalline/crystalline systems

have, so far, not received the attention that amorphous/amorphous or crystalline/amorphous systems have, perhaps due to the complexity of the morphology that often results within such systems. The polyamide-4,6, polyamide-6, polyamide-6,9, and polyamide-6,12 studied here are representative of the range of available linear polyamides. The organic diluents were initially chosen as potential hydrogen bond disruptors. Investigative techniques concentrated on Differential Scanning Calorimetry and Fourier transform infrared spectroscopy.


was also used for the determination of diluent levels in samples. It was found that melt blends where the crystallisation temperature of the undiluted polyamide is less than that of the neat diluent lead to characteristic solidification during slow cooling at 2 0C/min. Here, there is a linear drop off in enthalpy of crystallisation for the diluent component of the blend.

This reaches zero near the “eutectic” composition.

There are no

endothermic peaks at the “eutectic” temperature for blends where polyamide is in excess of the eutectic composition. There is depression of polymer and diluent crystallisation by each other in direct proportion to the concentration of the other material when the crystallisation temperature relationships are reversed. A wider scatter in the enthalpy of crystallisation of the diluent component ensues.

Prior

crystallisation close to the polyamide crystallisation temperature is more probable, especially when the cooling rate is greater.

viii

The range of polyamides studied shows the effects of polyamide repeat structure on eutectic compositions and on the maximum crystallisation and melting depressions for the three polyamides having relatively close melting temperatures. Choosing the diluents as probable model polyamide-polyamide hydrogen bond disruptors has potentially been important for a variety of applications. Hydrogen bond donor groups of the two diluents have been found not to interact with the carbonyl acceptors in the solidified polyamides.

Fourier

Transform Infra-Red spectroscopy (FTIR) carried out with photoacoustic spectroscopy in the Mid Range IR on all blend samples have found no evidence of N-H peak shifts that differ from simple summation of the spectra. This point is supported by Near Infra-Red experiments on all blends using the DRIFT technique in sensitive hydrogen bonding overtone regions. Some information in the Mid Range IR band 1700 to 400 cm-1 is ambiguous because of the non-linear response of photoacoustic measurements.

In

addition, very bright pink colours were observed in some samples, implying pi-conjugation or charge transfer processes. The thesis includes discussion in an appendix on the general applicability of Temperature Modulated Differential Scanning Calorimetry and the use of Lissajous figures to evaluate experimental conditions such as purge gas type and flow rates.

ix

DECLARATION

To the author’s knowledge and belief, the material in this thesis, except where due reference is made or where common knowledge is assumed, is original. No part of this work has been submitted for any other degree or award in any university.

John Pockett

x

ACKNOWLEDGMENTS

I wish to thank my supervisors Prof. Roger Horn, Prof. Jani Matisons and Dr. Namita Roy-Choudhury for their assistance during the course of this work. I would also like to thank the other (current and former) members of the Polymer sector of the Ian Wark Research Institute at the University of South Australia for support in various forms. Most notable amongst them are Dr. Stephen Clarke, Dr Mark Fisher and Dr. Leanne Britcher for a range of advice and Mr Clint Gamlin for experimental work he carried out at the University of Connecticut in the USA. Dr. Clarke contributed 10-15% of the text in Appendix B on Temperature Modulated Differential Scanning Calorimetry and Lissajous Figures in what we are intending submitting soon for publication in a journal article. Adj. Prof. Dennis Mulcahy and Dr. David Davey of Pharmaceutical, Molecular and Bio-Sciences of the University of South Australia, my wife, daughter and son have all made useful suggestions on the earlier chapters of the thesis and my daughter on the Conclusions chapter. The University of Connecticut is to be thanked for making its Small Angle X-Ray Scattering equipment available to Mr Clint Gamlin during his stay there in late 2000. Dr. Andrew Whittaker of the University of Queensland at St Lucia in Brisbane Qld, Australia has been kind enough to carry out many hours of Solid State Nuclear Magnetic Resonance Spectroscopy (NMR) experiments on my samples. The University of Queensland is to be thanked for the use of their NMR equipment. Dr Adam McCluskey of Environmental and Life Sciences-Chemistry at Newcastle University, NSW, Australia computed structures of carbazole and phenothiazine including electron density maps. Dr. Cor Koning (formerly) of DSM Research in the Netherlands provided a one-kilogram sample of polyamide-4,6 with no additives. He also provided some other small samples of different polyamides during a visit made to xi

DSM Research Laboratories in Geleen in the Netherlands during April 1999. Useful discussions took place with a number of people in various departments during that visit. Fruitful discussions also took place with Dr R J Gaymans and others at the University of Twente in the Netherlands and with Prof J A Subirana and others of Universidad Polytecnica Catalunya in Barcelona, Spain during visits to their establishments in April 1999. A number of authors from various Universities have provided reprints of articles they have (co-)authored when these have not been available at any of the libraries of the three Universities in South Australia. Some of these are J. Bou, G. Brooke, R.J. Gaymans, S. Montserrat, S. Munoz-Guerra, N Pace, P.C. Painter, J Puigalli, R. Scherrenberg, S. Simon, J. A. Subirana, and G. Ungar. Mr Craig Hackney, Dr. Barbara Brougham and Mr. Frank Peddie of the University of South Australia have helped immensely in IT areas, particularly in coping with the vagaries of Windows and the Office suite of software. Mr. Peddie has additionally discussed many matters at times over a cup of coffee. The thesis covers research work mainly carried out at the University of South Australia, Ian Wark Research Institute and that institute has provided a high quality environment in which to extend my skills. This research was carried out with the financial support of the University of South Australia in the form of a University of South Australia Postgraduate Research Award. The staff members of “Chem Tech” at the University of South Australia have provided a welcome respite from writing-up over short morning tea breaks in their Staff Room and have been very supportive of my studies. My wife Liesbeth, and children Monique and Lachlan, have provided unflagging support for my studies.

xii

Chapter 1

INTRODUCTION CONTENTS 1.1

Preamble

2

1.2

Review of background theory

4

1.2.1

Mixtures of materials

1.2.1.1 1.2.1.2 1.2.1.3 1.2.1.4 1.2.1.5 1.2.1.6 1.2.1.7 1.2.1.8

4

Phase diagrams Miscibility Partial Miscibility and changes in miscibility Phase separation, USCT and LCST Solid Solutions Eutectics and Monotectics Thermodynamics of Liquid Mixtures Kinetics of Phase Transformations

4 5 5 6 8 8 11 12

1.2.2

Hydrogen Bonding

13

1.2.3

Small organic molecules

14

1.2.3.1

1.2.4

14

Thermoplastics and Thermosets Basics including molecular weight and molecular shape Amorphous polymers Polymer crystallinity Lamellar melting Melting behaviour of semicrystalline polymers Spherulites Poorly and partially crystallising polymer types Polymer-polymer miscibility Polymer-diluent systems

Linear polyamides (Nylons)

1.2.5.1 1.2.5.2 1.2.5.3 1.2.5.4 1.2.5.5 1.2.5.6 1.2.5.7 1.2.5.8 1.2.5.9 1.2.5.10 1.2.5.11 1.2.5.12 1.2.5.13

1.3

14

Polymers

1.2.4.1 1.2.4.2 1.2.4.3 1.2.4.4 1.2.4.5 1.2.4.6 1.2.4.7 1.2.4.8 1.2.4.9 1.2.4.10

1.2.5

Melting and crystallisation

15 15 17 18 25 26 26 27 27 28

29

History of polyamides Strengths Weaknesses Chemical structure and polyamide types Biological-polyamide parallels Polyamide Hydrogen Bonding Polyamide Crystallinity Polyamide Crystalline Structures Effect of polyamide Type and Segment Length on Crystal Form Multiple crystalline forms are possible - Polymorphism Effect of pressure on crystallinity, melting temperature and crystal form Metastability Brill Temperature

Relevant papers in the area to be covered in the research

29 30 31 31 33 33 34 35 36 38 38 39 39

40

1.3.1

Small molecule-small molecule

40

1.3.2

Polymers with small molecules

40

1.3.3

Blend interactions and hydrogen bonding

41

1.3.4

Polyamides and Polymers

42

1

1.3.5

1.4

Polyamides and small molecules

43

The focus of the research project

45

1.4.1

Materials chosen

1.4.1.1 1.4.1.2

1.4.2

1.5

46

Polyamides Small molecules

46 47

Sample blending and notation used for blends

49

Experimental Techniques Used

50

1.5.1


50

1.5.2


51

1.5.2.1 1.5.2.2 1.5.2.3 1.5.2.4

Thermogram Overlays Thermograms expected from thermal events Assignment of “Spiky” Crystallisations to Carbazole or Phenothiazine Phase diagrams derived from thermograms

53 54 56 57

1.5.3

Simultaneous Differential Thermal Analysis/Thermogravimetric Analysis

61

1.5.4


62

1.5.4.1 1.5.4.2 1.5.4.3 1.5.4.4 1.5.4.5

General Mid Range IR and hydrogen bond Interactions Mid Range IR Frequencies of Interest Mid Infra-Red Data Analysis for Blends Near Infra –Red FTIR (NIR)

62 64 65 67 71

1.5.5


72

1.5.6

Solid state Nuclear Magnetic Resonance Spectroscopy

73

1.6

Structure of the Thesis

74

1.7

Summary

75

1.1 Preamble Linear polyamides, commonly known as Nylons, have a broad range of commercial applications.

They are used widely where their high melting

temperatures, high heat stability, toughness and abrasion resistance can be used to advantage. Detailed knowledge of their material properties is needed to optimise their processability and properties when blended with other materials for a variety of purposes such, as the formation of membranes. This thesis contributes to the understanding of high temperature solutions of semicrystalline linear polyamides melt blended with two different crystallisable

small-molecule

organic

compounds,

carbazole

and

phenothiazine. It also covers the crystallisation processes that take place during solidification to room temperature. It concludes that the major factor affecting the resulting nano- and microstructure of the solid is the relative crystallisation temperature of pure polyamide and compound. 2

There has been much work on semicrystalline polymers blended with amorphous polymers [1] .

There is not a great deal in the literature on

semicrystalline polymers blended with semicrystalline polymers [2-6]. The production of membranes with Thermally Induced Phase Separation (TIPS) by using amorphous polymers with crystallisable small molecule diluents has recently become an area of some interest for some people [7]. Little has been reported on the area of semicrystalline polyamides melt blended with small organic compounds [8]. Using the crystalline small organic molecules affects crystallisation strongly because the small molecules are highly mobile ahead of the crystallising polymer front. The work therefore makes an important contribution to a somewhat neglected area, particularly as it covers a range of polyamides with differing repeat units, melting temperatures and crystal structures. An investigation of these differences has led to a better understanding of high temperature solutions of polyamides with some small organic molecules and of the manner in which semicrystalline polyamides crystallise with normally highly crystalline small molecules. It will also enhance our knowledge of complex lamellar

formation

and

small

organic

molecule

crystallisation

in

a

semicrystalline combination of the two types of material. The major tools for the investigation are Differential Scanning Calorimetry (DSC), Thermogravimetric Analysis (TGA) and Fourier Transform Infra-Red spectroscopy (FTIR) in Mid-range Infra Red and in the Near-range Infra Red (NIR). The original purpose of this research had been to gain a better understanding of crystallinity in linear polyamides and an appreciation of how hydrogen bonding affects polymer crystallinity. The two types of small molecules used are potential hydrogen bond disruptors. The research has led to the different focus for the work because the two diluents were shown later by Fourier Transform Infra Red analysis not to interact in the solid state by hydrogen bonding with polyamides.

It was, however, recognised

that scientifically interesting questions arose from some of the experiments that had been undertaken. These had been found with material from the first sample made in an ampoule for producing bulk blends from the melt. These larger quantities were to be used for several characterisation 3

techniques requiring bigger samples than the few milligrams that could be produced in a DSC. The interesting results were the production of three very separate sections of the sample with quite different colours, brick red, white and fawn and of very different brittleness/hardness.

Thermogravimetric

Analysis (TGA) showed that the weight percentage polyamide was different in the samples and Differential Scanning (DSC) thermograms were also very different for the three. These results are discussed fully in Chapter 3.

1.2 Review of background theory Much of the information covered in the next few sections may be found in undergraduate textbooks on physical chemistry and materials science. It is, however, still worthwhile to refresh our memories and briefly draw all the basic concepts together to form the groundwork of the research environment of the project. The topics are covered in a relatively superficial manner and are meant only to lead the reader to the point where current research in the field is discussed. 1.2.1 Mixtures of materials What happens when two different materials are put together in the same environment at the same temperature and pressure will be explored. This will provide a basis for what happens when polyamides are heated with either carbazole or phenothiazine to the melt and then cooled down. 1.2.1.1

Phase diagrams

The main part of the research work covers mixtures of materials being studied by DSC in solid and liquid states at pressures near one atmosphere. In these conditions, the effects of ambient pressure are not intrusive in the measurements.

A simplification of the total equilibrium state is to only

consider solid and liquid phases, as will be done here.

The maximum

number of degrees of freedom for a system with two components is three so by fixing the ambient pressure at a nominal one atmosphere we can effectively consider only the two variables, temperature and composition. Equilibrium phase diagrams represent the different phases encountered in latter parts of the text. The phase diagrams are plots of temperature against composition with lines defining temperature/composition conditions that lead to regions where there is a common phase. The liquidus is the line that defines the lowest temperature where all material is in the liquid state. The 4

solidus is the line defining the highest temperatures where all the material is in the solid state. There are regions in addition to the all-liquid and all-solid states where there is a liquid coexisting with one or another solid material. The materials are considered to be in the ideal equilibrium conditions in the discussion immediately below. The reality of the experimental conditions is that the work has been done in non-equilibrium conditions and this will cause some modification of the outcomes.

The major experimental

technique used in the work was Differential Scanning Calorimeter (DSC) and from the DSC output we can observe some of the physical changes that take place with melting and crystallisation. The melting and crystallisation peaks can be interpreted to give an understanding of the underlying phase diagrams, although with the caveat that the observations may not lie at exactly the phase boundaries for equilibrium phase diagrams. 1.2.1.2

Miscibility

It is instructive to first discuss liquids before discussing multiphase solids. Two liquids are miscible in each other when the molecules of one material are completely dispersed in another at an atomic level for all concentrations, such as ethanol in water. It is common to talk of solvent and solute where one material (the solvent) is in substantial excess.

This becomes more

difficult in many of the cases to be discussed in this thesis because concentrations ranging from just over 20% polyamide to 83% polyamide are encountered as well as the pure materials. General usage of solvent is not necessarily the best because there will be a number of cases where the liquids are not in solution at specific temperatures. A more general word that will be used is diluent, which is suitable for all cases here.

In most

cases, the terms will not be raised and only the weight percentage polyamide referred to. 1.2.1.3

Partial Miscibility and changes in miscibility

Partial miscibility occurs when one liquid can only be added to another to a certain limit and then will not dissolve further.

The two materials will

separate out from one another into two layers if there is a difference in density or into droplets/blobs of one material in the other if there is little density difference. It should be noted here that there will be at least a small

5

amount of material A dissolved in material B and vice versa, even if the materials are essentially immiscible, such as water and oil. Liquid combinations that are immiscible at one temperature can often become miscible at other temperatures. For example if a mixture of phenol and water is heated to over 65 0C it becomes miscible. 1.2.1.4

Phase separation, USCT and LCST

A and B not in stable solution

Solution of A and B Temperature

Temperature

Upper Critical Solution Temperature (UCST)

Lower Critical Solution Temperature (LCST)


0

50 Percentage of Material B

Solution of A and B

100

0

Solution of A and B

Solution of A in B

A and B not in stable solution 50 Percentage of Material B

Metastable regions

Temperature

Temperature


0

100

Figure 1-3 Example of phase diagram with Lower Critical Solution Temperature.

Figure 1-1 Example of phase diagram with Upper Critical Solution Temperature.

Solution of B in A

50 Percentage of Material B

Binodal line Spinodal line

A and B unstable and spinodally decompose 0

100

50 Percentage of Material B Figure 1-4 Example of binodal and spinodal lines in a phase diagram.

Figure 1-2 Example of Phase diagram with immiscible region but no UCST or LCST.

6

100

The phenol/water case is an example of an Upper Critical Solution Temperature (UCST) where there is a maximum temperature at which the materials are not completely soluble. A typical phase diagram is given in Figure 1-1 The UCST does not have to lie at the centre of the concentration range but is often very strongly towards one or the other side of the phase diagram. There are also cases with some materials where there is a Lower Critical Solution Temperature (LCST) and once the temperature has been raised sufficiently the two materials begin to separate into separate phases. That can be seen in Figure 1-3. Another case is shown in Figure 1-2, where there is no upper or lower critical temperature but a region in mid concentrations where the materials are insoluble. Utracki [9] in his work on polymer/polymer miscibility states that UCST is more common in general with solvent-polymer and LCST with polymer-polymer systems.

There are

regions in composition-temperature space where the two materials cannot exist stably in a single, miscible, phase. This line is called the binodal curve. A metastable condition is often reached where the two materials still coexist without separating if the density of the two materials does not differ markedly. Another region exists within the binodal curve where the single phase nature of the liquid becomes completely impossible.

That inner

boundary is called the spinodal. Inside it the two materials will begin to phase separate spontaneously. An example showing the spinodal in a phase diagram is given in Figure 1-4.

Spinodal decomposition takes place throughout the mixed liquid

with very small volumes segregating themselves into like kinds of materials. This is energetically unfavourable because of the high interfacial surface area. Over a period of time, volumes of like material touch each other and reduce surface area by coalescing.

The entities of each material

progressively become bigger, as in Figure 1-5.

Figure 1-5 Ripening over time of small spinodally decomposed regions on the left to larger ones on the right.

7

This can happen by Ostwald ripening where domains at a greater radius than some critical value grow at a faster rate by diffusion from the surrounding medium. It can also happen by coalescence of droplets or by hydrodynamic effects [10]. The example shown has near equal amounts of each material but, when the two materials are there in different proportions, droplets of one material can exist in a matrix of the other material. Concentration changes encountered by addition of one material or crystallisation can lead to a phase inversion where the dominant body can become the droplets in the other. The two materials are in a metastable situation if they are quenched to a position on the phase diagram between the binodal and the spinodal. Statistical density fluctuations often lead to phase separation by nucleation and growth when the mixture is in the metastable region. 1.2.1.5

Solid Solutions

Solids can also form solutions in the same way that liquids do.

It is the

ability of the solids to mix in all proportions of the basic materials that is the criteria for a solid solution. In this case, though, the phase of the solution is a solid rather than a liquid. An example of this is copper with gold. 1.2.1.6

Eutectics and Monotectics

There are many cases where solids are not significantly soluble in each other but the liquids become soluble when the temperature is raised sufficiently. One example of this is common solder used in electronics where a eutectic is formed.

Eutectic is Greek for “easy melting”.

A

eutectic reaction is defined [11] to be: “An isothermal, reversible reaction between two (or more) solid phases during the heating of a system, as a result of which a single liquid phase is produced.” In these cases the phase diagram is similar to Figure 1-6. We see six regions in the figure. The first has both together as a solution (in the melt). The second and third regions α and β are solid solutions with one or the other of the materials in virtually pure solid form with a small amount of the other material dissolved in it up to the solubility limit. The fourth and fifth have one of α or β as excess solid in equilibrium with the solution. The sixth is where materials α and β are together in solid form. This last usually has a 8

finely divided matrix of α in β (or vice versa) with an overall concentration of the eutectic composition. Within that are larger domains of any excess α or β form. Consider solidified material after cooling from a molten mixture of materials A and B and having a concentration and temperature defined by point c in Figure 1-6. Material A is in excess so the solid will comprise nearly pure material A inclusions (of phase α) with the same composition as m solidified within a matrix of a solid eutectic mix of A and B (m and n with overall composition of e). The solid will reach d as it is heated. At that point, the eutectic portion will melt at the eutectic temperature (Te) into a liquid of the eutectic composition, leaving the inclusions of α in equilibrium with the liquid.

200

C)

j

180

liquid

Temperature (

0

160

h

g α

140

120

m

β+ liquid

α + liquid f

d

c

e

solid α + β

k

β

n

100 0

50

100

Percentage of Material B Figure 1-6 A simple phase diagram showing eutectic formation.

A further increase in temperature will see some of the solid inclusions changing composition along the line f-g as A dissolves into the liquid. This causes the composition of the liquid to move along the line e-h as the material α dissolves into it. The strong move of the liquid to the left with increasing temperature means that a considerable amount of α is dissolving into the liquid. Eventually the composition of the liquid will become the same as the original proportions of the two materials in the solid at the time the 9

last of the α phase of composition g dissolves into the liquid. Further heating maintains the composition at c-h and the liquid moves on the phase diagram in the direction of j. Consider the alternative of a solution having a concentration and temperature defined by point j in Figure 1-6. The state of the solution will reach h as it is cooled. At that point, material A will begin to crystallise out in nearly pure form as phase α with a composition given by point g across the tie line linking compositions in equilibrium at that temperature.

The

removal of phase α by crystallisation will naturally increase the relative concentration of B in the solution as the temperature is lowered slightly. The solution will thus follow the curved line towards point e as the temperature is lowered further with continuing crystallisation of phase α. The material crystallising will vary slightly in composition following the line g-f.

The lowest temperature where the liquid can coexist with solid is at

point e.

The liquid cannot exist below the eutectic temperature so the

remaining liquid (of the eutectic composition) will crystallise at that point in the cooling process. The final solid will incorporate solid, nearly pure A in a matrix. That matrix is phase separated A-rich and B-rich micro domains overall having the eutectic composition. On average, the solid will obviously have the composition proportions of the two materials in the original solution. The above descriptions for the phase diagram are for equilibrium at all times. Heating under practical conditions may mean that the composition of the α inclusions may not have time to change from f to that of g. There may be kinetic delays meaning that at faster heating rates some steps take place a little later (at higher temperatures).

The situation is more complicated

where we start with two powders A and B placed together. The powders will reach the eutectic temperature where the points of contact between the two types of powder will start to dissolve those grains of the powders. This will continue until all of B powder is consumed, leaving pure powder A (in this case) in the liquid of the eutectic composition. Further increases above the eutectic temperature will result in the dissolution of powder A into the liquid, moving the composition of the liquid along line

e-h, as previously.

The

practical implementation of eutectic formation from powders may result in further delays than when starting with an α-in-eutectic solid. We will see 10

later

that

polymers

often partly crystallise

forming nanometre-thick

crystallites that tend to exclude other molecules. It may be expected that the melting of a previously solidified mixture incorporating a semicrystalline polymer will behave in a fashion intermediate between two powders melting and that for eutectic mixes of small molecules or metals. The phase diagram seen in Figure 1-6 is one with a simple eutectic relationship between the materials. Many, more complicated, types of phase diagrams are found in practice with various material combinations.

A

simpler phase diagram occurs with the side regions disappearing when the solubility of one solid material in the other is totally insignificant. A monotectic reaction is similar to a eutectic reaction but here a solid and a liquid solidify (reversibly) from monotectic liquid. The compositions of the solid and liquid are both different from that of the originating liquid. It is possible to have multiple very small regions of liquid dispersed within a solid matrix as a result of a monotectic reaction. Those liquid domains can then solidify at lower temperatures. IUPAC [11] define a monotectic reaction as: “The reversible transition, on cooling, of a liquid to a mixture of a second liquid and a solid.” 1.2.1.7 Liquid

Thermodynamics of Liquid Mixtures mixtures,

like

other

systems,

are

characterised

by

normal

thermodynamic parameters such as Gibbs free energy G, internal energy U, enthalpy H, entropy S and volume V. The values found for real mixtures are not the sums of the values of the pure constituents. For example, mixing a volume of one liquid with an equal amount of another liquid will not give exactly twice the volume of the first material. The same applies to the Gibbs free energy. The difference between the actual free energy G and the sum of the Gibbs free energies of the pure components Gi mixing ∆Gmix.

is the free energy of

Similar comments apply to entropy and enthalpy with the

convention that the value for the mixture takes on the sign of the subtraction of the sum of the components from the value for the real system. This means. ∆Ymix

=

Y - (Y1 + Y2 +…+Yn), where Y is a thermodynamic

parameter and the values of Y with subscripts are those for the n pure materials in the system. 11

There is a partial molar property for any of the above in a system defined as the partial derivative of that property with respect to the number of moles of one constituent when temperature, pressure and the number of moles of all the other components are kept constant. The partial molar Gibbs free energy is the same as a parameter called the chemical potential of that constituent. It is usually given the symbol µi. These chemical potentials of the constituents are the quantities that determine phase equilibria and from them we can derive a range of other parameters.

It is worth mentioning here that the chemical potential of a

pure material will be the same as the molar Gibbs free energy of the material at the particular temperature and pressure of interest, ie. µi0 = Gi0. This discussion is general to mixtures of liquids but will play a part in the discussion later of the Flory-Huggins theory as it relates to polymers in solutions. 1.2.1.8

Kinetics of Phase Transformations

Most transformations from one state into another do not take place instantaneously because of impediments to the changes.

Often energy

barriers related to the phase boundaries have to be overcome for the molecules to be able to re-arrange themselves. There is usually a nucleation stage followed by a growth stage. The whole process is time dependent. The nucleation is the formation of stable microscopic particles of the new phase in the originating phase. This is followed by the growth of new material onto the nuclei. The growth of the new material proceeds by diffusion into the old phase. It occurs until all the volumes of new phase impinge on each other making the system wholly the new phase. The time taken for the change to take place is termed kinetics and is obviously important for production processes. The rate at which the volume of material changes from one state into another will be dependent upon how much of the old state remains if we hold the temperature constant.

The

outcome is an “S” shaped curve Figure 1-7 below that is described by the Avrami equation in Equation 1-1 [12-14].

12

Percentage Phase Transition

100 80 60 40 20 0 0

40

80

120

160

200

Time (sec) Figure 1-7 A typical Avrami plot for extent of crystallisation taking place.

Y = 1-exp(-ktn)

Eqn (1-1)

where Y is the volume fraction of crystalline material formed by time t at constant temperature, and k is a variable dependent on temperature. The exponent n should be an integer between 1 and 4, depending on the model used, according to the original theory. Nowadays it is regarded as a variable to match the data. The Avrami approach is a simple one that has found application across a wide variety of phase transitions. It is often used for crystallisation of metals but is used for polymers [15, 16] as well. 1.2.2

Hydrogen Bonding

We will now look at the hydrogen bonding that plays a strong part in the behaviour of polyamides and the different types of bonding encountered in chemistry. Most people are familiar with ionic and covalent bonds. Ionic bonds take place with the complete transfer of electrons from one atom to another and are very strong. Covalent bonds are directed between two atoms such as C-C or C-N and are also strong. Covalent bonds have energies in the order of 300kJ/mol. The much weaker van der Waals forces are of the order of 1kJ/mol and are non-specific in direction. Hydrogen bonds are intermediate in strength (around 30kJ/mol) and act between hydrogen atoms and two other electronegative atoms, usually from the group oxygen, nitrogen and the halides, particularly fluorine. They are essentially electrostatic in nature. They act when hydrogen has been covalently bonded to one of the above highly electronegative group that has 13

drawn some of the charge from the hydrogen atom.

This makes the

hydrogen atom partially positive in charge. An atom in another molecule or another part of the same molecule that is also electronegative will be weakly attracted to the hydrogen atom, forming a hydrogen bond (or hydrogen bridge). Probably the most important case of hydrogen bond formation is with water where an O-H from one water molecule is bonded to the O of another water molecule.

Hydrogen bonds are continually forming and reforming, even in

water near 100 0C.

The reason the water in our own bodies does not

evaporate at sub-zero temperatures is the hydrogen bonding that provides an energy barrier to evaporation. It is also hydrogen bonding forces that link the peptide groups of DNA into a double helix.

We will see later that

hydrogen bond formation is implicit in the physical properties of the polyamides (Nylons) of this research. 1.2.3

Small organic molecules

1.2.3.1


Small organic molecules in the solid state are arranged in a very regular, symmetrical manner, held in place by van der Waals and possibly hydrogen bonding forces. These forces will be stronger if the molecules can fit closely together with as many atoms of one molecule as close as possible to atoms of the next.

The covalent forces holding each molecule together are much

higher than the inter-molecular forces. The atoms gain in vibrational and rotational energy as the temperature is raised until the structure breaks down suddenly and a disordered liquid state results. The reverse occurs as the temperature is lowered and the molecules can nestle together in an ordered structure. A better physical fit between the molecules will result in stronger van der Waals forces between the molecules and a higher melting temperature. 1.2.4

Polymers

The Greek word poly means many and the word meros means part. Polymers are macromolecules (very large molecules) made from the combination of a large number of smaller repeating molecular units (monomers) to create a larger molecule. They occur naturally as in DNA or silk and can be manufactured synthetically as in Nylons and cured epoxy. This research covers synthetic homopolymers made from only one sort of 14

repeat unit unlike copolymers where there are polymer sections of differing types within the one molecular chain.

They can be structured as long

chains, with or without sidechains, networks, or dendrimers. This work is on long linear polymers without sidechains. 1.2.4.1

Thermoplastics and Thermosets

Polymers are of two types, thermoplastic and thermosetting. Thermosetting polymers result from the in-situ reaction of smaller molecules where covalent crosslinkages form between the molecules during polymerisation, resulting in a network. Subsequent reheating of the solid will not allow the material to liquefy once the polymerisation reaction has been performed. Eventually degradation occurs with the application of further heat.

Thermoplastic

polymers can undergo multiple cycles of the polymer softening and becoming liquid with heat and hardening on cooling.

The polymer chains gain

sufficient vibrational energy to break the weak van der Waals forces between the polymer molecules during this reversible process although this usually takes longer than with small molecules because of the greater number of molecular interactions involved with these large molecules.

This type of

polymer can be processed in the melt by injection moulding, casting, blow moulding and spinning to form solids of the required form on cooling. The research here is on thermoplastic polymers. 1.2.4.2

Basics including molecular weight and molecular shape

The number of repeat units in a polymer molecule affects the size of the total molecule in solution or the melt.

This, in turn, affects the viscosity in

solution and the melting temperature.

Commercial polymers often have

molecular weights of 20 to 50 kilodaltons. Reactions to make polymers from monomeric units normally do not mean that all molecules form at exactly the same molecular weight.

There is

usually a distribution of molecular weights from a polymerisation process. This means that the physical properties of a polymer such as viscosity are an average over all chain lengths represented in the sample. The degree of polymerisation, n, will be a distribution with a number average, Mn, that is less than the weight average, Mw.

The ratio of the two is called the

polydispersity and is a measure of the broadness of the molecular weight distribution. 15

Polymer chains have the opportunity to become entangled, spaghetti-like, when in solution or the melt if they are sufficiently long.

Viscosity is

increased markedly once entanglement has set in and the dynamics of crystallisation are also altered [17, 18]. The number of repeat units before entanglement becomes an issue is different between polymers and depends on whether the polymer is a plain linear chain or has side chains, and on other characteristics of the repeat units. The distance between atoms of different polymer chains is a balance between attractive van der Waals forces and Born repulsion between the clouds of electrons surrounding each atom.

The bond lengths between covalently

bonded atoms in the one molecular chain attempt to remain at their equilibrium distances. optimum angles.

At the same time, the bonds try to stay at their

Every atom in a polymer chain attempts to find an

energetically favourable position for itself under the constraints of bond angles and interatomic distances. The application of more heat to a system results in greater vibration of atoms around their optimum positions. We will see later that the multiple forces acting on an atom can be utilised in Fourier Transform Infra Red techniques to characterise the environments of atoms by their frequencies of vibration. Polymer chain molecules are not straight. Usually the bonds are at preferred angles other than 1800 and, unless sterically hindered by some of the atoms, are able to rotate when in the melt or in solution. This results in a threedimensional “random walk” if the path in space is followed from one atom to the next as displayed in two dimensions in Figure 1-8.

The length of the

chain from one end to the other can be seen to be much greater than the end-to-end distance along the straight line A-B. The size of the molecule can be characterised statistically for a given situation with the radius of gyration as a measure of how large the molecule is. That is determined by the mass average of the square root of the squares of distances of the atoms from the centre of mass of the molecule. A flexible molecule that is in thermal motion requires that a time average be taken over all configurations. The thermal vibrations in a melt at high temperature will tend to result in a larger radius of gyration. The size of a polymer chain in a poor solvent will be much smaller than with a good solvent because the chain segments tend to keep to similar chemical environments. They retract to a smaller volume 16

to exclude unfavourable solvent molecule interactions. This also occurs with proteins in an aqueous environment where the hydrophobic portions bury themselves at the centre away from the solvent and the hydrophilic portions extend into their watery surrounds.

Figure 1-8 Random walk between A and B, the ends of the polymer chain.

A linear polymer chain, such as with the Nylons of this text, will have a larger radius of gyration than one of the same molecular weight but with branches emanating from the main chain. 1.2.4.3

Amorphous polymers

Polymers can solidify into the amorphous state where, firstly, the longitudinal motion is locked in, making the polymer solid but rubbery. When the material reaches the glass transition temperature (Tg), it becomes like a glassy, supercooled liquid as it is cooled further.

Here, the motion

allowed by molecular vibrations and rotations becomes much more restricted.

Thermal conductivity, dielectric constants and mechanical

properties undergo significant changes at similar (but usually not quite identical) temperatures. The glass transition temperature (Tg) for a particular parameter is the point at which half the step change has taken place. Polymers are very stiff below their Tg but they become rubbery as the temperature is raised again.

Eventually they become viscous liquids that

quickly thin further as the temperature is progressively raised. The process is completely reversible unless the temperature has been raised so high that the polymer has degraded. An example of the viscosity behaviour as temperature is raised for an amorphous polymer is shown in Figure 1-9. 17

Viscosity

|

Tg

Temperature

Figure 1-9 Typical effect of temperature on viscosity for amorphous thermoplastic polymers. Viscosity is extremely high below the Glass Transition temperature Tg and rapidly drops with increasing temperature.

The hole theory of liquids requires there to be minute voids between the molecules that allow them to move from one position to another.

If we

extend this to polymers we must recognise that the molecules of a polymer chain have to move cooperatively.

This requires a minimum void size to

allow chain segments to move from one location to another. The free volume will increase rapidly with temperature above this critical temperature. The free volume of a polymer will remain relatively constant below this temperature as molecular motion is frozen. 1.2.4.4

Polymer crystallinity

The discussion in this and the following four sections is only meant to provide a general background for discussion of various aspects of polymer crystallinity in the main chapters dealing with specific polyamide-diluent combinations. Over fifty years ago it was found that approximately three quarters of different types of polymers are able to also enter a partially crystalline state. This can happen for those types of polymers if the cooling conditions are slower, or if solid amorphous polymer is taken through an appropriate thermal history that allows a solid-state crystallisation to take place. The degree of crystallinity will depend on the thermal and mechanical history of the sample and can range from zero to 90%. It is this crystallisation that adds to the mechanical stability of many manufactured plastic articles. Generally, as the cooling rate during crystallisation increases, the percentage in the amorphous state increases and crystallinity decreases. Polyethylene 18

and polyethylene oxide tend to have very high crystallinity and others have much less, ranging down to those that cannot be crystallised. Polyamide-4,6 being studied here can crystallise up to 70% by volume under favourable conditions. Lamellae are crystalline regions within the overall amorphous polymer. The order caused by polymer chains (and parts of polymer chains) aligning themselves means regularity in the structure of atoms, allowing Bragg reflections to be seen with X-Rays in the manner seen with crystalline mineralisations. Whether a crystallisable polymer solidifies purely with an amorphous structure or with a certain extent of crystallisation will depend upon a range of parameters that can also affect the crystallographic form.

The thermal

history as well as the molecular weight will also play a strong part. A final crystalline state, probably metastable, will depend on nucleation states and entropic barriers. Other factors that can affect the way in which the final crystal form develops are the degree of undercooling, recrystallisation and the lamellar thickening or thinning mentioned below. The crystallisation takes the form of lamellae (platelet like structures or crystallites several micrometres across and 5 to 10 nm thick) as seen in Figure 1-10.

Figure 1-10 Keller’s diagram [19] for the laying down of folded polymer chains along the edge faces of lamellae

The larger, flat surfaces are called the basal planes and the thin surfaces along the edges (and joining the basal planes) are called edge faces. Long polymer chains are generally considered [19] to fold backwards and forwards into place across the edge faces of lamellae crystallising from the melt or solution.

These energetically unfavourable assemblies come about mainly

because of the attempts of long polymer molecules to rearrange themselves 19

into energetically more favourable structures with greater order. They are, of course, restricted in their ability to “reptate” like snakes into ideal positions in reasonable timescales. Kinetics plays an important role in the perfection attained. Shorter polymer chains can crystallise in an extended chain form where molecules line up together, side by side, without chain folding.

This is a

lower energy configuration because there are no folds necessary.

For

example, PEG usually forms folded chains in the lamellae only when the molecular weight exceeds approximately 4,000 Daltons [20, 21].

The

extended chains become difficult to lay side by side when they are too long. Layers are added on the edge faces to build up the thin lamellae from their centre with chain folds in a consistent manner dependent upon steric effects with the atoms and energy minimisation. The folding is driven by kinetic factors because the initial nucleus has polymer chains locked into the folded form. The lamellae formed in solution are usually more perfect than those formed in the melt because there is more opportunity for polymer chains to easily orient themselves correctly by displacing the smaller solvent molecules. Later growth of the lamellae by secondary nucleation of chains onto the edge faces of the crystals continues the original folding form but often the thickness of the lamellae vary as they grow bigger.

It is common for

polymer lamellae to thicken if they are later annealed for some time at temperatures somewhat below the melting temperature.

That occurs by

polymer chains reptating like snakes in the lamellae to produce the more thermodynamically stable thicker configuration. The ends of molecules withdraw from their initial place in the lamella during the reptation process. It has been found [22] that lamellar thinning also occurred with some semirigid polymers, including polyamides. The lamellae can often form in different crystallographic structures. Generally the lamellar thickness increases with molecular weight and the melting temperature also increases. The

lamellar

thickness

crystallisation in polymers.

generally

depends

on

the

temperature

of

The thickness of a lamella is reduced when

crystallisation takes place at a larger undercooling.

The lamellae do not

have time to form in thicker, more energetically favourable forms when 20

driven by high supercooling. The significant surface energy tied up in thin lamellae makes them less stable, in general, than thicker ones. This leads to the melting temperature of thinner lamellae being reduced below that of an infinitely large crystal. The melting temperature approaches an asymptotic value as the molecular weight tends towards infinity. Gibbs and Duhem were able to relate the equilibrium melting temperature, Tm0 to the measured melting temperature Tm using the lamellar thickness l, the enthalpy of fusion ∆Hf and σe from the slope of the graph of Tm against 1/l. Hoffman and Weeks [23, 24] took this further by eliminating the need to know the lamellar thickness with the use of plots of Tm against the crystallisation temperature Tx under the same ramp rates. Measured values for the pair are extrapolated to the line Tm = Tx and that gives the equilibrium value Tm0 . The Hoffman-Weeks approach has been extended with linear and non-linear extrapolations by Marand, Xu and Srinivas [25]. On the other hand Welch and Muthukumar [26] believe that a reliable estimate of equilibrium melting temperatures cannot be obtained by this method. We must consider two aspects for crystallisation to take place, the nucleation of lamellae formation and the kinetics of lamellar growth. There is the primary nucleation of lamellae as a first stage and then and then the growth stage with secondary nucleation. The primary nucleation can be likened to atoms in a gas condensing with lowered temperature to form small groups with a high surface area to volume ratio. This situation is unfavourable with much energy tied up in the surface compared to the free energy gain by condensation to a liquid. The group will dissociate unless sufficient atoms can simultaneously coalesce to make a nucleus that can grow.

That is because the increase in volume

produces a lower energy than the increase in surface area. relies on statistical fluctuations at high temperature.

The process

These intermediate

groups then produce nuclei that can grow further. The formation of lamellar nuclei in molten polymer or in solution, are generally regarded as analogous to the gas-liquid condensation described above. An embryo lamella accretes and loses adjoining sections of polymer chain in a dynamic manner until it is large enough that the free energy gains outweigh the surface energy effects. 21

It is able to become a lamellar nucleus for further growth. This is a type of situation with a hurdle to overcome where the Avrami approach is applicable. The major theory or model put forward to cover the secondary nucleation and growth of lamellae once they have nucleated and begun to grow is that of Hoffman and Lauritzen [27, 28]. In that theory there is assumed to be a flat existing substrate. A new polymer chain lays down on the flat surface of the edge face beside the previous chain and becomes attached to both.

The

question then arises as to when it folds. An attempt to maximise the contact area and bonding for the new chain will lead to longer lengths between folds and result in thicker lamellae. The necessity to achieve this quickly when there are strong driving forces towards crystallisation means a shorter length between folds is desirable. The final lamellar thickness is thus dependent upon the interplay between these opposing tendencies. The outcome is that larger undercoolings result in thinner lamellae. Often crystallisation also takes place more slowly behind the main crystallisation front on the lamellae. This is called secondary crystallisation and results in increased densification of the solid, particularly because the slower rate of crystallisation results in more perfect (and denser) crystalline regions.

Diluent in the melt blend systems studied here will mean that

secondary crystallisation effects for the polymers will be promoted due to dilution

but

be

reduced

polymer-diluent material.

by

lower

viscosity

of

the

uncrystallised

The outcomes cannot be predicted and would

require other techniques such as time resolved SAXS measurements to determine. An analogous situation can be seen to occur for the diluent with mini

spikes

occasionally

being

seen

well

after

the

main

diluent

crystallisation peak such as with 65PA46Car in Figure 3-14 where there is a small diluent crystallisation spike 30 0C lower than the main carbazole crystallisation peak. Androsch and Wunderlich [29] showed with studies on poly(ethylene-co-octene) using Temperature Modulated Differential Scanning Calorimetry (TMDSC) that secondary crystallisation occurred with a delay of 5 min after primary crystallisation when cooling at 10 0C/min. There is, however, much controversy over the steps that take place in going through from the supercooled melt to the formation of crystals. 22

Olmsted et al. [30] suggest that there is liquid-liquid spinodal decomposition taking place prior to the formation of crystals. The spinodal decomposition was detected by Small Angle X-ray Scattering (SAXS).The actual crystal formation was detected by Wide Angle X-ray Diffraction (WAXD). They base this in part on work by others eg Ezquerra [31] on a variety of polymers where SAXS signals are seen to increase and partially decay before the WAXD signal appears in simultaneous SAXS-WAXD.

They propose that

there are statistical fluctuations in density and entropy (linked) which result in spinodal decomposition of the molten polymer into denser and less dense phases. It is the more ordered, dense regions (detected by SAXS) that later crystallise into the lamellae detected by WAXD.

The direct experimental

results are supported by Monte Carlo simulations by Toma, Toma and Subirana [32] where they investigate the formation of a compact globule state with a lamellar conformation prior to the creation of a crystal. More recently Jiang et al. [33] and other groups have found similar precursor activity with other polymers examined with Fourier Transform Infra-Red spectroscopy (FTIR) in situ during crystallisation and there are some parallels in the recent work of Rabani, Reichman, Geissler and Brus [34] on the formation of nanoparticle structures during drying. Welch and Muthukumar [26] suggest entropic barriers are involved initially, that chains attach themselves to the growing crystal in line with the existing chains and that lamellar thickening takes place at a later stage in a cooperative fashion. Wurm

and

Schick

[35]

heated

poly(ε-caprolactone)

and

syndiotactic

poly(propylene) with small laser pulses and presented evidence towards a model with crystallisation first taking place by a partially ordered metastable structure in the melt that becomes progressively more ordered into a lamellar crystal as it undergoes a stabilisation stage. Doye and Frenkel [36] disagree with the Hoffman-Lauritzen theory as to how it predicts lamellar thickness and attempt to improve aspects of the Sadler-Gilmer approach which is based on entropic barriers. There are a number of other theories that also try to overcome the weaknesses of the Hoffman-Lauritzen theory that was a major step forwards over forty years ago. It is the Olmsted et al. approach that we will later see 23

supported in one aspect of the research being presented in this thesis. That aspect is the minor phase separation seen in certain cases during crystallisation and later re-melting. The picture of lamellar structure and formation presented by Keller, Lauritzen and Hoffman and others is an ideal one. In practice the loops of the chain folds can either be close ones or they can re-enter the lamella some distance away as in a “telephone switchboard” model. Often a group of polymer chains will cluster together to form a lamella but some of the sections of some chains will also be incorporated in other lamellae.

The intervening sections will meander through the amorphous

region between the lamellae. A high undercooling below the melting temperature, either by rapid dropping of temperature to a desired isothermal crystallisation temperature or by fast dynamic cooling, leads to strong driving forces for crystallisation. The short crystallisation period, which results from the strong driving force, does not allow as much time for polymer chains to reorganise themselves into favourable configurations. A far from ideal structure is then locked in. The kinetics of crystallisation is strongly affected by the cooling rate, as mentioned above. The crystallisation temperature is lowered as the cooling rate is increased. Usually the amount of material that becomes crystalline is reduced and the speed of crystallisation increases [37] with faster cooling, resulting in a highly amorphous solid if the material is quenched, for example, into ice water or liquid nitrogen. It has been explained above that the question of crystallisation kinetics is a complex one, even in the situation where crystallisation takes place isothermally.

There

are

good

reasons

for

also

wanting

to

study

crystallisation in a non-isothermal or “dynamic” context. That is the way crystallisation takes place in almost every production environment so an understanding of the processes under conditions emulating real life is also needed [38]. There is also another consideration when crystallising a range of blends where there may be multiple crystallisations.

The temperature(s) for

isothermal crystallisation have to be chosen specifically for each particular blend [6] and a wrong choice will obscure the information being sought. It 24

would be difficult to obtain meaningful results from a comparison between a number of blended materials where the melting temperatures of the constituents differ markedly.

This is particularly so when differing

compositions of any pair being blended could lead to differing crystallisation temperature depressions. A practical solution is using non-isothermal crystallisation

so that

crystallisation takes place when the molecules are ready for crystallisation at the cooling rate used. A faster cooling ramp leads to a greater undercooling before the crystallisation takes place because of kinetic factors.

The

crystallisation does actually take place at near isothermal conditions because the self-generated temperature field from the latent heat of crystallisation does tend to maintain the local temperature in a pseudoisothermal condition during the crystallisation.

Some authors utilise this

self generated pseudo-isothermal crystallisation in their own manner to achieve specific undercoolings that would otherwise be difficult to achieve [39].

It is not an ideal situation from a theoretical perspective but does

provide a practical way of solving the conundrum in those cases. 1.2.4.5

Lamellar melting

The melting of lamellae of monodisperse polymers takes place over a greater temperature range than for small organic molecules or metals.

This is

because the long polymer chains have to reorganise themselves and dissociate themselves into the melt from their places attached to the side faces of the lamellae.

That process takes time and is at least partly

sequential with one layer being removed before the next one can also escape into the melt. The process of melting (some) individual chains from a number of lamellae has been detected with polymer chains partly disengaging themselves into the melt and recrystallising those sections onto the lamellae [40]. This was carried out by using quasi-isothermal TMDSC using a very small amplitude of the temperature oscillations.

The reason for the small oscillations in

temperature was to keep the chains partly tethered to the lamellae so that there was no barrier to recrystallising back onto the same lamellae. The energy barrier to re-nucleation just referred to is a contributor to the substantial difference in melting and crystallisation temperatures found for polymers.

It is necessary to have a substantial undercooling of perhaps 25

10-30 0C before the formation of lamellae takes place. That is the case even when there are nucleation sites present from nucleating agents that have been added to the polymer, or because the original melting of crystalline regions was incomplete. This is discussed below. Different

lengths

of

polymer

chain

have

slightly

different

melting

temperatures with higher molecular weight polymers having higher melting temperatures than lower molecular weight polymers. The influence of this is much greater at low molecular weights and negligible at high molecular weights where the predominant factors are lamellar thickness and crystal perfection.

For example, Smith and St.John Manley [41] point to quasi-

monodisperse fractions of polyethylene with Mw = 1000 having a melting temperature of 105 0C, that rising to 121 0C for Mw = 2000 but only rising further to 131 0C for a molecular weight of 20,000. The same relationship applies for crystallisation. Polydisperse polymers, as found in the real world, therefore have wider melting and crystallisation temperature ranges than monodisperse polymers and far wider than for small organic molecules. 1.2.4.6

Melting behaviour of semicrystalline polymers

We have now looked at both completely amorphous polymers and the melting and crystallisation of lamellae.

Even polymers in a very highly

crystalline state have 10% or more of amorphous material incorporated between the lamellae and many semicrystalline polymers are 50-80% amorphous. The glassy amorphous material will become rubbery at the glass transition temperature as the temperature is raised. The viscosity of a semicrystalline polymer will be higher than with a fully amorphous one because the lamellae act as relatively inert platelets restricting motion.

Eventually the

temperature reaches the lamellar melting temperature and the polymer segments that had been in the lamellae peel off the lamellae, becoming indistinguishable from those that had been in the amorphous part. There will be some drop in overall viscosity at the melting temperature (Tm) to that of the amorphous material at that temperature as the melting lamellae cease to restrict molecular motion in general. 1.2.4.7

Spherulites

Spherulites and other larger scale structures are made up of lamellae. The spherulites are lamellar structures that have grown, splayed out and twisted 26

to create spherical forms. They appear as Maltese crosses under crossed polarisation illumination. crystallised from the melt.

They are particularly important in polymers As early as 1888 Lehmann had made the

conclusion that they formed from long crystals that had forked as they grew and spread out to fill up the space until they impinged on other growing spherulites or until growth had stopped. The early observations of “twisted crystals” in the 1920s and 1930s were for relatively small natural macromolecules but by the mid to late 1940s they were being recognised in polyethylene.

Bryant identified that long polymer chains could partake in

multiple lamellae within a spherulite. That connection of the chains between lamellae means that there is a coupled growth front for the spherulite as a whole with the stacks of lamellae having amorphous material in between. Other structures of lamellae that are encountered are axialites and hedrites where crystals are attached to a common axis.

Axialites can occur in

crystallisation from the melt but will be observed differently depending on the direction of the axis relative to the observation direction. There are also fibrillar structures encountered with oriented growth under stress such as polymer fibre formation. 1.2.4.8

Poorly and partially crystallising polymer types

Aromatic groups on the main polymer backbone will have difficulty in crystallising into lamellae because of steric hindrance. Polymers with long branches will encounter difficulties in having the chains folding side by side in lamellae because the side chains will get in the way sterically.

Some

branched types will not crystallise in the main backbone but long side branches may form lamellae.

Copolymers often have one section that

crystallises and another part that does not. 1.2.4.9

Polymer-polymer miscibility

Often we want to utilise the strong points of two polymers to produce a better material. The problem is that different types of polymers are usually immiscible. There are a number of ways this can be overcome. One is to synthesise block copolymers that have long chain segments of each polymer. The synthesis in production quantities is often expensive.

A number of

alternative approaches are possible, such as to use bulk quantities of each polymer with a smaller amount of the relatively expensive copolymer to tie phase separated domains of each type together as a compatibliser. Another 27

is to use maleic anhydride to reactively compatibilise the two. An alternative method is to have two materials that can hydrogen bond together and use this to compatibilise the materials as described by Huang et al. [42]. 1.2.4.10

Polymer-diluent systems

The Flory-Huggins theory is an attempt to determine the ∆Gmix for polymer solutions.

Flory [43] and Huggins [44-46] independently put forward a

theory that has been modified by others in a variety of ways. The approach is an extension of an earlier one by van Laar in which he treated two types (1 and 2) of equally sized molecules in an ideal solution as occupying the sites of a three dimensional lattice. He then predicted the ∆Gmix as a function of the universal gas constant, absolute temperature, numbers of moles of each and the mole fractions of each material. That approach (which can be used to derive Raoult’s law) was a failure for the case of polymer solutions. It was extended by Flory and Huggins with the restraint that the segments of polymer molecules within the solution are interconnected.

It utilised an

interaction parameter Χ12 between the polymer and solvent molecules and led to the ability to predict melting temperature depressions and phase diagrams. Flory [47 p.569] shows that RVu 1 1 (v1 − X 12 v12 ) − 0 = Tm Tm ∆H uV1

Eqn. (1-2 )

where Tm is the equilibrium melting temperature of the mixture, Tm0 the equilibrium melting temperature of the pure polymer, R the Universal Gas Constant, v1 the volume fraction of the diluent, ∆Hu is the heat of fusion of the repeat unit, V1 and V2 the molar volumes of the diluent and unit respectively and X1, the interaction parameter. The assumptions made in the theory are that there is no volume change upon mixing, interactions of the different types of segments cause the enthalpy of mixing after same type interactions have been replaced, polymer repeat units and solvent molecules are the same size and the number of combinations of polymer configuration solely determines the entropy of mixing. The Flory-Huggins lattice model, mean field theory above and its large number of variants is suited to describing melting point depressions, plasticisation

and

liquid-liquid

(L-L) phase 28

separations

but

not

for

liquid-solid (L-S) phase transitions as taking place during crystallisation, as discussed by Hu, Frenkel and Mathot [49].

No models to derive the

diluent-polymer interaction parameter Χ12 from linear relationships between crystallisation temperature depressions and concentration have been located in the literature.

A number of the materials combinations studied in the

thesis do exhibit linear relationships between Tc and concentration for both polyamide and for diluent.

Calculations based on the above equation are

carried out in those chapters where melting depressions were found to occur with a linear relation between melting point and concentration found. Kelley and Bueche [50] use the additivity of free volume for the pure materials in a miscible blend to determine the glass transition temperature of a blend.

Their calculations lead to good predictions of the Tg versus

composition curve and to the so-called Kelley-Bueche line delineating vitrified and unvitrified blend regions.

The intersection of the vitrification

and cloudpoint curves (due to phase separation) is the Berghmans Point. 1.2.5 1.2.5.1

Linear polyamides (Nylons) History of polyamides

A linear polyamide was polymerised inadvertently at the end of last century but was not recognised as being a polymer. A gelatinous mass had resulted from experiments with amino carboxylic acids. Prior to the 1920’s organic chemists failed to recognise the importance of polymeric materials, concentrating their efforts on producing monomolecular weight compounds. During the 1920’s, Staudinger recognised the existence of polymeric material by relating solution viscosity to molecular weight. Wallace H. Carothers was a brilliant organic chemist and in 1928 was employed by DuPont to carry out research. He elected to continue in the polymeric field opened up by Staudinger. 1929 was a period where great controversy still existed as to whether polymers were long chain molecules, colloids, or aggregates of cyclic compounds. At about this time Carothers [51] wrote a short review that for the first time clearly identified the two main polymerisation reactions that we now know as “chain growth” (addition) polymerisation and

“step growth”

(condensation) polymerisation. He identified, in this review, that molecules containing an amino group and a carboxylic acid group could condense to form polymers such as polyamide-6. He also suggested that it may be 29

possible to condense diamine compounds with dicarboxylic acid compounds to form polymers such as polyamide-6,6. In the early 1930s, when linear polyamide-6 was being synthesised with caprolactam, he proceeded from polycondensation

with

ε-aminocaproic

acid

to

the

synthesis

using

hexamethyl diamine and adipic acid. By 1939, the US approach had developed to the extent that there were plants set up to commercially produce polyamide-6, which by then had acquired the commercial name Nylon.

This commercial activity was soon

subsumed by the war efforts. Germany pursued the investigation of optimal polyamide types using the amino acid condensation and did not produce polyamides commercially until after the war. Nylons were one of the early polymers developed commercially. Nowadays, they are manufactured industrially for a broad range of applications such as clothing, stockings, carpets, fishing lines, tyre reinforcers, seat belts, and in the components of a wide range of appliances and equipment. The fibre component alone of linear polyamide worldwide production is in the order of 4 million tonnes per annum.

This comprises nearly a quarter of total

synthetic fibre production, as noted by Elias [52]. Later developments lead to polyamides made with aromatic groups in the main

chain

(called

Aramids),

branched

polyamides

and

copolymers

incorporating polyamides in various forms. These later types are not covered in this research work and they represent a much smaller volume of commercial production. 1.2.5.2

Strengths

Polyamides are tough, impact resistant, flexible, abrasion resistant, heat stable materials [53, 54] whose characteristic physical properties are mainly determined as a result of hydrogen bonding.

There are a range of

polyamides with varying properties dependant upon molecular structure of the monomer repeat units. Some of the newer polyamides such as polyamide-4,6 [55] have very high melting temperatures, and mechanical stability that allow them to be used in automotive applications near the engine. This particular polyamide has the fast crystallisation that makes it attractive for injection moulders.

30

1.2.5.3

Weaknesses

Humidity plasticises and weakens polyamides.

Polyamides also become

brittle when dry. Both these characteristics result from hydrogen bonding. 1.2.5.4 Linear

Chemical structure and polyamide types polyamides

have

a

main

chain

with

repeated

amide

units

incorporating -CONH- sections as shown in Figure 1-12. The amide unit is always trans across the polymer backbone although it can sometimes be partly twisted. O

II –C–C–N–C–CI H Figure 1-12 CONH amide units found in polyamides showing the trans configuration of the bonds.

There are two basic types of linear polyamides, the polyamide-n type and the polyamide-m,n type where the m and n are numbers representing the number of carbon atoms in (parts of) the polymer repeat units. Polyamide-n types, with n carbon atoms per repeat unit, can be formed by condensation from amino acids such as in Carothers’ earlier work. Only one material is used as the monomeric substance.

An example is the ring

opening of caprolactam with its 6 carbon atoms and one nitrogen atom in a ring. The opened ring is polymerised end to end into long chains forming polyamide-6. Water is a by-product of the high temperature polymerisation reaction and is pumped away to drive the reaction forward. The polyamide-m,n types are obtained by the polycondensation of a diamine and a dicarboxylic acid (or diacid). The number of carbon atoms in the main chain due to the diamine gives the first number, m,

and the number of

carbon atoms in the diacid gives the second number, n.

For example,

hexamethyl diamine and adipic acid are used to synthesise polyamide-6,6. Sometimes the number is placed before the word polyamide and sometimes PA or Nylon is used. In some situations the name of the amino acid is used. It is common to see PA-6, PA6, Nylon-6, Nylon6, 6-Nylon, Polyamide 6 and poly(ε-caprolactam) for polyamide-6. There is an even greater variation of naming for the m,n (or mn) types. Sometimes the comma is left out with Nylon 612 meaning polyamide-6,12. The maximum length of polyamide-n 31

types is in the twenties and the maximum length of a polyamide-m,n is similar so there is usually little confusion in omitting the comma. When the numbers n or m+n are small then the repeat distance is shorter. These polyamides are often referred to as “short” or “lower” polyamides as distinct from “higher” polyamides.

This does not refer to the number of

repeat units in the total polymer chain length. It should be pointed out that polyamide-6 and polyamide-6,6 are quite different materials even though the density of amide bonds in a polymer chain is the same.

The melting temperature of polyamide-6 at 225 0C is

some 30 0C less than for polyamide-6,6. The reasons for this will become evident later. The simplified structure of polyamide-6 and polyamide-6,6 are shown below in Figure 1-13 with repeat units in bold font.

O O II II -C-C-C-C-C-C-N-C-C-C-C-C-C-NI I H H polyamide-6 O O II II - N-C-C-C-C-C-C-N-C-C-C-C-C-C– I I H H polyamide-6,6 Figure 1-13 Repeat units of the n type polyamide-6 with all amide groups in the same direction and m,n type polyamide-6,6 from diamine combined with diacid and having the amide groups in alternating directions.

Note that the amide group is asymmetric so that the polyamide-6 repeat unit fits head to tail along the molecule.

The molecule as a whole is

unidirectional. On the other hand, the polyamide-6,6 can be seen to have points of symmetry at the mid points of the amide and of the diacid groups. This is an important point and will be taken up later.

The “directional”

polyamide-n types will have antiparallel sections of molecules next to each other as the molecule loops back in a hairpin.

This happens as the

backward and forward laying of the molecule into place occurs on the lateral 32

faces of the lamellae. Sections of different molecules layered against them at later stages can be parallel or antiparallel in direction. It can also be seen that the polyamide-6 has a repeat length of 7 atoms in the backbone whereas the polyamide-6,6 has a repeat length of 14 atoms. Polyamide-6 and -6,6 are used for textiles because of their high tensile strength. Polyamide-6,10 and polyamide-11 have longer distance between amide groups are used for sutures and sporting goods requiring flexibility. 1.2.5.5

Biological-polyamide parallels

The biological fields touched on below are examples of, perhaps, the most exciting potential areas to which this research could contribute because the boundaries between biology and synthetic chemistry are breaking down and both disciplines are learning from each other. This can be seen in a recent review with over 160 references by Cunliffe, Pennadam and Alexander [56]. Linear polyamides are one of the most important natural polymers and are known by biochemists and biologists as proteins or polypeptides.

The

peptide linkage referred to by biologists is identical to the amide linkage that occurs in synthetic linear polyamides.

The molecular structure of

polyamide-2 forms a very simple model [57, 58] for a protein. Some of the parallels between polyamides and proteins can be pointed out.

A better

comprehension of polyamide crystallinity in different environments could potentially lead to improved understanding of the way in which proteins fold, recognised nowadays as a very important area of biology. Proteins can form “molten globules” before crystallising out fully [32, 58] and this concept may, in turn, be relevant to the way in which polyamides crystallise from the melt or solution, particularly in the light of the recent work of Olmsted et al. [30] on the formation of crystallites in molten polymers. 1.2.5.6

Polyamide Hydrogen Bonding

Polyamide crystallisation is more complicated than with many polymers because hydrogen bonding constrains the crystallographic possibilities further than just the steric considerations [59]. Hydrogen bonding, in general, was discussed earlier. It is now appropriate to look at hydrogen bonding specifically in polyamides. The nitrogen atoms in the amide sections are highly electronegative, withdrawing some of the charge from the attached hydrogen. Normally the oxygen from the carbonyl 33

bond in another amide group elsewhere in the polymer chain or from another molecule will be attracted to the hydrogen to form the N-H.…O hydrogen bond, as portrayed in Figure 1-14. O II –C–C–N–C– I H . . .

O II –C–C–N–C– I H Figure 1-14 Amide to amide hydrogen bonding found in polyamides showing the bridging from the electronegative oxygen of one amide group to the electron deficient hydrogen attached to the electronegative nitrogen atom of another amide group in the same polymer chain or another molecule

In general, there can be weak and strong hydrogen bonds. Those involved in polyamides are considered moderate to strong. These hydrogen bonds in polyamides are pervasive, being substantially consummated in the amorphous state and are even present at a significant level in the melt [60-62]. This makes the polyamides much more viscous in the 50 0C range above their melting temperature than many other polymers. They are the driving force that locks the crystallising lamella into one or another crystalline form. They are also the reason for the very high melting temperature of linear polyamides because they provide stability to the lamellar structures. Other molecules can be incorporated into the amorphous polyamide structure, such as water, which plasticises and weakens polyamides by displacing the hydrogen bonds. 1.2.5.7

Polyamide Crystallinity

Linear polyamide crystallinity is strongly affected by the exact type of linear polyamide because of the limited combinations of the way hydrogen bonds can be consummated within the constraints of the number of molecules between amide groups. The orientation of the non-symmetric amide groups in the chains also plays a strong role such as in the difference of 30 0C in melting temperatures of polyamide-6 and polyamide-6,6 referred to above. There are also steric limitations between the sections of molecular chains 34

lying next to each other and between different molecules in a lamella. These differences

between

polyamides

can

be

exploited

to

gain

a

better

understanding of polyamide crystallinity and the part hydrogen bonding plays in their properties. 1.2.5.8

Polyamide Crystalline Structures

This and most of the following few sections are included mainly to provide background understanding of the crystallographic forms polyamides can take in differing situations and on Brill transitions rather than raise expectations of the discussion of those in the experimental results. We will first describe the five major crystallographic forms encountered with polyamides as they crystallise from solution or the melt. There are: a) α, where the hydrogen bonds are in planes or “sheets” parallel to the edge faces of lamellae (often intra-molecular bonds) and layers are built up layer (sheet) upon layer [63]. With α there is an offset from the bonds of one layer to another resulting in the basal planes of the lamellae being inclined to the chain direction.

Wide Angle X-ray Diffraction (WAXD)

gives two peaks at approximately 0.44 and 0.38 nm respectively at room temperature. This is a stable crystalline form. b) β, identical to α except that the chains with their offsets are stacked one up and one down resulting in the (rougher at a molecular level) lamellar basal plane being more or less perpendicular to the chain direction [63]. This is a stable crystalline form. c) γ or pseudohexagonal and has inter-molecular hydrogen bonds between amide groups in separate layers (sheets). The energetics result in a slight offset between chains of different layers [63]. The chain spacing is nearly hexagonal with a spacing of approximately 0.41 nm. With equal intraand inter-sheet distances between chains it is possible now for the hydrogen bonds to be inter-sheet rather than intra-sheet. d) Hydrogen bonds with more than one direction. Here, the amide groups are twisted to give optimal energetics with one hydrogen bond to an amide group in a chain in the same sheet and the next hydrogen bond above or below being to the next sheet. 35

Recently, polyamide-6,9 was

found by Franco et al. [64] to belong to this overall group of polyamides. The groups that initiated this understanding are Subirana, Puiggali, Navarro and colleagues with collaboration from Atkins, Hill, Cooper and Jones [65-71]. e) Metastable pseudohexagonal forms [72] (broader single peak with X-rays) and other forms with imperfect α structures. There are also some other minor crystalline forms various authors have referred to, including the Atkins, Hill, Hong, Keller and Organ [73] work showing polyamide-4,6 has an α-like structure but with the chain direction completely perpendicular to the basal plane and amide groups in the chain fold. This is unlike the usual inclination to the basal plane, as found with polyamide-6 and polyamide-6,12. 1.2.5.9

Effect of polyamide Type and Segment Length on Crystal Form

The exact way that a crystallising polyamide molecule folds backwards and forwards to match up hydrogen bond acceptors and donors is very important [48 Section 1.3]. We know from earlier work by Roberts and Jenekhe [74] that virtually 100% of hydrogen bonds are consummated in the crystal, even if it requires bending of bonds or the backbone of the chain to link through from N-H to O.

Both the parallel and antiparallel chain alignments can

connect hydrogen bonds easily within the molecule if the polyamide is an “odd” numbered polyamide-n such as polyamide-7. Odd Nylon n types tend to be more stable in the α- or β-form. The stable form for “even” polyamide-n types, such as polyamide-6 is generally the α− or β-form with hydrogen bonds matching up parallel to each other and perpendicular to the overall polymer backbone. The angle in the lamellar basal plane between the intra-molecular hydrogen bonds and the corresponding chains of the next layer is at 67.50 to satisfy the steric and energy constraints. Distances between chains within the sheets are greater than between the van der Waals bonded sheets. The coefficient of thermal expansion is less in the plane of the molecules than the inter-planar direction. The hydrogen bonding constrains the molecular chains in a sheet much more than between molecular sheets, as these are only held together by the weaker van der Waals forces. The longer even polyamides can be more stable in the γ form. 36

Even

polyamide-n types in

the α- or β-form have

higher melting

temperatures than similar repeat length odd polyamide-n types by nearly 20 0C. The γ-form is regarded as being thermodynamically more unstable, which correlates with the lower Tm. The situation is further complicated with polyamide-m,n types because there can be odd with odd, even with even, odd with even and even with odd numbers of carbon atoms in the diamine and diacid sections respectively. Even the last two are different in the way the parts of a molecule or parts of different molecules can link together to consummate the hydrogen bonds. The requirements for crystallinity are that this all happens in a consistent way over (at least) regions of lamellae.

In some cases the crystal repeat

distances are two monomer repeat lengths. Odd-odd, odd-even and even-odd polyamide m,n types are usually more stable in the γ-form. Here the hydrogen bonds are made between amide groups in adjacent molecular sheets.

The γ configuration has the two

hydrogen bonds in a molecular repeat unit at an angle to each other, and neither is exactly perpendicular to the zigzagging backbone. This is because the bonds do not exactly match up opposite to each other. The energy of the total configuration must be minimised. It leads to hydrogen bonds holding the sections of molecules further apart than would be the case for the α- or β-form, and even further apart than for polyethylene.

It also leads to a

slightly shorter repeat length. The total outcome is a crystal with slightly lower density.

The angle in the basal plane between the intramolecular

plane and the chains of the next layer is close to 600 and this leads to a near hexagonal crystal structure, usually referred to as pseudohexagonal.

The

coefficient of thermal expansion is the same in both directions of the basal plane. Even-odd polyamide-6,7 would seem, at first sight, to be similar to the odd-even polyamide 7,6 but the bonds have to be twisted at different angles to make the O.…H-N connections.

The result is slightly different material

properties between the two polyamide types. There are a number of diverse characteristics that can be found in the different polyamide types.

Even-even polyamide m,n types are generally

more stable in the α- or β-form [75]. 37

“Shorter” Nylons have a higher

hydrogen bond density and have higher melting temperatures and densities than the same types with longer overall repeat lengths. Polyamide-6,6 has a 30 0C higher melting temperature than Nylon-6, although both have the same overall hydrogen bond density. Crystalline forms are different because of the different orientation of amide groups within the chains [76, 77]. Different Nylon types have differing levels of moisture uptake due to their various hydrogen bonding configurations [78 p. 324]. 1.2.5.10

Multiple crystalline forms are possible - Polymorphism

There is usually more than one form possible for a particular polyamide type but it often depends upon the thermal history as to which one is present in a sample. The α- or β-form is more stable for longer polyamide n types and the γ-form for shorter polyamide n types. Polyamide-6 appears to be equally likely to have both forms, and these can coexist in a lamella. Polyamide-4,6 can exist in both α- and β-forms at the same time [79]. The form that exists in a polyamide depends on steric restrictions and the most energetically favourable situation at a particular time.

Often a

metastable crystalline configuration will form first, and later the crystalline structure will change to another arrangement of hydrogen bonds, bond angles and crystal cell distances.

It can become energetically more

favourable to change to a different configuration as the thermal history of a crystal develops. Conversion between the two forms can be made to take place by temperature changes [80] and also by solvents or materials that make polyamide swell [81]. Sometimes a number of crystal forms will be present in the one sample [82] and for polyamide-12 [83], the crystal structure can be varied by pressure and cooling rates. 1.2.5.11 Pressure

Effect of pressure on crystallinity, melting temperature and crystal form often

affects

polymers

by

increasing

crystallinity

[84],

melting/crystallisation temperature and can change the crystallographic structure.

In particular, pressure affects the way in which polyamides

crystallise such as with the Ramesh work on polyamide-12 [83] and supported by the English abstracts of the Chinese language work by Lu Huang, Fan, Cai and Xie on polyamide-6 [85, 86]. Gogolewski and Pennings 38

show in their work on polyamide-6 [87, 88] that crystallisation under pressure increases the crystallinity, although a greater increase can be gained afterwards by annealing under pressure. 1.2.5.12

Metastability

Some materials go to metastable forms above or at the crystallisation temperature and then change to more stable forms as the temperature is lowered. Fast cooling can often trap crystal structures in a metastable form because the molecules quickly lose the energy to surmount an activation energy barrier.

“Cold crystallisation” can often only take place when

previously quickly cooled material is raised in temperature to near melting. A kinetic event takes place rather than a thermodynamic one with the melting and recrystallisation into a more stable form before melting of the stable form into liquid melt can take place. 1.2.5.13

Brill Temperature

The Brill transition occurs where a low temperature α form is heated so far that the inter-sheet spacing increases until it is the same as the intra-sheet spacing.

Some contraction of the intra-sheet spacing is required with

temperature increase for the energy minimisation of the structure. Eventually both d-spacings become 0.41 nm in a hexagonal structure. At this stage, the hydrogen bonds can easily change from intra-sheet to inter-sheet. The structure then becomes the γ form described above.

The

changes in d-spacings between crystalline planes can be followed with WAXD as temperature increases.

The Brill temperature (TB) is the point

where there is no difference in spacings. The Brill transition has been most extensively studied in polyamide-6,6 [64, 89-92] but does also occur in other even-even polyamides [79, 93-98]. The Brill transition is reversible and on cooling, the stable hexagonal γ form material reverts to α form. Some polyamides do not quite reach the Brill transition before they melt. The stable form of the crystal will remain α- or β-form in these cases. Kohan states [99 p. 143] that Brill transitions are usually not seen with DSC scans for melt crystallised samples, so they are not expected to be seen with DSC in the work carried out here.

39

This discussion has been included to alert to some of the complexities involved.

No further discussion of Brill temperatures is given in the text

because of the use of DSC results and absence of WAXD results.

1.3 Relevant papers in the area to be covered in the research There has been much done in the way of research on amorphous-amorphous and crystallisable-amorphous polymer systems (including polyamides) and methods to overcome miscibility problems. Much of that has been driven by the desire to improve the physical properties of polymers in a cost-effective manner. A few have done work on semicrystalline-semicrystalline blends, sometimes enhancing miscibility by hydrogen bond interactions [100] (although Qiu et al. only touch on those interactions) and a few have researched semicrystalline-(crystalline) small molecule blends. The area covered by the research is concerned with the melting and crystallisation of aliphatic polyamides with certain, potentially hydrogen bond disrupting, small molecules.

With the exception of water, this area

does not appear to have been covered by other researchers but there have been papers published in adjoining areas and these will be reviewed in this section. 1.3.1

Small molecule-small molecule

Sucrose is usually crystallised from anhydrous melts or highly concentrated solutions in a controlled manner to generate the specific textures or appearance required for the final product such as fudge, hard candies. It is shown in a paper by Bhandari and Hartel [101] by DSC and XRD results that it is possible to reduce the crystallinity from molten anhydrous sucrose to about a third by the addition of up to total weight 20% fructose, glucose or a mixture of the two in a co-crystallisation process. 1.3.2

Polymers with small molecules

Kristiansen et al. have studied sorbitol based nucleating agents(DMDBS) for removing haze

from

isotactic poly(propylene) (iPP) in their recent paper

[102]. They also studied a very much wider range of concentrations than the optimal clarifying concentration near 0.8% so that they could understand some of the mechanisms.

They refer to a regime III near the melting

temperature of DMDBS (higher than the iPP) where phase separation takes place (as determined with optical microscopy). At lower temperatures, there 40

is a partially crystallised fibrillar structure that solidifies below the eutectic temperature. Simek et al. [103] have studied the melting temperature depression of isotactic poly(propylene) (iPP) by alkanes. They have used the Flory Huggins relationship to explore the size of the effect. Kim and Kim [104] looked at liquid-liquid phase separation occurring with vinyl acetate and paraffin wax blends with poly(ethylene-ran-vinyl acetate) using DSC, cloud point determinations and wide angle X-ray diffraction. Their 1 0C/min and 10 0C/min cooling results are closest to the 2 0C/min and 25 0C/min cooling rates used in this work and show only slight differences in the DSC thermograms for reheating after nonisothermal crystallisation. 1.3.3

Blend interactions and hydrogen bonding

There are some similarities in a recent paper by Rocco et al. [105] to the original concept for hydrogen bond disruption by small molecules. In their case they were interested in suppressing crystallisation of poly(ethylene oxide) (PEO) by hydrogen bonding poly(bisphenol A-co-epichlorohydrin) (PBE) to enhance properties of PEO being used as polyelectrolytes in batteries. The blend interactions were observed with a shift for O-H from 3495 to 3348 cm-1 in going from the “free” (non-hydrogen bonded) state to the “bound” hydrogen bonded state. This change of 50 cm-1 seen in peak position (without Gaussian deconvolution) is indicative of what could be expected with polyamides and hydrogen bond disrupting diluents if there were any hydrogen bond interactions. Dormidontova and ten Brinke [106] tackle the influences of hydrogen bonding

on

perspective

microfor

and

comb

macro-phase

copolymers

separation

with

from

hydrogen

a

theoretical

bond

interacting

end-functionalised oligomers. Kobori et al. [107] looked at interfacial interactions of immiscible polymer blends (linear-low density polyethylene/poly(methyl methacrylate) with polyethylene) where hydrogen bonding did and did not play a role. The two combinations phenol)

respectively

containing

(LLDPE/PVPh

with

linear-low

density

polyethylene/poly(4-vinyl

polyethylene-block-poly(methyl PE-b-PMMA)

and 41

the

methacrylate)

non-associating

blend

LLDPE/PMMA with PE-b-PMMA were studied. FTIR measurements showed differences with an extra peak for hydrogen bonded carbonyl groups 30 cm-1 from the normal peak for unbonded carbonyl groups at 1730 cm-1. These were associated with differences in the phase boundaries demonstrating lower interfacial tension between the phases. 1.3.4

Polyamides and Polymers

Much of the work reported in this area is by researchers trying to overcome the abysmal performance of polyamide/other polymer blends using a variety of compatibilisers. Often comments are made about the uncompatibilised blends that give an idea of the normal situation. Moon, Ryoo and Park [108] discuss their work on using maleic anhydride grafted

polypropylene

as

a

compatibiliser

to

improve

polyamide/polypropylene blends that are a semicrystalline/semicrystalline combination. Jafari et al. [109] studied the crystallisation of polyamide-6/polypropylene blends using hot stage microscopy to look at the formation of polyamide spherulites and how the polypropylene crystallised at a later stage. It will be raised in a later chapter later that this paper may be relevant to the way polyamides crystallise in certain circumstances. Murthy

et

al.

non-crystallisable

[110]

took

aromatic

the

interesting

polyamide

with

approach (normally

of

blending

a

crystallisable)

polyamide-6 and used simultaneous small and wide-angle X-ray studies to probe the crystallinity of the polyamide-6 in the blends. They found that the polyamide-6 crystallinity was depressed by the presence of this other polyamide. Kim, Cho and Yoon [5] have recently studied the effects of compatibilisers on blends between polyamide-6 and poly(vinylidenefluoride) (PVDF) to improve the poor mechanical performance of these semicrystalline/semicrystalline blends. The uncompatibilised blends had strong phase separation. The two areas the uncompatibilised blends were noted for were poorer compatibility in the amorphous regions and faster crystallisation. PVDF is crystallised isothermally with polyamide-11 by Li and Kaito [111] and studied as uniaxially stretched films with SAXS and WAXD with or without annealing. There are limited DSC results with a peak for the 42

polyamide crystallising in the blend at a temperature higher than normal for polyamide.

It is possible that this experimental result is consistent in

mechanism with a couple of similar examples of this in this work, despite theirs being a polymer-polymer system. 1.3.5

Polyamides and small molecules

Cha et al. [8] have studied a system of polyamide-12 with poly(ethylene glycol) (PEG) of differing molecular weights in relation to the formation of membranes by thermally induced phase separation. This is chemically the closest of the available literature to the systems studied here but they have tackled it from a different perspective with a focus on the effects of molecular weight of the diluent and on diluent-rich domain size. Their work used light transmission changes to detect phase separation in dynamically cooled (10 0C/min) melt blends. Samples 200 µm thick between coverslips could be cooled quickly into the unstable region where droplets of polymer-poor material formed and solidified once the phase separation temperatures had been determined.

Sample thickness at 200 µm is less than half the

estimated thickness of the samples investigated in this thesis but that is not expected to induce significant differences due to dimensional constraints. Some samples were initially produced by first solvent casting (with heated vacuum drying) before forming the melted film. The authors claim that no differences were detected due to this procedure. Videomicroscopy was also used for phase separation temperatures.

Their study gives experimental

phase diagrams with cloud point curves delineating the liquid-liquid phase separation boundary and melting & crystallisation points for the Nylon at different polyamide concentrations. They do not describe how these latter data points were determined and whether they are the melting and crystallisation peak temperatures or onset temperatures. However part of the group describe in a later article how they use optical methods to measure melting and crystallisation temperatures [10]. Whether those are the methods used in the 1995 paper is not clear but that paper does not describe the use of equipment other than hot stage optical microscopy observations, videomicroscopy and SEM. devoted

to

the

development

of

The early part of the paper is phase

diagrams

giving

the

temperature-composition conditions where two-phase behaviour exists and the results of that are used for setting up experiments where mixtures are 43

quenched to 170

0C,

diluent rich domains develop and later the samples are

cooled to ambient temperatures. The focus here is on the size of PEG-rich domains that can be used to form membrane pores. Their study covered PEG with molecular weights of 200, 400, 600, 1000, 1540 and 3400 Daltons. The results for the PEG having a molecular weight of 200 Dalton are the closest to the carbazole and phenothiazine used in this study (167 and 199 Daltons respectively) and PEG has a quite different molecular

form

phenothiazine.

to

the

poly-aromatic

rings

of

the

carbazole

and

The carbazole and phenothiazine used here have quite

different molecular shapes to the PEG and only have molecular weights near 200. The molecular weight and form factors will affect the mobility of the diluent molecules in the amorphous and molten polyamide in different ways and the chemical potentials of the diluents with respect to that of the polyamides will differ. Polyamide-12 is a polyamide with a lower density of amide groups than the polyamide-6,12. It has a lower melting temperature than the polyamides studied here and would be expected to have a lower crystallinity also, approximating a polyolefin much more than them. The major findings of the paper were that UCST behaviour was seen, solid-liquid as well as liquid-liquid phase separation were seen, that the two phase region was larger with increasing diluent molecular weight. They also showed that both diluent molecular weight and content in the mixture affected the interaction energies derived with Flory-Huggins theory. Factors they found that are perhaps of lesser interest in the context of this thesis are that the domain size was larger for greater PEG molecular weight but this was not so strong an effect at low PEG content of the mixture. This thesis later discusses the relative crystallisation temperatures of polyamide and carbazole or phenothiazine diluent. It can be noted at this point

that

PEG200

has

a

higher

crystallisation

temperature

than

polyamide-12 in the study by Cha et al. and that this corresponds to the case with polyamide-6, polyamide-6,9 or polyamide-6,12 combined with carbazole.

44

1.4 The focus of the research project Aliphatic polyamides, or “Nylons”, are an important class of engineering polymers. They are characterised by relatively high melting temperatures, high impact strength and toughness. An overall study of the literature in this general field has not uncovered much work generally in the area of polymers melt blended with small organic molecules and only one relevant paper on polyamides with small organic molecules with that one looking at quite different aspects [8]. The literature has shown that some others have achieved hydrogen bond complexing between

poly(ethyleneoxide)

with

poly(bisphenol

A-co-epichlorohydrin)

whereas that was found not to occur with the materials chosen. The research problem is to understand the processes involved in forming high temperature solutions by melting linear polyamides with carbazole or phenothiazine (as examples of small molecules) and in their crystallisation. It

is

also

to

understand

the

resulting

morphologies

arising

from

crystallisation and the effect of polyamide type. The research had originally been planned to investigate the role of hydrogen bond formation on crystallinity in linear polyamides. The concept was partly based on the work of Damman, Point and coworkers [21, 112-114] in creating molecular complexes between poly(ethylene oxide) (PEO) and p-nitrophenol or resorcinol that are hydrogen bond complexed with them. There was also some (as yet unpublished) work done by others within our group at the University of South Australia on poly(ethylene glycol) (PEG) and resorcinol. The project had also been undertaken to study the effect of hydrogen bond complexing on the physical properties that make Nylons desirable to use in many applications.

The potential benefits of the project were to aid in

widening the manufacturing/processing window for Nylons, to provide options for adding dyes to Nylons in solution or the melt and the potential to assist in developing new ways to deliver drugs within the body by encapsulating them in polyamide excipients. A more fundamental reason for doing this research was to help the understanding of hydrogen bonding in polyamides in general. There was also the possibility of using a synthetic,

45

model compound to better understand protein folding because of the similarities between amides and the peptides found in proteins. The aim was to insert organic materials in the melt that would disrupt the polyamide-polyamide hydrogen bonding so that the strong hydrogen bonds would be destroyed and the material properties altered. This approach using organic hydrogen bond disruptors is quite different from the earlier inorganic approaches by others. Those had concentrated on iodination [115-119] or the use of metallic ions such as Ca++ [120] for the study of changes in crystalline structure.

This alternative approach was taken because of the

obvious close parallels with many biological systems. The work was done in the melt rather than room temperature solutions to avoid the three-way competition for hydrogen bonds that would arise from dissolving the polyamides in a solvent [42]. Polyamides require very strong solvents such as formic acid, concentrated sulphuric acid, m-cresol or special solvents such

as

1,1,1,3,3,3hexafluoroisopropanol [121] that have to destroy the

polyamide-polyamide hydrogen bonds in order to dissolve the solid polymer in the first place.

The intention was to use DSC as part of the material

property analysis. It will later be clear from FTIR results that the materials chosen did not result in hydrogen bond interactions with the polyamides along the lines expected [42, 112, 113]. 1.4.1 1.4.1.1

Materials chosen Polyamides

Four, quite different, representative polyamides were chosen for the study so that the conclusions could be as general as possible.

It has been shown

above that the melting temperatures and crystallography of polyamides are influenced strongly by the type of polyamide. The relevant parameters included whether they were polyamide-m,n or polyamide-n types, whether polyamide-m,n is even-even or odd-even and the density of amide groups in the backbone is high or low. One of the polyamide materials chosen was polyamide-4,6 [122, 123] which is an even-even polyamide with a high amide group density. It has much higher crystallinity and melting temperature than most other polyamides [55] due to the above factors. It provides one end of the scale of even-even polyamides studied. Polyamide-6,12 is towards the extreme of even-even 46

polyamides having low amide density and still readily available. The very common polyamide-6,6 was not chosen because there were no samples available that were not known to have fire retardants and other additives and because the two more extreme members of even-even polyamides were being studied, allowing estimates for the intermediate polyamide-6,6. Polyamide-6 was available in grades not known to have additives. Polyamide-6 and polyamide-6,6 are the most common of commercial polyamides so polyamide-6 is representative of both a mid amide density polyamide and a polyamide-n type. Polyamide-6,9 was also available in a grade not known to have additives and is representative of an even-odd polyamide-m,n. Its melting temperature is lower than polyamide-6,12, with even lower amide density, due to the even-odd configuration having unfavourable hydrogen bond linkages. It is also a member of the group of polyamides now known to have hydrogen bonds in multiple directions [64]. These four polyamides have melting temperatures between 209 and 290 0C and provide a compact group with a suitable range in repeat unit types, stable hydrogen bond structures and melting temperatures.

This should

allow us to draw some general conclusions about the interactions and phase behaviour of polyamides melted with the two chosen materials. 1.4.1.2

Small molecules

Work started originally with 2-methyl resorcinal and p-dihydroxybenzene (hydroquinone) as these had been hydrogen bond complexed with PEG in Paternostre, Damman and Dosiere’s work [124, 125]. Evaporation was an immediate problem because the polyamides melt at such high temperatures, so other potential materials such as benzophenone with higher boiling temperatures were also tested. There were several determining factors in the choice of small hydrogen bond disrupting molecules to be used in the originally planned melt complexing project. A list of criteria was then drawn up. Polyamides start to degrade (scission of the polymer chain at the amide groups above approximately 325 0C as seen in the TGA thermogram later (Figure 1-16 in Section.1.5.1). There is usually further polymerisation of polyamides at temperatures near the melt and above [48]. Extended periods 47

at elevated temperatures above 300 0C would result in a marked increase in polydispersity from scission and further polymerisation that would detract from the validity of the work because of the uncontrolled molecular weight distribution. The small molecule melting temperature upper limit became 300 0C. a) Trials need to run substantially above the melting temperature of both the Nylon and the potential disruptor so that self-seeding nuclei from either material would not remain in the melt to cause premature crystallisation.

In particular, the polyamide should have over five

minutes fully in the melt to remove the previous crystalline state. b) The potential disruptor should not evaporate or decompose at the temperature of the trials, ie. more than 300 0C in the case of experiments with polyamide-4,6. It was preferable to have the same material(s) for all polyamides so that valid comparisons could be made. c) The affinity of a hydrogen bond disruptor for the Nylon hydrogen bonds should preferably be greater than the strength of polyamide-polyamide hydrogen bonds. d) There should only be one potential hydrogen bonding site on the molecule so that bridging between several polyamide chains

(or within a chain)

would be avoided. These criteria are quite difficult to meet. For example the common Nylon plasticisers N-ethyl o- or p-toluenesulfonamide boil at 196 0C and have multiple

potential

hydrogen

bonding

sites

per

molecule.

Another,

N-butylbenzenesulfonamide, boils at 314 0C Nearly a dozen potential compounds were found that seemed to be suitable and each one had a single N-H or C=O bond available for hydrogen bonding. Some of those were not commercially available and could also not be obtained via contacts in various laboratories.

A handful of the rest

remained. Many of those were evaluated with Simultaneous DTA-TGA (SDT) to eliminate poor performers on the critical evaporation criterion. That left only two, carbazole and phenothiazine, that were reasonably suitable.

The melting temperature of carbazole is 246 0C and its boiling

temperature is 355 0C whilst the melting temperature of phenothiazine is 48

186 0C and its boiling temperature is 371 0C [126] . It was found that the boiling temperature is not as critical as the vapour pressure at the working temperatures near 310 0C. It will be seen in later chapters that carbazole with the lower boiling temperature gave less trouble in the trials than the higher boiling phenothiazine. Problems were still encountered for carbazole and phenothiazine with evaporation, even for polyamide-6,9, with its low melting temperature, and even with high heating and cooling rates to minimise evaporation. The structures of the two materials are shown below in Figure 1-15.

H

H

N

N

S Figure 1-15 The structures of Carbazole and Phenothiazine.

These are both relatively flat molecules although the phenothiazine has a slight curvature from top to bottom as computed for us by Dr. Adam McCluskey at Newcastle University in New South Wales, Australia.

Both

have pi electron clouds above and below the benzene rings. 1.4.2

Sample blending and notation used for blends

Small samples could be made up from powders in Differential Scanning Calorimeter (DSC) pans to understand the initial melting (plus crystallisation and later remelting) in the DSC. Larger blend samples were mandatory to study properties using a variety of the techniques described below in Section 1.6 and these could be made in ampoules. A consistent notation is used within the thesis for melt blend samples. Polyamides are often described in the literature in various forms.

For

example, polyamide-4,6 is seen in articles as Polyamide-4,6 polyamide-46 Polyamide4,6 polyamide46 Nylon4,6 Nylon46

PA-4,6

PA-46 PA46

Nylon-4,6

Nylon-46

4,6-Nylon and some other variants. The versions that

will be used here for melt blending are PA46, PA6, PA69, and PA612 for polyamide-4,6, polyamide-6, polyamide-6,9 and polyamide-6,12 respectively when combined with Car

for carbazole, or PTh for phenothiazine (This

should perhaps have been PhTh but the aim was to keep it to three letters signifying which diluent was involved in the blend.). Polyamide-4,6 blended 49

with carbazole is generally noted as PA46Car. Specific samples with known percentages of polyamide are preceded by the weight percentage of polyamide eg 39PA69PTh for a sample of 39% polyamide-6,9 in combination with phenothiazine. The value of 39% would be calculated from the few milligrams of each material used when blending in pans in the DSC or from the TGA results where a sample is taken from the bulk material made in larger quantities in an ampoule. The samples for TGA are taken from next to the DSC samples.

This notation provides an easily recognisable and

compact descriptor for each sample.

1.5 Experimental Techniques Used This section includes results that will be used to illustrate certain recurring features that will be discussed throughout the thesis. The major focus of the work rests on the results of Differential Scanning Calorimetry and Fourier Transform Infra-Red spectroscopy (in Mid and Near Infra-Red ranges) with the support of Thermogravimetric Analysis for determining polyamide concentration in ampoule samples. 1.5.1


Thermogravimetric Analysis (TGA) is used in this project to determine the weight percentage of polyamide in a bulk sample where the composition may vary markedly from the average for the whole sample.

It is a technique

where a sample of material is heated in a gas stream with a furnace and the weight is monitored accurately with an extremely sensitive balance.

Figure 1-16 Evaporation of carbazole followed by degradation of polyamide-4,6 in TGA.

50

This technique can be used because carbazole (or phenothiazine) in a blend sample will evaporate in an inert gas stream (nitrogen) before the polyamide begins to degrade. Evaporation usually takes place (at 10 0C/min ramping rate) in the range 175-275 0C but the polyamide does not begin to degrade at that ramp rate until well into the molten state over 325 0C. It means there is a plateau in the TGA thermogram of remaining percentage of the samples’ weight vs. temperature at least in the range 275-325 0C. A small amount of degradation products from the polyamide usually remains by 600 0C [127]. The plateau is clearly observable in the typical TGA thermogram depicted in Figure 1-16. 1.5.2


Differential Scanning Calorimetry (DSC) was the main technique used in the experimental work. This is because it was able to provide information on melting and crystallisation temperatures and crystallinity of samples when they were being heated into the molten state and crystallised during cooling to room temperature. Additionally, the DSC was used as a ‘furnace” to take small samples of polyamide and diluent powders to the melt to study the high temperature solutions.

Monitoring could take place in situ whilst

carrying out this preparatory process. It enabled a better understanding of the initial eutectic formation from the raw mixes of powders. DSC measures the flow of heat into or out of samples when they are heated or cooled. Thermal transitions as a function of temperature and time give quantitative and qualitative information regarding physical (and chemical) changes such as melting, crystallisation, recrystallisation glass transition temperatures, cold crystallisations, polymerisation, degradation reactions, volatilisation or changes in heat capacity.


temperatures can be determined. The amount of crystalline material that has melted or crystallised can be determined and, by comparison with literature values for 100% crystalline material, the crystallinity can be found. There are two types of instrument, “Heat Flow” and “Power Compensated”. In the first type, the temperature difference between a reference and sample pan is measured as both are heated in similar situations in a DSC cell. The other type determines the amount of power required to keep the sample at the same temperature as a reference as they are both heated in a cell. 51

There are a number of quite different designs for the cells used with both types of instrument [128 p. 129]. The type used in this work is a Heat Flux instrument. Figure 1-17presented here shows a cross-section of a DSC cell for a TA Instruments calorimeter Model 2920 DSC.

Figure 1-17 Cross section of DSC cell (taken from [129 p. 4-5].

There are two methods of treating heating ramps for differential scanning calorimetry.

There is “standard” DSC with a constant ramping rate and

“temperature modulated” DSC (TMDSC) where a sinusoidal or sawtooth [130] modulation is superimposed on the constant ramping rate. This later method, developed since 1993 [131], was put forward as having a number of experimental advantages.

It has, however, been more recently recognised

that there can be limitations in the interpretations [132-134], especially with melting and crystallisation events. The work in this thesis was done under TMDSC conditions (with the extra calibration required) to utilise the smaller sample size, increased resolution and sensitivity. Analysis could then still be done at a Reversing/Non-Reversing level where it was required and appropriate.

Glass transition temperatures are also obtainable where the

crystallinity is not too high.

Unfortunately, polyamides are often high in

crystallinity, leading to weak glass transitions. There is a more extensive discussion of standard DSC, TMDSC and the use of Lissajous figures to better understand thermal events during TMDSC in Appendix B.

The caveats placed on the use of TMDSC described in this

appendix mean that it was inappropriate to analyse the melting and crystallisation processes of highly crystalline diluents and very crystalline polyamides from a TMDSC perspective. 52

Small-molecule diluents remain solidified until the temperature is raised sufficiently that molecular motion catastrophically breaks down the crystal structure.

The amorphous part of a pure polymer will be reduced in

viscosity with heating to the viscosity at melting and polymer chains comprising the lamellae will “melt” into this fluid of the same composition. Blends differ from both of these in that the amorphous part of a polymer is highly plasticised by the diluent, forming a solution that is of lower viscosity than the normal polymer melt. liquid.

The lamellae essentially “dissolve” in this

Technically the correct usage throughout the thesis should be

dissolution but in many cases there are pure materials melting and blends dissolving under the same section heading or in the same thermogram. The common term “melting” has been used for both headings and figure captions as well as text describing the melting of pure materials.

Usage of

“dissolution” has generally only been followed in the text for blends where there is specific discussion of polymer chains being removed from lamellae into the liquid. 1.5.2.1

Thermogram Overlays

In general, the DSC thermograms are displayed as overlays with several thermograms together in a figure. That is done to make better comparisons between different compositions investigated under the same conditions. The thermograms are all shown as heat flow in J/g against temperature in degrees Celsius. All thermograms have exotherms pointing upwards. The thermograms are spaced out vertically and coloured in a consistent manner to aid clarity.

The colour scheme can be seen in Figure 3-16 in

Section 3.4.4.1. The peak with largest amplitude starts with the endotherm or exotherm near zero.

In practice, that will be either carbazole or

phenothiazine. The other thermograms are placed in order of concentration through to the polyamide so that trends can easily be seen as they relate to polyamide concentration. It means that the polyamide curve will be at the bottom for melting and at the top for crystallisation. The legends are always with the pure polyamide (100% polyamide) at the top and range down to the diluent (0% polyamide) at the bottom. In some cases, the phenothiazine or carbazole peak is extremely large in amplitude compared to the thermograms of the polyamide or the blends. In 53

those cases the very large peaks have been truncated in the figure so that the detail of the other materials and/or combinations can be clearly seen. First time thermograms of blends are generally drawn with a dash and the repeat runs in the DSC are drawn with a solid line. The exceptions are the few cases where there is more than one thermogram in the same concentration range and other line types have been used. 1.5.2.2

Thermograms expected from thermal events

We will now consider the general forms of thermograms resulting from different types of thermal events. This will facilitate discussion of results in later chapters. The DSC thermogram will have a single peak for melting or crystallisation if the percentage of polyamide is exactly that for the eutectic composition because at the eutectic triple point the solid changes at one time through from the solid to the liquid phase or vice versa.

The temperature of that

peak will be close to the equilibrium eutectic temperature but will be modified by the dynamic heating or cooling not exactly being at equilibrium. There will also be differences in heating and cooling eutectic peaks because polymers are involved in this study and the

normal melting and

crystallisation of polymers do not take place at exactly the same temperature. Polydispersity of the commercial polymers used will also have an influence on the outcome. Consider now the case of polyamide/diluent with a polyamide concentration different from the eutectic concentration and being heated. In the first stage of heating, polyamide and the diluent melt up to the limit of solubility of one material in the other in a eutectic melting peak. That peak temperature is virtually constant across a wide range of total composition in samples. There is now a residual of one or other material because the two materials are not present at exactly the eutectic composition.

The solubility of the

excess material will generally increase rapidly at higher temperatures. The endothermic curve in the thermogram above the eutectic melting peak is due to the progressive melting of more and more of the residual material as the temperature is increased in the heating ramp.

Eventually the excess is

consumed and the sample is completely liquid with no further melting

54

activity.

This can be seen in Berghmans’ chapter of Mathot’s book [135

p. 214 Fig.8.7]. This process results in an endothermic curve above the eutectic melting peak in the thermogram that takes the shape seen in the second peaks of Figure 1-18.

These peaks have a similar form due to a similar process

taking place, however, the second peak for 25PA6Car extends further as more carbazole has to be dissolved into the liquid, requiring higher temperatures to increase the solubility.

A higher level of carbazole again

would require even higher temperatures to dissolve all the diluent.

This

form of curve will be referred to in the text as a Temperature Limited Solubility (or TLS) peak. It is interesting that the peak temperature is just a few degrees before the end of the melting process that defines the totally liquid state. A plot of melting peak temperatures against composition will be seen in later chapters to take on the general form of the eutectic phase diagram (Figure 1-6

of Section 1.2.1.6).

There are, however, differences

because the peak temperatures are not the end of melting but peak melting rate and because the system is not in an equilibrium state. The forms of the curves are slightly different between excess of diluent and for excess of polyamide but the principles are the same.

Figure 1-18 Two examples are shown to illustrate this general feature. The upper curve is for 25% polyamide-6 in carbazole and the lower curve for 64% polyamide-6 in carbazole. The first peaks near 195 0C corresponds to melting the eutectic composition and the second peak to melting the “excess” mixture of which there is more in the 25% sample.

55

We have seen in Section 1.2.5.12 that we can have metastable crystalline forms locked in to lamellae, particularly by fast cooling. These metastable lamellae have lower melting temperatures than the stable form and undergo a melting and recrystallisation into the more stable form before the final melting of the stable form. That can be observed in several variants. We can see the first melting of the existing metastable crystals absorbing energy, and later the heat given off in crystallisation of the metastable lamellae prior to the main peak endothermic melting of the stable form for the polyamide-6,12 sample in the thermogram below. We can also see a minor version of these processes taking place for the 60PA612PTh thermogram in Figure 1-19.

This latter thermogram only shows a shoulder early in the

main melting peak.

Figure 1-19 Melting and recrystallisation of metastable crystals before melting the stable crystals. This can either be an extensive endotherm and exotherm pair, as with the pure polyamide, or there can be a subtle dip before the main peak and a shoulder on the leading edge of it for the blend.

1.5.2.3

Assignment of “Spiky” Crystallisations to Carbazole or Phenothiazine

The crystallisation of carbazole and phenothiazine take place extremely rapidly because the molecules are quite small compared to long polymer chains. The heat released in crystallising often makes the peak temperature of crystallisation appear higher than the crystallisation onset temperature. The form of the crystallisation peak is very distinctive, as can be seen in Figure 1-20.

It is very easy to identify a crystallisation as being from nearly

56

pure carbazole (or phenothiazine), unlike the situation during melting. The following discussion about carbazole applies equally well to phenothiazine. The distinctive slight rise in temperature is due to the sample thermocouple being on the underside of the constantan dimple where the sample pan rests.

The pan contains the molten carbazole that is being cooled.

The

carbazole is still molten at the time the thermocouple reduces in temperature to below the carbazole freezing temperature due to slight thermal lag in the system. This is because of small but noticeable thermal resistances between thermocouple and carbazole.

The freezing carbazole

within the sample maintains it at the carbazole crystallising temperature so the thermocouple soon rises again to match that temperature.

It takes

approximately 6 s for the heat flow to reach a maximum. In that time the “Sample” temperature measurement “appears” to increase by 0.44 0C.

Figure 1-20 Displaying the radically different forms of crystallisation peaks for polyamide and diluents allowing identification of the material crystallising. Phenothiazine has a similar crystallisation thermogram form.

The crystallisation of polyamide, however, approximates a broader Gaussian distribution because it is a polymer crystallising and because the polymer is polydisperse (Section 1.2.4.5). 1.5.2.4

Phase diagrams derived from thermograms

Three examples are given here of experimental non-equilibrium phase diagrams. Figure 1-21, is from Cha et al. [8] referred to earlier as closest to the systems studied here. It used PEG (Mw = 200 Dalton) as the diluent with 57

polyamide-12 as the polyamide with cloud point measurements on Liquid-Liquid

phase

separation-and

some

(undefined)

melting

and

crystallisation measurements.

Figure 1-21 Experimental phase diagrams measured under the condition

of 10°C/min cooling rate: (a) nylon 12/PEG2 00 blend from Char et al.

[8], where Tcloud is from cloudpoint measurements, and Tm & Tc are

melting and crystallisation temperatures respectively.

260 250 240

Liquid

Solid Liquid

0

Temperature ( C)

230 220 210 200 Liquid & solid

190 180

TmPA69Pure TmCarDepr TmEut TcPA69Pure TcCarDepr TcEut

170 160 150 140

Solid Liquid

Solid Liquid Solid

130 0

10

20

30 40 50 60 70 Polyamide concentration (%wt)

80

Figure 1-22 Example of eutectic style non-equilibrium phase diagrams for heating to the liquid state and cooling raw materials and blends of polyamide-69 (PA69) and carbazole (Car) Melting peaks are noted with Tm and crystallisation by Tc. Eutectic points are denoted by TmEut or TcEut respectively and blends having peaks depressed are denoted in the legend by Depr.

58

90

100

The other two, Figure 1-22 and Figure 1-23, are typical of those seen in later chapters, one being a eutectic crystallisation and the other being a Flory-Huggins style crystallisation. There is some uncertainty in the phase diagrams having Flory-Huggins crystallisation as to whether the melting having near-constant melting temperature is a true eutectic or not but the term eutectic will be used in the text. Figure 1-22 takes the same form as Figure 1-21 except that Liquid-Liquid phase separation is replaced by melting and Liquid-Solid phase separation for the crystallisation of the diluent. 230

Liquid

220

Liquid Liquid & crystallites

210 200 0

Temperature ( C)

Liquid & crystallites Liquid

Solid & liquid Liquid


190 180


170

Solid Solid & liquid

160 150

Solid & liquid Solid Liquid & crystallites Solid Solid & liquid

TmPA69Pure TmPA69Depr TmPThDepr TmEut TcPA69Pure TcPA69Depr TcPThDepr

140 130

Solid

120 110 0

10

20


80

90

100

Figure 1-23 Example of Flory-Huggins style non-equilibrium phase diagrams for heating to the liquid state and cooling raw materials and blends of polyamide-69 (PA69) and carbazole (Car) Melting peaks are noted with Tm and crystallisation by Tc. Blends having peaks depressed are denoted in the legend by Depr.

The eutectic style and Flory-Huggins style crystallisations presented in the various chapters are consistent with the above two types of phase diagrams in that samples are from ampoule material and temperatures are peak temperatures. Red coloured text and graphics refer to heating at 5 0C/min whilst blue is for cooling at 2 0C/min. Phase regions described in black are common to heating and cooling. Reference to crystallites is where there have been a small amount of near pure crystallites of polymer with melting and crystallisation temperatures very close to the pure polyamide.

The phase

regions for them are delineated in the liquid region by fine broken lines. Heavy long dashed lines delineate the melting/crystallisation of polyamide. 59

Heavy short dashes delineate the melting/crystallisation of diluent.

Solid

lines delineate the melting/crystallisation of eutectics. Data points are solid diamonds for pure polyamide, solid squares for polyamide in blends depressed in peak temperatures, triangles for the diluent and solid circles for eutectics. It can be seen in many of the phase diagrams that, where polyamide and diluent where there has been Flory-Huggins style crystallisation to be also defined as melt or crystallise almost simultaneously, there is a slight depression of the transition temperature in a manner similar to that described by Berghmans [135]. The crystallisation of diluents giving “spiky” peaks allows definitive assignment of crystallisation peaks. That is not the case for melting where peaks are Gaussian in form.

The assignment of

melting peaks to polyamide or diluent has often been determined for melting phase diagrams from the order in which the crystallisation has taken place, recognising that whichever has the higher crystallisation temperature will also have the higher melting temperature for the same blend. Account is also taken with this of the area of the peaks in relation to the amount of material of each in the blend. This has allowed phase diagrams for melting in the case Flory-Huggins melting rather than as eutectic melting because often the first melting peak is taking place at near-constant temperature. There are three alternatives for phase diagrams.

One is to consider the

onset of eutectic melting as the solidus and the end of a TLS peak as the liquidus but this leaves an unusual gap between them right at the eutectic point caused by the difference between start and ending of eutectic melting. Another is to do as Visjager, Tervoort and Smith [136] and combine peak temperatures of eutectic melting with the end of the TLS peak but this is an unusual combination and would require a different approach where Flory-Huggins style melting or crystallisation takes place.

It has been

decided for consistency to use peak temperatures throughout the phase diagrams, giving consistency in presentation of information throughout the whole thesis.

As a caveat, it should be recognised that the experimental

conditions differ strongly from equilibrium and that the use of peak temperatures is not giving the temperatures at which all material is in the liquid (or solid) state.

60

1.5.3

Simultaneous Differential Thermal Analysis/Thermogravimetric Analysis

Simultaneous Differential Thermal Analysis/Thermogravimetric Analysis (SDT) allows combined TGA and DSC to be run on a sample at the same time.

It does not have quite the same TGA or DSC sensitivity as the

individual instruments.

It does have major benefits in allowing rapid

assessment of materials from a large selection of materials.

Melting,

evaporation and degradation can be assessed from a single fast experimental run. This allowed the efficient selection of candidate materials for the work based on having high melting temperatures and the material not evaporating too quickly in the working range to 300 0C. A typical analysis curve obtained from this technique is shown below in Figure 1-24.

Figure 1-24 Example of Simultaneous Differential Thermal Analysis with Thermogravimetric Analysis (SDT) for Flourenone.

The dashed curve of the temperature difference between the sample and an inert reference shows the melting of fluorenone just above 80 0C followed by the evaporation of the molten fluorenone. The solid curve does not show any significant weight loss at the melting temperature but the sample mass has been reduced to zero by 255 0C due to evaporation in the nitrogen gas stream. The maximum feasible working temperature for combination with polyamides in a DSC pan would be just above 150 0C, too low for polyamides that melt at 209-290 0C.and may have to be taken up to 307 0C to remove residual lamellar nuclei. Approximately 20 candidate materials were quickly 61

evaluated this way, leaving carbazole and phenothiazine as the only reasonable ones left. 1.5.4 1.5.4.1

Fourier Transform Infra-Red Spectroscopy General

Infra-red spectroscopy is a technique that measures the interaction between a material and infra-red (IR) frequency electromagnetic radiation.

It is

commonly used in the mid range frequencies of 4000 to 400 cm-1. This is the region where molecular vibrations and rotations show absorbance bands that are characteristic of the atoms involved in the bonds. Atomic bonds in a molecule can generate absorbance bands in the far, mid and near infra-red, regions. Energy is absorbed when the frequency of the irradiating electromagnetic waves is in resonance with characteristic modes of molecular movements such as bond stretching, vibrations and rotations. Changes in the states of groups of atoms in a polymer molecule can cause shifts in the absorbance bands. These can contribute to our understanding of what is happening to the molecules or parts of molecules and it is this area that is important when looking for the influence of changed hydrogen bond environments [42, 112, 113]. Each specific chemical environment will have characteristic frequencies where infra-red radiation is absorbed.

In

principle, this allows the determination of molecular structures from the infra-red “fingerprint”, however full interpretation can be difficult because of the myriad of different possibilities with reasonably sized organic molecules. It is quite easy, for example, to discriminate at four or five places across the mid IR spectrum between the various polyamides used in the trials. Various authors have discriminated between amorphous and crystalline states of polyamides [82] and between various crystallographic forms of polyamides [82, 137, 138] including Brill transitions [91] using a variety of FTIR techniques. Polyamide interactions with liquid crystal oligomers have also been detected [139], as have those with fibre reinforcement [140]. The majority of work on Fourier Transform Infra Red spectroscopy FTIR) is in the mid range of the IR spectrum but the Near IR (NIR) in the range 11,000 to 4000 cm-1 is very sensitive to subtle differences in hydrogen bonding. We will see later that the original premise of hydrogen bond destruction with polyamides by the potential hydrogen bond acceptors, carbazole and phenothiazine, was cast into doubt by the mid range IR work. This prompted 62

validation by NIR investigations in general bands identified as being related to hydrogen bond interaction with polyamides [141]. Earlier instruments were dispersive but these have largely been supplanted by

inexpensive,

rugged

Fourier

Transform

Infra

Red

spectroscopy

instruments that allow relatively quick measurements to be made with extremely high signal-to-noise ratios. A Nicolet 750 FTIR instrument was used for the work. There are a variety of FTIR techniques available to use: a) Attenuated Total Reflection(ATR) b) Transmission of solutions c) Transmission of cast thin films d)

Diffuse Reflectance Infra-red Fourier Transform (DRIFT) spectroscopy where the Infra Red sample beam is deflected downwards onto the surface of a sample with an elliptic mirror. Any diffuse IR reflection from the surface is collected with another elliptical mirror and focussed back into a beam incident on the detector. One advantage of this technique is that samples directly as formed may be examined without disrupting the morphology or chemical interactions between molecules.

e)

Photoacoustic Spectroscopy (PAS) studies utilise the generation of thermal waves in the sample upon infra-red absorption. This leads to acoustic waves being propagated within the sample and into the surrounding gas. A sensitive microphone picks up the acoustic signal and amplifies it to give spectra as the IR frequency is swept across the mid infrared spectrum. The original principle dates back to the 1800s. It has been applied in FTIR for the last dozen years or so.

The first three techniques were not utilised because they involved modification of the bulk samples in ways that would alter their morphology at the detecting surface or in the bulk. Photoacoustic (PAS) detection was used for the Mid IR range experiments because it is more suitable than DRIFT when looking at small differences in frequency.

The photoacoustic approach does have a disadvantage in that

the heights and areas of peaks do not necessarily represent the relative intensities of the absorption of IR. There can be an attenuation of strong 63

signals.

This will be discussed below.

In this particular case, the

advantages of using material in its native morphology and having very accurate peak frequencies outweigh the drawbacks due to non-linearity. DRIFT was used for the NIR experiments because the instrument signal was far superior to the photoacoustic signal with that part of the IR spectrum. 1.5.4.2

Mid Range IR and hydrogen bond Interactions

Polyamides have an N-H stretch with a large peak near 3300 cm-1.

The

normal situation for polyamides is to be strongly hydrogen bonded from the carbonyl oxygen through the amide hydrogen to the nitrogen of another amide group on the same polymer molecule or another molecule. The large peak near 3300 cm-1 represents the bound state because the vast majority of potential hydrogen bonds are consummated at room temperature [48 p. 270]. The state of the N-H bond in carbazole material is normally unbound. There is a major N-H peak at 3441 cm-1. There should be a shift in the IR peaks for the N-H from polyamide-4,6 and the N-H peak from the carbazole if the carbazole molecules replace polyamide N-H in the hydrogen bond structure. The carbazole N-H will then become bound and the polyamide N-H will become unbound. There should have been shifts in both towards each other of about 10 cm-1 if there were any substantial complexing of the two materials with hydrogen bonding. Guerra et al. [142] found shifts of 58 cm-1 in N-H stretching band maxima as they altered the percentages in their hydrogen bond interacting blends. The N-H stretch for polyamide-6 film increases by 18 cm-1 in being heated from 50 to 227.5 0C in work by Xu et al. [143] due to the reduction in bound hydrogen bonds and a move to less restricted N-H bonds. The same paper shows a shift to the right in the melt of polyamide-6/LiBr compared with pure polyamide-6 because the amide-amide hydrogen bonds are supplanted by the intense ionic bonds with the salt. Gao

and

Scheinbeim

studied

interactions

between

Nylon-11

and

poly(vinylidene fluoride) (PVF2) [144]. They found a shift in the N-H stretch by up to 8 cm-1 as the level of PVF2 was increased. This shift was to lower wavenumbers because the F…N-H hydrogen bond was stronger than the C=O…N-H bond. That is obviously in the opposite direction to that expected 64

if the N-H of carbazole or phenothiazine were to supplant the amide N-H bond to O=C on another amide group thus freeing up an amide N-H. It is therefore a useful benchmark for the type of change expected. Skrovanek et al. also looked at semicrystalline Nylon-11 considering the effects of temperature increases leading to the melt [61]. They found a shift in the peak of the main N-H stretch of 32 cm-1 to higher wavenumbers in that process as the temperature was raised and the hydrogen bonds weakened.

In their case the normalised area of the peaks reduced in

sympathy with the temperature increase. Wang, Ma and Wu [145] solution blended polyamide-6 or polyamide-6,6 into “Novolac”, a phenolic resin.

The aim was to reduce the brittleness of the

Novolac by using intermolecular hydrogen bonding of the materials in the blends. They did not specify explicitly in the paper the extent of the FTIR frequency shifts they found but their figures 5 and 6 plotting the spectra for various blends make it clear that substantial shifts have, in fact, taken place. The O-H of the Novolac has changed by something in the order of 70 cm-1. The focus of this Mid-Range IR work will be on the N-H stretch as that is the major area where the disruption of C=O….H-N(amide) with “free” N-H(diluent) to produce C=O….H-N (diluent,-“bound”) and “free” N-H(amide) would be expected to have an effect. Peak frequency was used as the determinant of N-H changes rather than the more risky deconvolution of non-linear PAS signals of composite spectra from different materials (vide infra).

The

results, above, from other authors’ work gave the confidence that this would be suitable to discriminate changes in hydrogen bonding activity. 1.5.4.3

Mid Range IR Frequencies of Interest

The relevant absorption frequencies for FTIR investigations described in Chapters

3

to

10

for

the

different

combinations

polyamide-6, polyamide-6,9 or polyamide-6,12

of

polyamide-4,6

with carbazole and with

phenothiazine are brought together in Appendix C: FTIR Assignments to avoid undue repetition. The Photoacoustic (PAS) spectrum of polyamide-4,6 is seen in Figure 1-25 below. The other polyamides in the study have peaks that are close but not

65

quite identical to the above. The slight differences across several absorbing bands can be used to positively identify polyamide types. 65

Polyamide-4,6

60 55

Photoacoustic

50 45 40 35 30 25 20 15 10 5 3500

3000

2500

2000

1500

-1

1000

500

Wavenumbers (cm ) Figure 1-25 Mid Range IR spectrum of polyamide-4,6 from an ampoule

The bands for carbazole are shown in the PAS spectrum of Figure 1-26 and for phenothiazine are shown in Figure 1-27. 50

Carbazole

Photoacoustic

40

30

20

10

3500

3000

2500

2000

-1

Wavenumbers (cm )

1500

1000

Figure 1-26 Carbazole photoacoustic FTIR peaks in the Mid Range IR

66

500

Phenothiazine

Photoacoustic

200

160

120

80

40

3500

3000

2500

2000

-1

1500

1000

500

Wavenumbers (cm ) Figure 1-27 Phenothiazine photoacoustic FTIR peaks in the Mid Range IR.

1.5.4.4

Mid Infra-Red Data Analysis for Blends

The original PAS spectra of polyamide-4,6 (Ampoule 64) and carbazole (Ampoule 63) are overlaid in Figure 1-28 with the spectrum of the ampoule material 66PA46Car (Ampoule 31) to demonstrate an FTIR analysis problem. 65 60

Carbazole Polyamide-4,6 66PA46Car

55 50

Photoacoustic

45 40 35 30 25 20 15 10 5 4000

3500

3000

2500

2000 -1

1500

1000

500

Wavenumbers (cm ) Figure 1-28 PAS spectra in Mid IR for carbazole, polyamide-4,6 and a blend.

It can be seen from Figure 1-28 that each of the spectra for the raw materials has a large number of sharp peaks. The spectrum for ampoule 67

material from a blend takes on approximately the combined peaks of the two raw materials. The peak from one constituent material of a blend may lie on a sharply rising or falling portion of the other material’s spectrum.

The

combined effect can result in a shift in peak frequency even if there are no changes in hydrogen bond interactions or morphology due to blending. The Photoacoustic technique is usually non-linear for strong peaks.

Spectral

additivity cannot normally be expected. There was a conundrum, however, because there was a large problem here with the interpretation of spectra. The following mathematical modelling was employed in an attempt to see if the Infra-Red spectra indicated interaction between the polyamide and the small molecules. The spectra of the two constituents were mathematically added in a proportion that mimicked the salient features of the spectrum of the blend material. It was done in order to look for regions where the blend spectrum was different from that expected for no interactions involved. Differences between the model and experimental results could potentially be indicative of frequency shifts. It was reasonably strong evidence for no hydrogen bond interaction or crystallographic/morphology changes to have taken place if the “model” and experimental peaks matched up precisely. Any artefact caused by the simple model would have to be exactly the same magnitude but of the opposite direction to actual chemical shifts, an unlikely scenario.

The importance of relative height changes of double and treble

peaks was considered low. Regions of each constituent spectrum were chosen where there was a significant peak in one material but not in the other. The spectra of the two materials were mathematically added and scaled to match the spectrum of the ampoule material at both points. Sometimes more points were selected to assist in the match.

Generally the peaks that were chosen were ones

where the signal for one material was reasonably high and the other material had a low signal at that point. The highest peaks were not chosen as primary ones as they were likely to be non-linear due to sensor saturation.

The

match to the actual identified peaks would thus be due to having the correct proportions of each material in the model. Examples of the spectral regions chosen are given in Figure 1-29. Secondary peaks have the spectrum of the other material is rising or falling strongly in that region or the peak is a very high one which is likely to be truncated by signal saturation. 68

30

Carbazole

28

Polyamide-4,6

Polyamide-4,6 peak and low carbazole absorbance

26 24

Photoacoustic

22 20 18 16 14 12 10

Confirmatory peak for Polyamide-4,6

8 6

Carbazole peak and low Polyamide-4,6 absorbance

4

Carbazole confirmatory peak

2 4000

3500

3000

2500

2000

1500

1000

500

Wavenumbers (cm-1) Figure 1-29 Peaks used for modelling polyamide-4,6/carbazole blend spectra in Mid IR.

An example of a model compared with a measured spectrum for a blend is given in Figure 1-30. 28

*Addition* model 23PA46Car

26 24

Photoacoustic

22 20 18 16 14 12 10 8 6 4 4000

3500

3000

2500

2000

1500

1000

500

Wavenumbers (cm-1) Figure 1-30 Mid IR PAS spectrum of 23PA46Car from Ampoule 57 with the model constructed from the spectra from polyamide-4,6 and carbazole.

This is expanded for the carbazole N-H stretching peak and the peak heights equalised in Figure 1-31. 69

*Addition* model 26

23PA46Car

24

Photoacoustic

22

20

18

16

14

12

3450

3440

3430

3420

3410

Wavenumbers (cm-1)

3400

3390

3380

Figure 1-31 Carbazole N-H stretch for 23PA46Car and model expanded from the comparison in Figure 1-30 and equalised in height.

PAS does have a disadvantage that strong peaks will cause signal saturation resulting in some truncation of peak heights. That applies particularly to the

major

peaks

for

spectra

of

the

constituent

materials.

The

mathematically combined spectra will therefore also give partly compressed tops of major peaks. These peaks are then expanded vertically in the graphs to give peaks that can be easily compared for peak frequency with the measured ones for the polyamide/diluent blend material from the ampoules. The rounding of the peak tips, as seen in Figure 1-31, is an artefact of expanding the models based on non-linear spectra to give the same peak heights for comparison.

The reason the models are built primarily on

significant peaks that are not the largest peaks for each material is to construct the model minimising the effects of signal saturation. The spectra for polyamide/diluent blend samples are less compressed at peak tops compared with the constituent spectra that the models were based on. Each highly absorbing peak is less saturated in the measured blend spectrum. Peak signals for blend materials are reduced in intensity by dilution because of the presence of the other material in the sample. The peak signals are then more linear because there is less detector saturation. The positions of peaks are the critical issues rather than the rounding. 70

It can be seen above that analysis without mathematical modelling would be extremely difficult but simple mathematical addition of spectra can result in artefacts due to non-linearity in the major spectral peaks for the constituent materials.

A comparison between a measured blend spectral peak and a

model from spectra of the constituent materials with no change in the peak frequency should be strong evidence for the materials not interacting. Slight height differences between sharp measured peaks and a model based on the spectra of the constituents may or may not be indicative of an interaction, given the non-linearity of the detector system. 1.5.4.5

Near Infra –Red FTIR (NIR)

The NIR region is noted in the literature [141] for being sensitive to the hydrogen bond status. The broad FTIR peaks for hydrogen bonding can be examined in the Near Infra-Red (NIR) at moderate sensitivity and resolution. The area of interest is in the 7500 to 4000 cm-1 region.

The Nicolet 750

instrument can be set up for the Near Infra Red region with wavenumbers between 11,000 and 3,000 cm-1. This requires some changes to the physical configuration of the instrument regarding the beam splitter, light source and detector. The DRIFT technique was the most appropriate. Wu and Siesler [141] studied polyamide-11 in this band. The frequencies in the Near Infra-Red they found were 6912, 6600,6390, 6290, 6180, 4940, 4846, 4580 and 4560 cm-1.

Values found with the polyamides here were

Absorbance

close to those values. A typical polyamide spectrum is shown in Figure 1-32.

7500

-1

Wavenumbers (cm )

4500

Figure 1-32 Typical polyamide DRIFT spectrum in the NIR region.

No NIR peak values were located in the literature for carbazole or phenothiazine. 71

The spectra for these two are shown below in Figure 1-33

Absorbance

carbazole phenothiazine

7500

Wavenumbers (cm-1)

4500

Figure 1-33 Carbazole and phenothiazine Near Infra Red spectra as measured with DRIFT.

The peaks for the polyamides compared to those of the diluents are quite separate and are not so steep as in parts of the Mid Range IR so it was not necessary to resort to the mathematical additions of spectra. There was a facility in the Omnic software used to drag an individual spectrum from a group up or down to match the height on screen of another spectrum peak. That facility was used to move each peak from the spectrum for a blend to match the equivalent peaks for those of the spectra for the constituents. 1.5.5


Small Angle X-Ray Scattering (SAXS) can be used to gain information about semicrystalline polymers at a dimensional range around the size of lamellae and relates to their stacking within a solid.

Periodically stacked lamellae

reflect X-rays more strongly than amorphous regions because of the higher density (and therefore electron density). This occurs with Bragg reflections at scattering angles of 2Θ according to the well known relationship of sin Θ = mλ/2d where m is an integer giving the order of scattering, λ is the

wavelength of the X-rays (in this case) and d is the periodicity distance of the scatterers.

For example, an X-ray wavelength of 0.154 nm for first order

scattering giving a peak at 2Θ of 0.80 equates to a scattering periodicity near 12 nm. An opportunity arose early in the making of ampoules to have some samples measured with SAXS by another University of South Australia PhD student who was working briefly at Connecticut University. He was able to carry out a very restricted number of trials because, at that stage, only a handful of 72

ampoules had been produced. The conclusions from these are limited but the results have been included for the benefit of other researchers. 1.5.6

Solid state Nuclear Magnetic Resonance Spectroscopy

FTIR, discussed above, looks at how the frequencies of interatomic bonds are influenced in their various modes of vibration by the atoms at the ends of the bonds and by the near-neighbour environment.

Nuclear Magnetic

Resonance (NMR) is interested in the nuclei of atoms and how they are influenced by near-neighbours. The nuclei of 1H,

13C 15N

and some isotopes

of other elements act as if they are spinning. It is possible, by placing them within strong magnetic fields and irradiating them with radio frequency (rf) electromagnetic radiation, to have them absorb energy.

They appear to

precess in the way the top of a spinning top gyrates in slower circles as the top spins. The combinations of magnetic field and rf frequency where the absorption occurs can be used to determine the environments of the nuclei. For example, the five hydrogen atoms attached directly to a benzene ring having an attached O-H group will have absorptions at three slightly different frequencies, one for the H atom directly opposite the O-H bond, there will be a peak twice as large and at a slightly different frequency for the two atoms next to it and another peak the same size as the previous one for the two hydrogen atoms attached to carbon atoms either side of the O-H group. There will also be a peak of the same size as the original peak, but at a noticeably different frequency, for the hydrogen atom attached to the oxygen atom of the O-H group. produce different

13C

Similarly differing

13C

environments will

peaks depending on the environments of the nuclei.

The measurements can either be carried out by keeping the rf irradiation constant and modulating the strong magnetic field or by keeping the magnetic field constant and varying the rf field. Usually the latter approach is taken nowadays with the advent of strong superconducting magnets and the ability to give a sharp pulse of rf radiation which populates all the atoms at the same time. The response as a function of time is deconvoluted to give the final output. These measurements can be carried out in solution or of solid materials. Solidified

small

organic

molecules

have

sharp

peaks

because

the

environments around the nuclei are quite regular. Even highly crystalline polymers always have a considerable amount of disordered amorphous 73

polymer material outside the spherulitic regions and in the interlamellar spaces. This leads to the absorption peaks being smeared out. It is often possible, however, to infer things from solid state NMR where the material is being left in its original morphological state. It was for this reason, that the opportunity to have solid state NMR measurements made on some blend material from ampoules was taken up in order to better understand the materials we were working with. A brief window in time occurred after the production of Ampoule 1 for NMR measurements to be carried out by Dr. Andrew Whittaker at Queensland University on red and white sections from the ampoule plus polyamide-4,6 and carbazole powders. The white sample proved too hard to make into a fine powder at the time so only the red blend could be measured along with the constituent powders. The results proved ambiguous. They have been included in Chapter 3 to make them available for other researchers.

1.6 Structure of the Thesis The thesis covers much work that is of the same structure from chapter to chapter covering different material combinations. A brief description of the various chapters follows. Chapter 2: Experimental The second chapter covers all experimental details and scant reference is made in other chapters to these details. Chapters 3 to 10: The polyamides combined with the diluents These chapters cover the experimental work on various combinations of polyamides with either carbazole or phenothiazine. A chapter is devoted to each

combination.

The

polyamides

polyamide-4,6,

polyamide-6,

polyamide-6,9 and polyamide-6,12 in that order are combined firstly with carbazole.

The last four chapters of this block cover the above four

polyamides combined with phenothiazine. There is much that is similar from chapter to chapter within this group of chapters but there is also much information that is different from one material combination to another.

As a result of the consistent approach

taken in the experimental work and in analysing the data, the chapters could appear repetitive. This will not be helped where the outcomes from one material combination to the next happen to be similar. 74

The chapters give only low-level conclusions on the results seen. This is the most appropriate because often the outcomes of experimental work on one or more combinations of the various polyamides with either of carbazole or phenothiazine are best evaluated together in the final conclusions chapter. Chapter 11 General Conclusions This chapter draws the previous eight chapters together in overall conclusions. The differences between the polyamides in the way they interact in high temperature solution and in crystallising to solid blends are covered in the context of the molecular structure of the individual polyamides. The common aspects are more oriented to how polyamides generally interact with these specific small molecules. This work has covered a reasonable tract with non-isothermal DSC and FTIR and could be considered as a pilot study. There are a number of aspects that could be pursued to further the scientific understanding and to pursue applications of this research.

A list of questions that the work raises is

provided in the hope that the opportunity will arise for them to be investigated. Appendices Appendix A: Further details from DSC thermograms Appendix B: Lissajous Figures for Understanding Temperature Modulated Differential Scanning Calorimetry of Nylons Appendix C: Mid

Range

Fourier

Transform

Infra

Red

Spectroscopy

Assignments Appendix D on CD: Fourier Transform Infra-Red spectra (PDF format) of blends

with

mathematical models

or

with

spectra

of

constituent materials. The CD contains the whole thesis Bibliography Bibliography of all references in the thesis.

1.7 Summary This introductory chapter to the thesis explained how aliphatic polyamides, commonly called nylons, are an important class of engineering polymers, that it is important to understand their properties more fully to utilise them 75

to best advantage and how this work contributes to the virtually untapped knowledge of their characteristics in high temperature solutions with low molecular mass diluents. It went on to take the reader through from the background information on mixtures of materials, hydrogen bonding, the structure, melting and crystallisation of semicrystalline polymers, and led to the specific case of aliphatic polyamides.

This continued into a survey of

some of the recent literature in areas adjacent to the specific area of interest, leading to a description of the research problem.

A description was then

given of techniques suited to investigating the research problem and the reasons. The details of expected outcomes of certain techniques were provided in some cases including the “TLS peak” often found when heating blends in the DSC and of mathematical modelling blend spectra in the Mid IR range using the spectra of the constituent materials. It is now time to look at Chapter 2 with its description of the experimental conditions used for the techniques.

76

Chapter 2

EXPERIMENTAL CONTENTS 2.1

Introduction

77

2.2

Materials

78

2.3

Materials handling

78

2.4

Preparation of Melt Blends

81

2.4.1

Melt Blending in DSC Pans

81

2.4.2

Melt Blending in Ampoules

82

2.5


86

2.6


87

2.6.1

Calibration materials and their preparation

87

2.6.2

Calibration

88

2.6.3

Experimental trials

89

2.6.3.1 2.6.3.2 2.6.3.3

2.6.4

2.8

90 90 91

DSC Thermogram evaluation

2.6.4.1 2.6.4.2 2.6.4.3

2.7

Melt Blending in Pans Melt Blending in Ampoules Additional trials covered in Appendix B

91

Melting and Crystallisation temperatures Enthalpies from Graphical Analysis Crystallinity calculations

91 91 92

Simultaneous Differential Thermal Analysis-Thermogravimetric Analysis

92


93

2.8.1

Mid Infra-Red Range Measurements

93

2.8.2

Near Infra-Red Range

94


95

2.9

2.10 Cross Polarised/Magic Angle Spinning Solid State 13C Nuclear Magnetic Resonance Spectroscopy 95

2.1 Introduction Many of the experimental details covered here are common to all combinations of the four polyamides with the two compounds melt blended with

them.

The

eight

major 77

chapters

covering

the

various

polyamide/compound combinations do not have experimental detail sections because everything is covered in this section.

2.2 Materials Polyamide-6

(Catalogue

number

18,110,0),

polyamide-6,9

(Catalogue

number 18,806,9) and polyamide-6,12 (Catalogue number 18,114,5) were obtained as commercial grades from Sigma Aldrich. None of these grades were known to have any additives, which could potentially affect the blending process. These polyamides came as pellets approximately 3 mm in diameter and 3 mm long. The size was not amenable to many of the ways the polyamides were to be used. A coarse powder would be more useful. The granules were very tough, being polyamides. Various attempts using different methods were tried in order to make smaller particles. The most successful was to freeze the granules with liquid nitrogen and grind at high speed in a Breville CG-2 electric coffee & spice grinder having a flat metal blade. The coffee grinder had the lower inside casing of metal with a transparent polymer lid. A stainless steel plate was cut to just fit on top of the lower metal part in order to provide protection to the polymer lid. The fit of the plate was not perfect. This allowed the fine particles swirling around during grinding to slip through and settle on the metal plate.

Automatic

separation of coarse and fine powder from the larger particles and granule pieces being ground below was thus achieved. It was necessary to continually add small amounts of liquid nitrogen to keep the granules brittle. Dr. Cor Koning (formerly) of DSM Research Laboratories was kind enough to provide 1 kg gratis of polyamide-4,6 having no additives. This material was provided as a coarse powder suitable for (near) direct use. A range of potential hydrogen bond disruptors was obtained mainly from Sigma Aldrich.

The two which were pursued in the main part of this

research project were carbazole (Sigma Aldrich catalogue number C310,3) and phenothiazine (Sigma Aldrich catalogue number P1,483,1).

These

materials were provided in fine powder form.

2.3 Materials handling Polyamides are well known [99 p. 324] to absorb moisture into the amorphous portion of their bulk. The amount absorbed is 1-15%, depending upon the type of polyamide, level of crystallinity, temperature and relative 78

humidity [48 p. 358]. portions of samples.

This moisture resides solely in the amorphous Moisture is unable to properly penetrate existing

lamellae due to their tight interlinking by hydrogen bonding [122, 144, 145]. The moisture causes hydrogen bond disruption in the amorphous part, acting on either the carbonyl or N-H groups [48 p. 360]. Absorption into the amorphous portion can have very strong effects on the mechanical properties of polyamides, acting as a plasticiser [146, 147]. The above characteristic of polyamides was the reason that all samples were handled right through the various processes so that moisture ingress would be at a minimum. Materials were stored in a glovebox with low moisture content. The materials were weighed out in the glovebox. Any handling of raw materials or samples outside the glovebox was done in a way to minimise contact with (moist) air. Thermogravimetric Analysis (TGA) results on polyamide-6 in wet and dry states are presented in Chapter 4. It provides the necessary supporting evidence for the approach taken. A Vacuum Atmospheres Company’s Drilab double glovebox was available for storing materials and preparing samples. The glovebox was maintained at less than 7 ppm moisture during the whole of the period of this research. Initially it had been filled with Ultra High Purity argon gas (99.999% pure) from BOC but during the last two months of the project it was filled with High Purity nitrogen (99.99% pure) from the same supplier. This change is not considered to pose a threat to the validity of any of the experimental work carried out here. The process, mentioned above, of creating powder from granules was carried out for convenience outside the glovebox.

Polyamide powders were then

dried under vacuum in the vacuum oven antechamber of the glovebox. The materials were placed in the vacuum oven in glass jars and held at close to 110 0C for a protracted period. The chamber was pumped down with an Edwards RV8 two-stage rotary vane pump until temperature and the vacuum level were achieved and then the vacuum pump turned off after the valve to the chamber was closed.

This meant that there was no

backstreaming of oil vapours from the pump into the vacuum oven chamber and onto the samples being vacuum-dried. Backstreaming could otherwise have occurred during the extended time the powders were being dried. The 79

heater was controlled at the required temperature.

Cooling water was

passed through the chamber walls and door to keep them cool. This had the effect of capturing any moisture released from the material being dried. A pressure of approximately 150 Pa was maintained during the drying process. Vacuum drying was initially carried out for 96 hours. That was dropped to 16 hours in later polyamide drying after discussions held in April 1999 with Dr. R. Gaymans of Twente University in the Netherlands.

Vacuum dried

materials were transferred directly inside the glovebox from the vacuum oven antechamber.

`The materials were stored in the glovebox until required.

Caps of the jars were left loose to allow equilibration with the glovebox interior. Polyamide-4,6 powder from DSM was opened in the glovebox. The material was vacuum dried in the same manner as the other polyamides so that a consistent environment was common for all polyamides.

The dried

polyamide-4,6 powder was stored afterwards in the glovebox. Seals of all potential hydrogen bond disruptors were opened in the glovebox and the material containers left open for a week with a cover approximately 1 cm above the rim to allow moisture stabilisation and reduce potential ingress of any dust. An AND model HA-180M balance was available within the glovebox for weighing materials. It was generally stable to approximately 0.2 mg but at times drifted by several milligrams over a short period. It was found best to allow it to stabilise for at least several hours when the balance started drifting substantially. An hermetic pan crimp was available within the glovebox. Pans and lids for the DSC were weighed on a Cahn microbalance Model C-34 outside the glovebox before being brought into the glovebox. The Cahn instrument gave reproducible measurements to within 10 µg. The mass was marked with a fine permanent ink marker on a stoppered glass vial container used for transport of the pan.

Materials were added in the glovebox with the less

sensitive AND balance there, giving approximately the desired amount. Extra care was taken for the cases where more than one material was being added to a pan.

The balance was checked for zero before and after

measurements to ensure that drift had not occurred. Measurements were 80

repeated until stable results could be obtained. Pans were crimped in the glovebox and reweighed later on the more accurate Cahn microbalance. The weight of the sample prior to DSC was also marked on the vial. TGA samples of 10 to 15 mg were weighed on the AND balance. Sidecutters were used to cut TGA samples from larger pieces. Tweezers were used for handling the small pieces involved. A scalpel was not used for cutting small pieces because of a risk of damage to gloves and injury to the operator. The sidecutters also had the advantage that they tended to contain the small pieces when cutting hard materials. Pieces for TGA were stored in vials with tops and the vials marked with the origin of the sample. These samples in airtight vials could safely be stored outside the glovebox until ready for use because the gas inside was dry from the glovebox environment.

2.4 Preparation of Melt Blends There are two basic ways to make blends, from solution or from the melt. There were initial attempts to create complexes from solution. This had been on the advice that melt complexing would be difficult due to problems in finding suitable materials where evaporation or degradation were not serious issues at the working temperatures involved for melt blending. The solution work was not successful in creating complexes. That work is not covered in the thesis. The melt blending proceeded firstly with experiments in DSC pans to evaluate the possibilities on a small scale and then to larger scale experiments in glass ampoules once the small-scale work looked promising. 2.4.1

Melt Blending in DSC Pans

Initial work utilised the temperature controlling facilities of a DSC in order to carry out melt blending. The advantage was the ability to thermally monitor potential complexation with DSC during the actual melt blending process. There was a disadvantage in the small amount of blended material being produced (approximately 0.01 g). The hermetic pans were accurately weighed prior to introduction into the glovebox.

As close as possible to 5 mg of the relevant vacuum dried

polyamide powder was weighed into the pan using the balance in the glovebox. Either 1, 3, 5 or 8 mg of the potential hydrogen bond disruptor was added on top of the polyamide. The pan was crimped hermetically in 81

the glovebox using the appropriate closure tool. The pan was then weighed accurately again outside the glovebox using the more accurate balance outside. Melt blending in the DSC was initially carried out at 5 0C/min heating rate followed, after 7 min in the molten state, by crystallisation at 25 0C/min. This was done for some of the blend ratios mentioned above and also for the individual starting materials as a reference. Some DSC runs were also carried out for a cooling rate of 2 0C/min instead of 25 0C/min in order to give the molten mixture more time to form crystals. This was under the more adverse conditions of having the diluent evaporating from the “hermetic” pan for longer periods at elevated temperatures. This was not carried out extensively because the project had already moved from blending in pans to making greater amounts of material at the same time in ampoules. The disadvantages of carrying out the process in the DSC are that the amount that is produced by this method (6 to 13 mg) is far too small for most analysis techniques such as Fourier Transform Infra Red, Nuclear Magnetic Resonance and Small Angle X-ray Scattering.

There is also a

problem in that the different analysis techniques use differently configured samples in terms of particle size and form. To avoid these difficulties, blends of larger size were made in ampoules. 2.4.2


The ampoules used were made from thick walled glass tubing 16 mm Outer Diameter (OD), 2.5 mm wall thickness and sealed at one end. They were 250 to 300 mm long with the middle 80 to 100 mm necked down to approx 3 to 4 mm Internal Diameter at the narrowest. The polyamides used had very low percentages of moisture because raw materials were kept in a glovebox with either dry argon or nitrogen. This meant that the polyamide tended to become electrostatically charged and stick to the walls of the glass ampoules.

A solution was found to the

problem of filling ampoules by first putting the diluent in the ampoule. That left a very fine layer of powder on the glass. The polyamide powder being added later would slip easily through to the bottom section and not cause problems in adhesion to the glass walls. 82

The required quantities of diluent and polyamide were weighed out into containers using the AND balance in the glovebox. The materials could then be fed into the top section of the ampoule to make the blending charge. The vast majority of powders adhering to the inner walls of the top section could later be scraped down through to the neck section using a thin spatula with a bent, rounded end. There was usually a very small amount of material adhering to the inner walls of the top and neck sections. This would have not substantially altered the ratios of materials in the charge, as that was relatively large at 1.5 g.

A suitable polyethylene vial stopper was fitted to

the top to seal against moisture ingress when the ampoule assembly was later removed from the dry glovebox. Sealing the ampoule neck was done outside the glovebox. The stopper was quickly removed and a vacuum hose attached to the top opening. The hose led, via a liquid nitrogen cold trap, to a rotary vane vacuum pump.

The

ampoule was evacuated for at least 10 minutes despite the small volume of the ampoule.

That was because the small neck in the ampoule would

restrict achieving a good vacuum within the volume where the charge was situated.

The bottom section of the ampoule (containing the charge) was

lowered into a thermos flask filled to the very top with liquid nitrogen. The charge was held at least 20 mm below the liquid nitrogen surface by a clamp so that glass above the charge was at liquid nitrogen temperatures. That meant the ampoule sealing operation would not degrade the charge materials.

A large natural gas/compressed air Bunsen burner was then

used to heat the ampoule neck whilst the system was still attached to the vacuum system. The glass eventually softened sufficiently for the neck to collapse under the vacuum. The ampoule body with charge could then be sealed off and removed.

The materials in the charge were then manually

shaken and turned over for five minutes to mix them thoroughly after the ampoule body had returned to room temperature. A

custom-made

furnace,

manufactured

by

Scientific

Equipment

Manufacturers, and capable of over 1000 0C was available for the trials. The furnace cavity was 150 mm wide, 270 mm deep and 90 mm high. There was a thermocouple controlled BTC-8080 PID controller with the capability of heating at a predetermined rate to a predetermined temperature. There was 83

also a separate sensor with power cut-out for thermal runaway events and an exhaust vent. The controller could not be set to cool apart from manually being set to lower temperatures in a stepwise manner. The first ampoule (see Chapter 3) was essentially carried out as a “dry run” to determine where all the problems with blending in ampoules would lie. The ampoule was mounted in a metal tube (against a possible explosion) in a nearly horizontal position supported above the furnace floor. That ampoule had overshot the 300 0C set temperature and had gone to 320 0C due to thermal inertia in the system. The temperature setpoint was quickly lowered manually to 280 0C and set 20 0C lower every 10 minutes. The furnace had dropped quickly in the initial cooling stages but soon lagged well behind the regular manual drops in setpoint. The outcome with this ampoule was the solid being coloured quite differently in three sections along its length, as described and shown in Chapter 3. Attempts over the next seven ampoules to repeat the separation into differently coloured sections of differing compositions were unsuccessful. The alternative strategy was to make the materials deliberately with the desired compositions in a very controlled manner.

The first step was to ensure that the ampoule was exactly

horizontal so that both ends would cool at the same rate. Thermal control over the process also had to be improved.

This led to

replacing the controller with a 7-step Temptron controller that could be programmed with FCLink software.

The software was run on a laptop

computer connected to the controller with an RS-232C serial connection. The controller chosen had the possibility of switching an alarm function on or off for each programmable step. That possibility allowed the alarm signal to be connected to a solenoid. Compressed air was run to the furnace via the solenoid and introduced to the furnace cavity via a metal pipe placed in the exhaust vent. The air entered one side of the furnace and circumvented a baffle.

Air left on the other side via the exhaust vent after passing the

ampoule.

The compressed air could work against the heater to provide

better heating and cooling control.

The maximum temperature during

furnace runs was 300 0C and the furnace was designed for quick heating to over 1000 0C so forced air cooling at 2 0C/min was not considered deleterious to the ceramic furnace interior. 84

Constant cooling at 2 0C/min

with the compressed air could be maintained down to less than 100 0C although it was usually only continued at that rate to 140 0C in order to shorten blending runs. Crystallisation activity was normally completed by 140 0C. The existing encased thermocouple was replaced with a thin wire one placed close to the ampoule to ensure that the controlled temperature represented the temperature of the material in the ampoule as closely as possible.

The use of a closed, thick metal tube to protect people from

potentially exploding ampoules was also not in the interests of good thermal responsiveness and control. The use of the tube was dropped after the first few ampoules were produced, as the risk of damage to the ceramic furnace liner appeared low and better thermal control could ensue. Instead, a lock and a notice were put on the door opening mechanism of the furnace during the melt blending process. It can be noted that none of the ampoules broke in the furnace although the potential for breakage or explosion still existed. The profiles used normally with the furnace were to heat the ampoules at 5 0C/min, hold the temperature at 260 0C for 1 hr before a sharp excursion to 325 0C in the melt for 5 min and cooling at 2 0C/min. Initial ampoules were varied and carried out with various profile variants and manual cooling. Results presented in the various chapters (with the exception of Ampoule 1 in Chapter 3) are from later ampoules using the temperature controller and forced-air cooling. Profiles were varied in the very early trials to achieve the most consistent composition throughout the sample as determined by Thermogravimetric Analysis (TGA). The results of the first ampoules had showed that having a circular cross section to the ampoule made the bulk material harder to present samples for FTIR and other techniques. The cooling of the material would also have been uneven, with the middle part of the cross-section thicker and therefore cooling at a different rate. It was better to have a relatively thin bulk sample of even thickness and spread out over a larger area if consistent material were to be formed. This led to most of the material coming from the trials being made with ampoules having one side flattened during manufacture. The outcome with an accurately horizontal ampoule in the furnace was a piece of near rectangular bulk sample with a nearly consistent 2 to 3 mm thickness from 1.5 g of constituent materials. 85

The ampoules from melt blending had a vacuum inside when back at room temperature because of the original sealing process. An inrush of moist air at opening would most likely cause adsorption of moisture that would later be absorbed into the material. Opening the ampoules could not be done in the glovebox because of the difficulty in breaking the ampoules open. Opening was carried out in a plastic bucket filled with cold dry nitrogen and covered with a sheet of plastic. The nitrogen was from liquid nitrogen in the bottom of the bucket. The ampoule had previously been scored around the middle with the edge of a triangular file. The ampoule was held under cover in

the

bucket

splinters/shards.

with

rubber

gloves

on

as

protection

from

glass

It was hit with a hammer to break the glass and the

pieces of blend material placed immediately in a jar that was quickly closed. Jars were put as quickly as possible in the glovebox via the antechamber and opened inside to the dry atmosphere. Material was stored in stoppered vials in the glovebox after a stabilisation period of several days. The pieces of material from the various trials could be kept permanently in the glovebox until cut up with sidecutters for the various trials.

2.5 Thermogravimetric Analysis Thermogravimetric Analysis (TGA) to determine the polyamide weight percentage in ampoule samples was carried out with a TA Instruments Hi-Res

Modulated

TGA

model

2950

Thermogravimetric

Analyser.

Experimental work was carried out over an extended period and several upgrades of the Instrument Control software were made during that period. Software changes were unlikely to have affected the experimental outcomes. Samples of material from the ampoules were cut to 10 to 15 mg total weight in the dry glovebox where the ampoule material was stored.

Pieces were

placed in small, stoppered glass vials marked with the contents.

Some

samples were crumbly and some were difficult to cut to near the desired weight range because they were so hard. Attempts were made to have the TGA samples in one (or two) pieces but that was not always possible. TGA samples were taken from bulk ampoule material next to where DSC samples were removed so that the effects of any composition variations in the bulk sample would be minimised. An AND model HR-180 balance in the glovebox was used for weighing samples to better than 1 mg accuracy. The vials were 86

removed from the glovebox and TGA was carried out within 24 hours to avoid any potential moisture leak past the seal on the plastic top. The TGA stirrup with an open Standard DSC Aluminium pan was tared just prior to the experiment.

The vial was opened minutes before TGA was

carried out, the sample placed in the DSC pan on the stirrup (on the loading table) and the experiment run. The DSC pan was used to minimise contamination of the stirrup. The purge gas was High Purity nitrogen (99.99%) from BOC Gases used at a flow rate of 50 ml/min in both gas streams.

TGA was run from room temperature to 500 0C at 10 0C/min.

Initially all measurements were carried out in High Resolution mode with settings

Resolution = 4

and Sensitivity = 4.

Later

in

the

series

of

experiments, duplicate samples were taken so that TGA could also be carried out ramping straight through from approximately 25 to 500 0C at a fixed 10 0C/min.

The two samples were taken from either side of where DSC

samples were removed from the bulk Ampoule material.

The High

Resolution runs are identified as “R4S4” and the others as “Straight” in the text. Generally, phenothiazine and carbazole material started evaporating at 175-200 0C in a TGA instrument. The weight remaining reached a plateau before 275 0C as the carbazole or phenothiazine evaporated from the sample. Degradation of the polyamide began above 325 0C at the ramp rates involved.

The weight percentage remaining at 300 0C was taken as the

percentage of polyamide in the sample.

2.6 Differential Scanning Calorimetry Differential Scanning calorimetry (DSC) was carried out on a TA Instruments DSC 2920 Modulated DSC instrument with their Instrument Control software running on a PC. 2.6.1

Calibration materials and their preparation

Pure indium wire for Cell Constant and Temperature calibration was obtained from TA Instruments. 10.515 mg of indium wire was then weighed into an aluminium hermetic DSC pan as supplied by TA Instruments. Fine Al2O3 powder from TA Instruments was dried by holding it in a furnace at 600 0C for 30 min under an oxygen atmosphere. Approximately 6 mg of this Al2O3 powder was quickly placed in an hermetic DSC pan and the lid 87

crimped. The powder was lightly pressed by the lid to the bottom of the pan for good thermal contact. 2.6.2

Calibration

Calibrations were run before every series of experimental runs. A baseline was run at the same ramp rate as experiments were to be run (5 0C/min) over the range 25 to 320 0C.

Cell Constants were calibrated at 5 0C/min

using indium with its known heat of fusion 28.45 J/g [128 p. 144] . This calibration ensures the correct enthalpy value when integrating peaks in thermograms of samples. The very sharp melting peak of indium was also used to calibrate temperature.

Experiments were run using Temperature

Modulated DSC so that “Reversing” and “Non-reversing” signals could also be extracted if needed. Calibration of the Heat Capacity Constant for the experiments was done using dried Al2O3 as a calibrant. The range of interest was from 100 to 300 0C. The calibration coefficient varies by 5% even over the very limited range 170-250 0C. Calibration was done under the same conditions as those used for experiments, viz 5 0C/min average ramp rate with a period of 30 s and amplitude of 0.41 0C. The short cyclic period was necessary because some transitions had been found to occur sharply and there should be four cycles over a transition [129 p. C-32, 148] to avoid artefacts in the thermograms. The amplitude was the maximum that kept a heating ramp from temporarily becoming cooling during a portion of each cycle. The maximum amplitude was chosen in order to enhance sensitivity for extraction of the “Reversing” signal. Helium gas was chosen for the purge gas because nitrogen has too slow a heat transfer to the pans. See Appendix B for a discussion of the approach taken to TMDSC. The flow rate in both gas streams was 50 ml/min. There was no opportunity within the TA Instruments DSC Instrument Control software used to change calibration settings during the actual course of the runs.

The cooling parts of the runs were therefore done

without using temperature modulation as it is recognised that the cooling parts are not calibrated properly. That applies, in general, where TMDSC work is carried out with different heating and cooling rates. Cooling from the melt was at 25 0C/min for early experimental runs. Experiments testing the instrument “process capability” showed that 88

25 0C/min was the fastest cooling rate that could be maintained under proper instrument control over the temperature range of interest.

That

required the heater switch on the Liquid Nitrogen Cooling Accessory (LNCA) to be in position 3. The experiments with the DSC using ampoule material and some with melt blends in pans were carried out incorporating cooling at 2 0C/min to give more opportunity for slow crystallisation to take place whilst retaining practical times for experimental runs. 2.6.3

Experimental trials

DSC runs were carried out in hermetic pans with sample masses in the range 7 to 15 mg (and predominantly between 8 and 10 mg). This is taking into consideration TA Instruments’ recommendations of 5 to 10 mg for DSC investigations of melting/crystallisation temperatures and crystallinity [129 p. 3-12], that their recommendations for TMDSC with polymers are 10 to 15 mg [129 p. C-74] and that we have two materials in the sample being heated and cooled through their transitions.

The larger sample weights

tended to be used where there was a smaller percentage of polyamide in order to keep the polyamide transition signals at a reasonable level. Hermetic pans can ideally contain pressures up to 4 Bar.

Some loss in

weight was experienced in each heating/cooling cycle. The loss was most likely due to the hermetic pans being less than perfectly sealed. It was not due to the degradation of the polyamide or loss of any very slight residual moisture as the loss was not experienced with pure polyamide samples. Carbazole has a melting temperature of 246 0C and a boiling temperature of 355 0C. There would be a reasonable vapour pressure at 300 0C. Any loss of vapour diffusing out of an imperfectly sealed pan (and removed in the helium gas stream) would be replaced from the solid diluent in the pan in order to maintain the equilibrium vapour pressure for that temperature. This would result in a continuous loss of carbazole whenever the vapour pressure was high and the pan not perfectly sealed. Obviously a higher temperature and longer period at the high temperature will lead to a greater the loss of carbazole

from

the

“hermetic”

pan.

phenothiazine.

89

Similar

comments

apply

to

Polyamide-4,6 melts at 290-300 0C but it degrades if left at temperatures over 300 0C for extended periods. The degradation can be seen afterwards by slight yellowing. The polyamide will also experience an increase in molecular weight if left for long periods near the melting temperature [48 p. 307].

The period in the melt had to be long enough to melt all the

lamellae but these restraints on extended periods at high temperatures meant that there was little flexibility with temperatures and times. Polyamide-4,6 thus had a narrow processing window for the trials. The procedure was to ramp the sample from 25 0C to 307 0C at 5 0C/min. The sample was held at the set maximum temperature for 5 min in order to Cooling took place at 25 0C/min or

allow the sample to melt properly.

2 0C/min rate to 25 0C. This cycle was repeated within the trial run after a 7 min delay at 25 0C. Usually the actual temperature was approx. 3 0C behind the setpoint when the heating ramp finished.

The sample then

stabilised rapidly at the set maximum over the holding period. The temperature profile used for polyamide-4,6/phenothiazine was the same one despite the much lower melting temperature for phenothiazine because the polyamide-4,6 still had to be fully melted. Ramp

profiles

were

similar

for

polyamide-6,

polyamide-6,9

and

polyamide-6,12 with the difference that the maximum temperature was set at 260 0C. melted.

That ensured all carbazole in polyamide/carbazole trials was

Polyamide/phenothiazine trials were carried out using the same

profiles for consistency in approach. 2.6.3.1

Melt Blending in Pans

The blending process of heating the dried powders at 5 0C/min (modulated as above) until they had both melted and the crystallisation at 25 0C/min could be monitored in situ. This was followed by a repeat run on the blended material to monitor how the melt-formed material performed thermally after solidification had been completed in the ampoule. 2.6.3.2


No thermal monitoring of the blending process could be performed directly where it took place in an ampoule. This means that the first run on material from an ampoule is equivalent to the second run with blending in DSC pans. A repeat monitoring run was additionally performed on the material from 90

ampoules.

These were all carried out with a heating rate of 5 0C/min

(modulated as above) and cooling at 2 0C/min. 2.6.3.3

Additional trials covered in Appendix B

The results referred to in Appendix B differed from the main part of the thesis in that they were carried out early in the project with Temperature Modulated Differential Scanning Calorimetry (TMDSC) at a ramp rate of 2 0C/min with amplitude of 0.212 0C and a period of 40 s. This results in a minimum heating rate of zero during each cycle.

Further detail on the

choice of TMDSC conditions for that work is covered in the appendix, along with other information pertinent to TMDSC experiments. Baseline,

Cell

Constant,

Temperature

and

Heat

Capacity

Constant

calibrations were carried out prior to this work under the same conditions. The work in the appendix explored the effects in the case of Al2O3 of helium versus nitrogen as the purge gas and of reducing the helium flow rate. Lissajous figures are used to evaluate thermal delays in the equipment and these are described in the appendix. There is also a trial melting polyamide-4,6 carried out with TMDSC under the same ramp and cycle conditions as the Al2O3 and using helium at 50 ml/min as purge gas. Sample material and preparation for this work was identical to that for the main part of the thesis. 2.6.4 2.6.4.1

DSC Thermogram evaluation Melting and Crystallisation temperatures

Melting and crystallisation temperatures reported in the text are taken as the peak temperatures unless otherwise noted. 2.6.4.2

Enthalpies from Graphical Analysis

Melting and crystallisation enthalpies were measured using the “Integrate Peak” capabilities of TA Instruments' Universal Analysis program. The Sigmoidal Tangential option was consistently used in Enthalpy of Fusion (∆Hf) determinations because sometimes the two sides of a peak were not exactly parallel.

Occasionally, two peaks partly overlapped.

The double

peak was integrated using the “Sigmoidal Tangential” option and then a “Perpendicular Drop” was placed to separate the two areas. The placement of the Drop was positioned so that the estimated tail from one curve was equal in area to the estimated tail from the other. One peak was sometimes 91

a shoulder on another other peak.

It was more difficult to accurately

estimate the appropriate areas attributable to each. Usually it was possible to make estimates based on different premises. These were then averaged to give the best estimate.

There was occasionally a ‘spiky” crystallisation of

carbazole or phenothiazine part way through another crystallisation with a Gaussian

form.

The

leading

edge

of

the

carbazole/phenothiazine

crystallisation was non-Gaussian. The most appropriate way of tackling that was to take the “Sigmoidal Tangential” integral over the double peak and to subtract the “Linear Integral” over the carbazole or phenothiazine peak from it to give the integral over the Gaussian peak. 2.6.4.3

Crystallinity calculations

Total crystallinity determination was made by comparing the measured total enthalpy from melting with that expected on a weight proportional basis from literature values for 100% crystallinity of the polyamide and the diluent. The weight percentages for samples blended in pans were taken from the weights of the constituent powders. For the ampoule samples, these were taken from the high resolution TGA “R4S4” results from samples taken next to the DSC samples.

2.7 Simultaneous Differential Thermal Analysis-Thermogravimetric Analysis The equipment used for Simultaneous Differential Thermal Analysis – Thermogravimetric Analysis (SDT) was a TA Instruments Model 2960. The furnace with an SDT instrument is horizontal as distinct from the vertical furnace of a TGA instrument. This means that the gas flow around a sample will be different from the TGA and different again to the DSC. DSC has the sample in a closed pan. SDT does provide a means of quickly comparing various materials. Trials were run with nitrogen at 50 ml/min. Heating ramps were carried out at 10 0C/min.

92

2.8 Fourier Transform Infra-Red Spectroscopy 2.8.1

Mid Infra-Red Range Measurements

Diffuse Reflectance Fourier Transform (DRIFT) Infra-Red spectroscopy was initially carried out on a Nicolet Magna Spectrometer Model 750 with DRIFT attachment.

A few samples from Ampoule 1 (PA46Car), carbazole powder

and polyamide-4,6 powder were tried. This led to some inconclusive results with possible frequency shifts when comparing the resulting spectra of the blends to those of the starting materials. It was suggested that the narrower peaks associated with Photoacoustic spectroscopy (PAS) could alleviate the problems. Consequently, the results presented in the text for Mid Range IR were carried out on the Nicolet 750 but with an MTEC model 300 PAS attachment. This was operated in the frequency range 4000 to 400 cm-1 at a resolution of 8 cm-1 with an IR source, KBr beamsplitter, a mirror velocity of 0.1581 cm/s and an aperture setting of 130. Ultra High Purity (99.999%) helium gas from BOC was used at a flow rate of 20 ml/min. Background calibration prior to the measurements on blends was with 512 scans of a carbon black standard from MTEC.

Trials showed that it was

necessary to re-calibrate every 4 to 5 hours. The entry of the carbon black sample holder in the PAS attachment and flushing took place at a much slower rate than the manufacturers’ recommendations on the advice of a colleague [149] who had had extensive PAS experience. The first stage of flushing was held for 10 minutes, and the later ones for two minutes before complete closure took place and measurements began. There was usually little evidence of H2O peaks in the background spectra.

Background

calibration was repeated, if necessary, until satisfactory elimination of moisture peaks was achieved. Samples of approximately 7x7x2 mm were cut in the glovebox from the material extracted from the ampoule and then placed (with dry atmosphere) in small glass vials with airtight stoppers. The samples could then be kept completely dry until they were put in the photoacoustic holder. This was in order to minimise the potential effects of moisture ingress. Samples were fitted to the PAS holder and placed immediately in the dry helium environment of the photoacoustic detector. The pieces of material from the various ampoules were usually placed bottom side (from the ampoule) 93

upwards in the sample holder as this surface was flat. They had to be fitted in several pieces to the holder in some cases where the material had been too crumbly to use a single piece. The samples were set at approximately 1 to 1.5 mm under the lip of the holder using shims under the sample. This height gave the best results and was in line with the manufacturer’s recommendations [150 p. 6]. Signal levels were checked prior to each trial run to ensure that they were not so low that noisy results ensued and that there was no risk of saturation of the detector amplifier. The first stage was held for 5 minutes and the later ones for a minute before complete closure took place. Samples were run for at least 256 scans. 2.8.2

Near Infra-Red Range

It is specifically in the Near Infra-Red (NIR) range 6900 to 6100 cm-1 and 5000 to 4500 cm-1 that hydrogen bonding effects are to be expected [141]. Measurements were carried out on the same Nicolet 750 instrument as for the mid-range experiments but with a white light source, a CaF2 beamsplitter and a PbSe detector. That configuration for NIR experiments covered the range 11,000 to 2100 cm-1. The changeover was easy to make but required 24 hours stabilisation of the system before measurements could be made because the whole instrument had to be opened to the atmosphere during the changeover process. The best results were obtained with the sample in a Diffuse Reflectance Infra-red Fourier Transform (DRIFT) attachment rather than the PAS attachment. A sample size of approximately 7x7x2 mm was used, as with the mid-range experiments. The sample height needed to be adjusted for each sample to give the maximum signal, or the signal/noise ratio was poor. A stabilising period prior to measurement of approximately 20 min was necessary to eliminate moisture from the measurement chamber after having opened it for sample insertion and height adjustment. Background and sample runs were at least 256 scans at a resolution of 4 cm-1 and a mirror velocity of 0.9494 cm/s.

Background

calibration was with dry KBr. The spectra for the ampoule material from blends could be directly compared with those from ampoules of the constituent materials without the modelling used in the Mid Range because of the broader peaks involved.

94

2.9 Small Angle X-Ray Scattering Mr Clint Gamlin of the Ian Wark Research Institute at the University of South Australia spent some time in late 2000 at the University of Connecticut, USA carrying out Small Angle X-ray Scattering (SAXS) for his own project. He was kind enough to run some of the samples available from ampoules whilst there. The samples were sent sealed against moisture to Connecticut. The equipment had been custom built in the University of Connecticut. Conditions used were 627 ±5 mm sample to detector distance.

A 40 kV

100 mA Cu source was used. The main wavelength is 0.154 nm. Scattered X-Rays co-operatively interfering from reflection planes in the sample were collected using a 2-D area detector. Material was pulverised in a mortar and pestle and then contained between Kapton films.

Vacuum was achieved with a rotary oil pump for 5 min

(approximately 0.150 Pa) before measurement runs having a duration of 1 hr. A blank run was made with the Kapton film but no sample. There was little scattering from the Kapton film. No peaks were evident at the angles used.

Results were obtained after determining the centre of the circle by

changing the radius and centre of an overlaid circle. The spectrum was then determined by integrating around the signal circle.

Custom software was

used for this. These results were saved as a tab delimited text file. Making a running average over three points smoothed the raw results and did not affect accuracy as peaks were broad. The background scatter results were subtracted from the measured values to give corrected smoothed curves.

2.10 Cross Polarised/Magic Angle Spinning Solid State 13C Nuclear Magnetic Resonance Spectroscopy All spectra were run on a Bruker MSL300 operating at 300.13 MHz for 1H and 75.482 MHz for

13C.

Experiments were performed with a standard

Bruker 4 mm Magic Angle Spinning (MAS) probe. The MAS spinning speed was approx. 5 kHz. The 900 pulse times for both 1H and

13C

were 4.1 µs.

The spectra were recorded using the standard cross polarisation sequence with a 3 s recycle delay and high power proton decoupling during acquisition. The cross-polarisation contact time was 1 ms. The spectrum width was 38.5 kHz, and 2000 data points were collected. 95

Single pulse

excitation spectra were collected under the same conditions and with a recycle delay of 3 s. On processing, a line broadening of 50 Hz was used, and the files were zero-filled to 4k data points.

Chemical shifts are

referenced to TMS via the external standard adamatane.

96

Chapter 3

POLYAMIDE-4,6 WITH CARBAZOLE CONTENTS 3.1

Introduction

98

3.2

Materials, Handling, Sample Preparation and Techniques Used

98

3.3


100

3.3.1

Reproducibility

100

3.3.2

Ramp methods

100

3.3.3

Ampoule 1

102

3.3.4

Later Ampoules

103

3.4


3.4.1

Thermograms for blends made in DSC pans

3.4.1.1 3.4.1.2 3.4.1.3 3.4.1.4 3.4.1.5

3.4.2

3.4.4.2 3.4.4.3 3.4.4.4 3.4.4.5

3.5

0

First DSC heating ramp of Ampoule 1 materials at 5 C/min. 0 Crystallisation for first DSC cooling ramp of molten blends at 2 C/min 0 Second DSC Melting at 5 C/min of samples from Ampoule 1 Thermograms for Ampoule 1 samples on second DSC cooling ramp at 0 2 C/min.

Later Ampoules Under Controlled Cooling Conditions

3.4.4.1

0

Melting Temperatures from first DSC melt at 5 C/min of samples from ampoules Overall Crystallinity 0 Crystallisation Temperatures from first crystallisation in DSC at 2 C/min PA46Car material from a variety of ampoules Crystallinity from first cooling Phase Diagrams for first time heating, and first time cooling ampoule material in DSC with later ampoules


3.5.1

Mid Infra-Red – Photoacoustic

3.5.1.1

3.5.2

104 104 107 110 113 114

115

Accuracy of Melting & Crystallisation Enthalpy Measurements

Ampoule 1

3.4.3.1 3.4.3.2 3.4.3.3 3.4.3.4

3.4.4

0

Melting for first heating ramp of the dry powders at 5 C/min in pans. 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min 0 Second Melting at 5 C/min of materials blended in pans 0 Crystallisation for second cooling ramp at 25 C/min of pan blended polyamide-4,6/carbazole mixtures. Evaporation of Carbazole

Melt blends from Ampoules

3.4.2.1

3.4.3

104

N-H Stretches

115

116 116 118 119 120

121 121 123 125 126 128

129 129 129

Near Infra-Red – Diffuse Reflectance Infra-red Fourier Transform

133

3.6


135

3.7

Nuclear Magnetic Resonance Spectroscopy

136

97

3.8

Summary

139

3.1 Introduction The observation that prompted this work was the discovery that a mixture of polyamide-4,6 with carbazole in an ampoule had unexpectedly solidified from the melt into three separate sections as material cooled at differing rates.

Those three sections as seen in Figure 3-1 had markedly different

colours (red, white and fawn), hardness, concentrations of polyamide and melting/crystallisation characteristics.

Figure 3-1 Red, white and fawn parts of polyamide-4,6/carbazole material from the first ampoule. The material had already had samples of the red, white and fawn material removed before this photo was taken.

This chapter covers the experimental work in melt blending polyamide-4,6 with carbazole and the characterisation of those materials using a variety of techniques. What will be shown from that experimental work is that there is usually phase separation of the two materials into crystalline phases embedded in the amorphous mix. Under certain conditions, the two materials can crystallise together giving characteristics different from either of the two raw materials. The characterisation places a strong emphasis on the DSC results. It will determine at what temperatures and to what extent crystallinity is occurring and whether phase separation takes place.

It will also be demonstrated,

based on FTIR in the Mid Range and with Near Infra-Red, that the formation of crystalline areas comprising both materials does not involve hydrogen bond interactions between the two materials.

3.2 Materials, Handling, Sample Preparation and Techniques Used Polyamide-4,6 and carbazole materials, their handling, the preparation of blended samples in the DSC and the preparation of samples in ampoules has been described in Chapter 2. 98

Blending of small amounts in pans took place for 5 mg of polyamide-4,6 with 1 mg, 3 mg, 5 mg and 8 mg of carbazole. These samples are referred to in the text as 83PA46Car, 63PA46Car, 50PA46Car and 38PA46Car, depending on the percentage polyamide in the blend, in line with the general notation referred to in Chapter 1.

DSC monitoring took place during the blending

process. Larger amounts, at least 1.5 g at a time, for investigation with some techniques were melt blended in ampoules using the methods described in Chapter 2. The samples made in DSC pans were crystallised at 25 0C/min in order to minimise the time at elevated temperatures where the carbazole was more likely to evaporate. The ampoules (apart from very early ones) had experienced controlled cooling at 2 0C/min.

Subsequent evaluation of

samples made in ampoules was undertaken at a cooling rate of 2 0C/min to match the process used for ampoules. This difference is useful in that it allows us to look for differences in samples caused by the fast or slow cooling rates. The first material from an ampoule was made with equal amounts of polyamide-4,6 and carbazole, but with heating and cooling only poorly controlled by manually changing the temperature setting on the furnace. There was a very brief overshoot to 320 0C and uncontrolled cooling in several steps. The ampoule had a circular cross-section and had not been entirely horizontal. Different parts of the solidifying material had cooled at different rates. The “Red” part was investigated with Solid-State NMR. The “Red”, “White” and “Fawn” parts were investigated with TGA and DSC. Later ampoules had flat bottoms and were set accurately horizontal so that cooling of the blend mass would be more even.

Relatively consistent

material could be made within each ampoule. A seven-step programmable oven controller was obtained after Ampoule 7 and compressed air cooling was installed to give close control over oven temperature during the critical crystallisation periods. Ampoule 15 (with 50% of each material) was used in SAXS. That material was made at a time the thermal control of the blending furnace was being 99

refined.

The cooling profile was slightly different from that used for later

ampoules described here. Melt blending of all other ampoules was carried out as in Chapter 2 to produce a series of ampoules. The blend weight percentages were chosen to give 1:1, 2:1 and 3:1 ratios between the polyamide and the carbazole with respect to potential hydrogen bond sites, both materials potentially being able to form hydrogen bond interactions. The above samples prepared in ampoules were investigated with TGA, DSC, SAXS and NMR as described in Chapter 2. Each different ampoule sample was not necessarily investigated with all techniques. In some cases, more than one sample was taken from an ampoule where there were regions with noticeably different colours.

3.3 Thermogravimetric Analysis 3.3.1

Reproducibility

It was found that variations in measured “composition” of solidified polyamide-4,6/ carbazole material from ampoules occurred when taking different samples from similarly coloured sections and using the same TGA method to determine the percentage of polyamide in the sample. The results on the few pairs of duplicate samples using the same ramp method were not more than 5% different from one another.

This should be seen in the

context of the accuracy of TGA instrument accuracy.

Samples of

approximately 10 mg should be accurate to approximately 1% so the extra variation is most likely due to local variations in composition. 3.3.2

Ramp methods

The effects of High Resolution, High Sensitivity “R4S4” versus the constant ramp rate “straight” ramp method evaluated for plateau levels at 300 0C in Figure 3-2 shows that there is a reasonably good correlation between the two ramp methods. Variations differ at most by 5% from a 1:1 correspondence. There

is

no

significant

preponderance

on

either

side

of

the

1:1

correspondence line. The results are in line with the cases mentioned in the previous section where multiple samples were taken from next to one another and 100

investigated with TGA using the same ramp method. This shows that the difference in polyamide percentage between using the two different methods from two different samples is no worse than overall accuracy of the TGA

Polyamide concentration by "straight" method (%wt)

method for taking two different samples. 100 PA46Car 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration by "R4S4" method (%wt) Figure 3-2Comparison of TGA results of polyamide levels determined by R4S4 and "straight" heating methods at 10 0C/min for polyamide-4,6/carbazole blends demonstrating similar results.

The TGA curves from the different ramp methods did not necessarily overlay each other and the temperature ranges where the weight percent dropped from near 100% to near the plateau level varied noticeably.

This is most

likely due to differences in the physical forms of the samples.

It was

sometimes not possible to have the same sized lumps or small pieces for both samples. This factor meant there would be a variation in the ease with which the carbazole could diffuse to the surface and evaporate. The actual percentages of polyamide in samples, as determined by TGA, often varied markedly from the percentages of the constituents weighed into the ampoules. The accuracy of weighing and introduction into the ampoules would not be a significant source of error, given that amounts weighed into the ampoules always exceeded 400 mg.

The variation in colour seen in

different parts of some of the ampoules shows some non-uniform solidification from the melt.

101

The TGA results were used later with the DSC results to provide accurate crystallinity calculations because the percentage polyamide is needed in those calculations, and that differed from the amounts weighed into the ampoules. This was the major use of TGA results. 3.3.3

Ampoule 1

The very first ampoule resulted in almost perfect separation of the solidified material into a brick-red, crumbly crystalline portion, a whitish hard material and a tough fawn material that cut with side-cutters in a manner similar to polyamides. It can also be discerned in Figure 3-1 that the height of the red section is noticeably lower than the remainder. That run, with manual adjustment of set temperatures, had originally just been intended as a “dry run” to determine where there were likely to be problems to overcome in making blends in ampoules.

Figure 3-3 TGA thermograms of the Red, White and Fawn sections of Ampoule 1 material with “R4S4 heating method at 10 0C/min Plateau levels at 300 0C give the percentages of polyamide in the samples.

Various later ampoules unsuccessfully attempted to emulate the original separation into the three radically different materials in the same ampoule. Another alternative was to produce each type independently under controlled conditions.

That was the approach taken later but there were sometimes 102

areas of differing colour within each ampoule. These differing colours were usually associated with differing compositions. Figure 3-3 shows the TGA results from samples taken from the red, white and fawn sections of the first ampoule material with percentages of polyamide being 35, 65 and 72% respectively. It was noteworthy that the first two percentages corresponded closely with blends where there was a 1:1 and 3:1 ratio between potential hydrogen bonding sites on the two constituent materials. That prompted the choice of weight percentages in later ampoules that were 1:1, 2:1 and 3:1 in the ratios between potential hydrogen bond sites.

Actual polyamide concentration (%wt)

3.3.4

Later Ampoules 100 PA46Car 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 3-4 Actual R4S4 TGA percentage of polyamide from plateau levels with R4S4 heating method at 10 0C/min against the expected percentage polyamide in polyamide-4,6/Carbazole ampoule samples based on the weights used in the ampoule. This displays the variability of concentration possible at a local level within the ampoules.

Figure 3-4 for samples taken from a number of ampoules, shows how the measured percentage compositions differed from the expected ones based upon the weight of each constituent put in an ampoule. It can be seen from the figure that the local average concentration of polyamide is often very different from the theoretical.

The encountered spread was due to some

self-association of the molten polyamide-4,6/carbazole and some uneven cooling of the material in ampoules.

103

The powders had been introduced

separately through the narrow neck of the ampoule before sealing and had been partly mixed by some shaking before heating in the furnace. The molten fluid had not been actively mixed at high temperature in the furnace due to the impracticality with the available equipment.

Diffusion of the

molten materials at high temperature had to be relied upon. Widely different polyamide percentages with samples from the same ampoule were usually associated with noticeably different colouring or hardness/flakiness of the samples.

3.4 Differential Scanning Calorimetry DSC experiments can tell much about what materials have melted from a crystalline phase upon heating and have crystallised into a phase on cooling. The peak temperatures are indicative of the material or material phase involved and the integrals of the peaks give information about the level of crystallinity because amorphous polymer material will not have clear melting peaks.

The DSC results here will explore the available information to

provide a reasonable understanding of what is happening as blends of polyamide-4,6 and carbazole are heated together, melted, crystallised into solids and remelted. Polyamide-4,6 melts at 293 0C. The polyamide crystallises near 274 0C for a cooling rate of 2 0C/min and near 259 0C for a cooling rate of 25 0C/min. Carbazole melts at 246 0C and crystallises at 237 0C for a cooling rate of 2 0C/min and near 233 0C for a cooling rate of 25 0C/min.

3.4.1 3.4.1.1

Thermograms for blends made in DSC pans Melting for first heating ramp of the dry powders at 5 0C/min in pans.

Figure 3-5 displays the DSC thermograms of polyamide-4,6, carbazole and various mixes of powders taken in DSC pans up to the melt.

The

thermograms have been placed on the same graph for best comparison of the different melting temperatures. The curves have been separated vertically for clarity. Colour is used to identify the individual thermograms and a consistent approach is taken with fixed colours for each concentration decile as displayed in Figure 3-16.

104

Figure 3-5 Polyamide-4,6, carbazole and various powder mixes in DSC pans heated for the first time in the DSC at 5 0C/min.

a) The 83PA46Car sample has a broad peak below the carbazole melting temperature and a second peak with a long lead-in below the polyamide melting temperature.

The first peak is consistent with amorphous

polyamide from the grains of polyamide in contact with carbazole powder dissolving molecules from the carbazole crystals to produce a eutectic solution of polyamide-4,6/carbazole.

The endotherm for the normal

melting of the small amount of carbazole in the sample has clearly been moved lower by more than 10 0C due to the presence of polyamide. Lamellae in the polyamide part of the sample will normally be impermeable to foreign molecules in the initial stages. The amorphous polyamide is well over the Tg (approximately 70 0C) at that stage so the polyamide chain segments are vibrating more strongly.

Carbazole

molecules in contact with amorphous polyamide are then in a position to be dislodged from carbazole crystals and to diffuse rapidly into the amorphous polyamide, reducing its viscosity by plasticisation and allowing stronger segmental motion.

It is probable that at higher

temperatures the polyamide/carbazole solution will be able to slowly 105

dissolve the more stable polyamide lamellae.

The amount dissolved by

240 0C is just enough to form a saturated solution of the two materials at that temperature, leaving an excess of polyamide-4,6 not in solution. The solubility of the polyamide in the solution increases as the temperature is increased in the DSC and some more can dissolve to form a saturated solution at that higher temperature. Eventually the temperature is high enough that all the excess polyamide can dissolve. In this case that process reaches a peak before the temperature has reached the normal starting temperature for the main melting of polyamide-4,6. This process of dissolving more and more of the higher melting material in the saturated solution occurs with virtually all the combinations of materials studied in this thesis. It has been referred to in Chapter 1 and, for want of a known term for physical chemists, has been given the name here of Temperature Limited Solubility (TLS). The DSC endotherm resulting from this type of process will be referred to throughout the text as a TLS peak even though it is often a flat, long shoulder at the end of a main peak. It can also be a large peak when there is a large excess of material to dissolve into the saturated solution. b) The 63PA46Car thermogram has a similar first peak to the first one in 83PA46Car but is larger because there is more carbazole in the sample to dissolve together with some of the polyamide-4,6. The very rounded first peak is quite common with first melting of powders, as will be seen in other chapters, because the samples are powders and not intimate molecular mixes of the two materials. The small amount of polyamide remaining, once a saturated solution at 245 0C is reached, requires only a small temperature increase to complete the dissolution of residual polyamide by an increase in temperature. The process is complete with the TLS “peak” finished by 265 0C. c) The 52PA46Car thermogram is similar to the 63PA46Car except that the first peak is larger and the TLS peak is smaller because of the smaller percentage of polyamide in the mixture. d) The 38PA46Car curve is similar again except that the amount of carbazole is larger.

There is a difference here in that some of the

polyamide-4,6 does not dissolve into the solution until the normal 106

polyamide-4,6 melting temperature.

There is a broadening of the

depressed carbazole melting at the normal melting temperature of carbazole. It cannot be said at this stage whether the observations here are due to kinetic effects with the slow dissolution of materials into the saturated solution or to the composition being in a region of phase space where there is phase separation, perhaps having a Lower Critical Solution Temperature (LCST). It can be seen from these curves that there are broad melting peaks where the carbazole appears to have dissolved into the polyamide at temperatures depressed below the normal carbazole melting range.

The broad peaks

appear to encompass multiple thermodynamic processes because they are multi-peaked endotherms. The middle concentrations can be seen to have no further peaks near the normal polyamide melting temperature.

This

infers that, in those cases, virtually all the polyamide has been dissolved into solution by close to the normal carbazole melting temperature. There is evidence at high temperature and low polyamide-4,6 concentration that there is an impediment to all of the polyamide being able to dissolve into the saturated solution before the normal melting temperature of the polyamide. This is indicating a forbidden region of phase space where phase separation most likely would take place.

3.4.1.2

Crystallisation for first cooling ramp of the molten blend at 25 0C/min

The crystallisation processes are depicted in the thermograms of Figure 3-6. The distinctive crystallisation thermogram of carbazole, as seen in Chapter 1, allows definitive assignment of its crystallisation, unlike the situation with melting where a peak may, in general, be due to either material or the dissolution of one material into the other. The crystallisation peaks are “spiky” due to their rapid occurrence and can lean backwards during cooling at angles greater than 900. The crystallisation can take place in a process where there is some polyamide crystallising at the same time as the carbazole and that will lead to a slightly less spiky form.

We will see

examples in other chapters where crystallisation of the carbazole begins to take place and is overridden in the same overall crystallisation peak by later crystallising of carbazole together with polyamide. Phenothiazine also has this distinctive “spiky” crystallisation peak. 107

Figure 3-6 Polyamide-4,6, carbazole and some PA46Car high temperature blends from DSC pan blended samples solidifying from the melt for the first time in the DSC at a cooling rate of 25 0C/min.

Peaks within 10-15 0C of the polyamide-4,6 crystallisation temperature can be assigned to the crystallisation depression of portions that form lamellae of polyamide-4,6 without incorporating any carbazole within the lamellae although there may be a small amount of carbazole in the interlamellar space [151 pp. 99, 208]. There are some peaks at lower temperatures having a decidedly skewed Gaussian

form.

They

are

very

much

lower

than

those

for

crystallisation-depressed polyamide and are of some form of combination between the two basic materials or are from a concurrent crystallisation of the two materials that begins with a high concentration of carbazole in the crystallisation. a) We have the depressed crystallisation of polyamide occurring within an amorphous phase at high polyamide concentrations in the 83PA46Car thermogram.

Essentially,

this

crystallisation.

108

is

a

“contaminated”

polyamide

b) At middle levels of polyamide, with 63PA46Car, there is only the one Gaussian peak approximately 50 0C lower than the normal crystallisation temperature of polyamide-4,6 and nearly 20 0C below that of carbazole. This is a combination of carbazole crystallising with polyamide-4,6. c) The sample 52PA46Car with almost equal weight concentrations of each constituent has a similar, if slightly more skewed, peak 2 – 3 0C higher. At this stage the depression of both crystallisation temperatures is slightly less because the interactions between the two materials is not quite so favourable at that concentration. d) Finally, the curve for 38PA46Car is showing a very slight crystallising of polyamide

just

under

the

normal

polyamide-4,6

crystallisation

temperature, probably indicating some phase separation. A double peak follows just over 10 0C below the carbazole crystallisation temperature. The first peak of the two shows crystallisation-depressed carbazole with its distinctive shape followed just after the peak by a Gaussian curve similar to the concurrent crystallisation curves but at a slightly higher temperature. In this last thermogram we find three phases crystallising within the remaining amorphous material, a small polyamide part, a carbazole part and a part where the concentrations of polyamide-4,6 and carbazole are such that they can crystallise at the same time. In the set of curves in Figure 3-6 we can see either 1 of 3 or all 3 forms crystallising in each cooling ramp at 25 0C/min. Overall, we see interactions between polyamide-4,6 and carbazole being sufficient that the crystallisation temperature for the polyamide is depressed by a minimum of 50 0C and carbazole by approximately 20 0C for a blend near 63% polyamide.

The interaction between the two materials is less

favourable as the concentration of polyamide is made lower than 52%. At 38% polyamide-4,6, there is a small amount of the polyamide crystallising in nearly pure form very near the polyamide crystallisation temperature. This is most likely due to phase separation. At that concentration, the carbazole is starting to crystallise independently with less depression of the crystallisation temperature.

It is followed by the remaining crystallisable

polyamide-4,6 and carbazole crystallising together at slightly higher temperatures than the samples with 52 and 63% polyamide. 109

3.4.1.3

Second Melting at 5 0C/min of materials blended in pans

Samples made in DSC pans underwent further DSC runs to ascertain the thermal characteristics of the solidified material from the first crystallisation. Examples are given in Figure 3-7 of the re-melting thermograms compared to those of the raw materials.

Figure 3-7 DSC thermograms of polyamide-4,6, carbazole and several powder mixes from the first heating in pans being remelted in the DSC at 5 0C/min after prior crystallisation at 25 0C/min.

a) The thermogram for 83PA46Car demonstrates again the single peak of the carbazole-“contaminated” polyamide in the lowest concentration of carbazole. The melting temperature on heating the solidified material is now slightly higher, most likely due to the concentration of carbazole being lower because of evaporation.

There is no evidence in this

thermogram of melting at temperatures under 270 0C. b) The remelting curves for the 63PA46Car sample displays four peaks. The first two peaks are at approximately 230 and 236 0C.

Double peaks

where there was a single Gaussian peak in the crystallisation are most likely due to the polyamide-4,6 part of the blend originally crystallising in a metastable form and then melting/recrystallising here into a stable 110

form just prior to melting on re-heating. The much lower temperature of the prior crystallisation at fast cooling rates has reduced the mobility of the polymer chains and so made the formation of lamellae far less than ideal.

Ramesh [152] shows the effects of isothermally crystallised

polyamide-4,6 changing from a high temperature α-form to a room temperature α-form at 120 0C.

The fast dynamic crystallisation at

25 0C/min is likely to have constrained the polyamide to remain in the high temperature form that had melted/recrystallised into the more stable form just prior to remelting. This is similar, in principle, to the melting of the unstable form of polyamide-6 into the stable form.

The

similar form of the melting curve for polyamide-6 blends, when remelted after a fast crystallisation, can also be seen in Chapter 4 in Figure 4-6. We will see in trials with polyamide-4,6/carbazole blends from ampoules that the initial (re-)melting curves are different because the cooling rate during the solidification process in the ampoule was at a much slower 2 0C/min. The small third peak is at 250-253 0C. That is a TLS peak for the dissolution of excess polyamide into the high temperature solution. The small fourth peak in this thermogram is close to the polyamide-4,6 melting temperature and shows a residual amount of polyamide that has not been able to melt earlier into the high temperature solution due to limited solubility in that part of the phase diagram. c) The thermogram for 52PA46Car is very similar to that for 63PA46Car except that the double peak is larger and the TLS peak smaller because of there being less excess polyamide-4,6 per milligram in the sample.

In

this case there was no small peak near the polyamide-4,6 melting temperature. There was, however, a small peak near 265 0C for which there is no available explanation. d) The sample with the highest carbazole concentration, 38PA46Car, has an extremely small deflection at the end of the double peak. This is a vestige of the TLS peaks seen in the previous two descriptions but it is not clear which material it is. The last peak, just under the melting temperature of pure polyamide-4,6, is similar to that with the 63PA46Car sample. All of the remelting curves have lost the broadness seen in the first melting of the powders. In all except the lowest polyamide concentration curve we 111

find that the remelting is taking place with the major peak approximately 3 0C lower than previously.

This is considered to be due to the more

intimate contact between the materials thermally rather than having two separate powders with only small points of contact until the carbazole has melted. The materials are already mixed rather than existing as separate powders. We also see double melting peaks for the main melting peak due to melting and re-crystallisation of metastable high temperature form into the more stable room temperature form before the main polyamide melting. The mid and lower polyamide concentration samples showed very small amounts of polyamide melting just under the normal polyamide melting temperature. The 38PA46Car sample was, again, demonstrating the limited solubility of polyamide-4,6 in carbazole at high concentrations of carbazole in a consistent manner to the previous two melting and crystallisation thermograms for the sample. A comparison between Figure 3-5 and Figure 3-7 shows that main melting peaks are generally similar between the powder melting and the remelting of previously melted material, although there are some small peaks evident on the second run that were not there on the first run.

There is a similar

pattern of melt depression of the polyamide melting and of carbazole melting in forming what appear to be eutectic melts. Other peaks are not associated with either of the main melting temperature ranges. The initial melting with polyamide and carbazole powders was only partially conducive to forming eutectic mixes during the initial melting. The materials were able to demonstrate melting of the various components in a cleaner manner than on the first run once they had solidified into a single mass of crystallised carbazole regions, crystallised polyamide regions, polyamide lamellae

with

inter-lamellar

carbazole

and

amorphous

polyamide-4,6/carbazole. The information from this thermogram is also portrayed as a scattergram in Figure 3-8 for later comparison with the ampoule scattergrams. Very small peaks for polyamide-4,6 are presented as smaller sized data points on the scattergram.

112

310 Metastable Eutectic TLS carbazole polyamide-4,6

300

Melting peaks ( 0C)

290 280 270 260 250 240 230 220 210 200 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 3-8 Second melt in pan at 5 0C/min for solidified polyamide-4,6 & carbazole: Melting peak temperatures vs. Percentage polyamide showing temperatures of peaks for metastable forms of polyamide-4,6, the melting of eutectic mixes, the temperature limited solubility(TLS) peak and near-pure polyamide-4,6. The smaller symbol size for some polyamide-4,6 data points signify very faint peaks.

3.4.1.4

Crystallisation for second cooling ramp at 25 0C/min of pan blended polyamide-4,6/carbazole mixtures.

Figure 3-9 Second crystallisation of pan blended samples of polyamide-4,6, carbazole and several powder mixes in the DSC at 25 0C/min.

113

The peaks in Figure 3-9 for the second crystallisation are almost identical to those of the first crystallisation except that they are 1 – 2 0C lower. A high level of similarity would be expected between the first and second crystallisations as the main difference is the slightly different concentrations due to some carbazole evaporation.

Crystallisation temperatures 1 – 2 0C

lower are perhaps due to a greater perfection of the second melt. There have possibly been some odd nuclei left after the first melt even though the samples had been held at temperatures over 300 0C for more than 6.5 min. 290 polyamide-46 carbazole Other Peak

Crystallisation peaks (0 C)

280 270 260 250 240 230 220 210 200 190 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 3-10 Second crystallisation in DSC at 25 0C/min of pan material from polyamide-4,6, carbazole and several powder mixes.: Crystallisation peak temperatures vs. weight percentage polyamide with small peaks having smaller symbols.

Figure 3-10 above is presented to show the scattergram of crystallisation temperatures versus composition for later comparison with ampoule results. The smaller green data points are for very small peaks.

3.4.1.5

Evaporation of Carbazole

The weight of samples reduced every time they were taken through a melt-crystallisation cycle in the DSC. This was despite the use of “hermetic” pans. Hermetic pans are rated to 4 Bar in an ideal situation. Lids had been crimped with the tool supplied by TA Instruments in the recommended way, although there was no practical manner of monitoring prior to the DSC run as to how perfect this was. There would be a substantial vapour pressure at the working temperatures (300-310 0C) despite the melt temperature of 114

carbazole being 246 0C and its boiling temperature being approximately 350 0C.

Any less than perfect seal meant that carbazole vapour could

diffuse out of the pan and be carried away in the gas stream. This is similar to the sublimation of carbazole samples in the open stirrup of the TGA already being 50% complete by 230 0C, as seen in Chapter 4. It can also be seen in Figure 3-3 that the evaporation of carbazole from blends can be substantially complete by 240 0C. The loss of carbazole by evaporation in a first DSC run meant that the concentration of polyamide in the subsequent run became closer to a pure polyamide sample than it had been. This was evident in some cases with high polyamide concentration where the melting temperature was not depressed as far in the second run.

A similar effect will be seen with

phenothiazine in later chapters.

3.4.2

Melt blends from Ampoules

3.4.2.1

Accuracy of Melting & Crystallisation Enthalpy Measurements

The total enthalpy of crystallisation from the melt to room temperature should be close to the enthalpy of melting in taking the sample from room temperature back into the melt. 140 PA46Car 1:1 line

∆Hf Second melt(J/g)

120 100 80 60 40 20 0 0

20

40

60

80

100

120

∆Hf First crystallisation (J/g) Figure 3-11 Relationship between the Enthalpy of Fusion ∆Hf for first crystallisation and second melt with polyamide-4,6/carbazole blends from ampoules demonstrating peak integration measurement accuracy. Total heat of fusion for second melt is plotted against Total Heat of the first crystallisation.

115

140

For full accuracy, the exotherms would have to be integrated in going from the melt to room temperature and combined with those from room temperature to just under melting. This would have to be compared with the integral of the endotherm(s) in taking the heated solid further through into the liquid state again. The peak areas for crystallisation during cooling to 100 0C were, in fact, plotted against the melt peaks from 100 0C to the melt because of the difficulty in accurately integrating small changes over long time spans. This comparison can be seen in Figure 3-11 along with the line for a one-to-one correspondence. It can be seen that there is generally a good correlation.

This forms a good verification that the integrals are

accurately measured because the shapes of the crystallisation curves are quite different from the melting curves.

3.4.3

Ampoule 1

Ampoule 1 was originally crystallised under uncontrolled conditions but it is still instructive to look at the thermal performance in the DSC of samples taken from the red, white and fawn sections as these gave the initial insights into the composite materials that resulted.

3.4.3.1

First DSC heating ramp of Ampoule 1 materials at 5 0C/min.

Figure 3-12 shows the (re)melting of the three distinct materials found in the first ampoule when samples of the ampoule material are ramped at 5 0C/min. It is shown along with the thermograms of ampoule samples from the raw materials. a) The thermogram for the “Fawn” material, 72PA46Car, is similar in form and close in temperatures to that for the 84PA46Car pan blended sample during its second DSC ramp. b) The very hard, white material, 65PA46Car, has a peak at the same temperature as the main peak of the remelted 63PA46Car of the pan blended series. In that trial, we had seen on the second melt, that there was a double peak approximately 10 0C below the carbazole melting temperature. In this case we do not see the double peak. This is most likely because the ampoule crystallisation took place at a slower rate creating a more stable lamella.

A change in lamellar form to a more

stable state was now not necessary. The peak between 240 and 270 0C is a TLS peak, also similar to the 63PA46Car in Figure 3-7 116

for the pan

blended series but extending a little longer before all the excess polyamide-4,6 is dissolved into the saturated solution.

Figure 3-12 First DSC melting at 5 0C/min of Red, White and Fawn sections melt blended in Ampoule 1 compared with melt curves of polyamide-4,6 and carbazole samples previously melted/crystallised in ampoules.

c) The flaky and crumbly brick-red crystalline part of Ampoule 1, 35PA46Car, has a single melting peak at 236 0C, approximately 10 0C below the melting temperature of carbazole. The level of polyamide, at 35%, is similar to that in the 38PA46Car pan blended material.

The

results should be comparable to the second DSC melt of that material because the ampoule sample has already gone through powder melting and crystallisation in the ampoule. There is no peak here near 230 0C prior to the main peak from melting the metastable form of polyamide-4,6 because the sample has been cooled relatively slowly, allowing the polyamide-4,6 in the blend to crystallise in a more stable form. The lack of a small peak near that of polyamide is because there is no excess of relatively pure polyamide lamellae with inter-lamellar polyamide still to melt. That is most likely due to no phase separation having occurred in this case during crystallisation rather than from slightly differing 117

concentrations between the two samples. The first explanation is more likely as will emerge from information to be presented in other chapters. The results seen here are very similar to those of the pan blended series except that there is no double peak approximately 10 0C below the carbazole melting temperature. This is because the sample had originally crystallised slowly into the more stable form rather than at 25 0C/min.

The absence

here of a small peak at the polyamide melting temperature is due to there having been no phase separation at the slower cooling rate.

3.4.3.2

Crystallisation for first DSC cooling ramp of molten blends at 2 0C/min

Figure 3-13

shows the re-crystallisation of the different materials from

Ampoule 1 after they have been heated to the melt and cooled at 2 0C/min.

Figure 3-13 First (re)crystallisation in DSC at 2 0C/min of Red, White and Fawn samples from Ampoule 1 and ampoule samples of polyamide-4,6 and carbazole.

a) The

“Fawn”

material

dilution-induced

again

freezing

shows

depression

the

characteristics

consistent

with

its

of

a

high

concentration of polyamide from the TGA, its (re)melting in DSC and similar performance of the samples with low carbazole concentration in the pan blended samples. 118

b) The white material displays quite a different re-crystallisation to that of the samples with similar carbazole concentration in the pan blended series. In that case we had a single crystallisation peak at approximately 210 0C but here there is a small peak just above that for carbazole and a carbazole crystallisation peak at approximately 220 0C. The reason for the difference is unknown but may be due to the different cooling rates used. c) The two crystallisation peaks of the red material are similar to the double peak of recrystallisation of the pan blended 38PA46Car

but

are slightly displaced from one another. The percentages of polyamide in the samples were similar (as mentioned previously) but prior cooling had not occurred in the same way.

3.4.3.3

Second DSC Melting at 5 0C/min of samples from Ampoule 1

Figure 3-14 Second melting in DSC of Ampoule 1 material at 5 0C/min heating rate and samples of polyamide-4,6 and carbazole from ampoules. The immediate prior thermal history of the blends was crystallisation in DSC at 2 0C/min.

The thermograms of Figure 3-14 for the second heating ramp of the three Ampoule 1 samples and the constituent materials from ampoules are similar to their first melting in the DSC. The exception is the sample 35PA46Car. 119

Here, there is a double main peak that comprises both eutectic melting and the closely followed dissolution of excess carbazole in a TLS peak. A small melting peak of polyamide-4,6 lamellae is found only 5 0C below the normal polyamide-4,6 melting temperature, signifying that the lamellae are in a virtually pure polyamide-4,6 environment.

All three peaks of this

thermogram are consistent with the prior crystallisation thermogram which had showed phase separation at high temperature of some of the polyamide-4,6 followed by major carbazole crystallisation with a final crystallisation of a eutectic mix of polyamide-4,6/carbazole.

3.4.3.4

Thermograms for Ampoule 1 samples on second DSC cooling ramp at 2 0C/min.

Figure 3-15 Second crystallisation in DSC of Ampoule 1 materials and polyamide-4,6 & carbazole from ampoules at 2 0C/min.

The set of thermograms in Figure 3-15

are, as expected, very similar to the

set in Figure 3-13 except that peak temperatures for the blends are usually 1-2 0C lower.

The small s form near 260 0C in 65PA46Car appears to be

noise in the instrument.

120

3.4.4

Later Ampoules Under Controlled Cooling Conditions

In this section, the later ampoules made under the programmed cooling control are investigated with melting and cooling in the DSC. The number of samples investigated is reasonably large so melting/crystallisation peak temperatures of all are provided in scattergrams and the thermograms of a selection are assembled in a separate figure to elucidate the scattergrams.

3.4.4.1

Melting Temperatures from first DSC melt at 5 0C/min of samples from ampoules

Figure 3-16 shows a selection of the DSC thermograms for melting polyamide-4,6/carbazole ampoule material.

It shows clearly the constant

eutectic melting peaks and TLS peaks for either polyamide-4,6 or carbazole. The main exception is the 48PA46Car that has both a separate melting peak.

Figure 3-16 Thermograms from first melting in DSC at 5 0C/min of ampoule blends of polyamide-4,6/carbazole, and displaying the consistent graph colouring scheme for different polyamide concentration ranges 20-29%, 30-39%, 40-49% etc. Figure 3-17

is the scattergram of the melting temperatures encountered during the first heating ramp of the DSC for samples from ampoules with various percentages of polyamide. 121

0

Melting peak temperatures ( C)

310 FirstPeak SecPeak ThirdPeak LastPeak

300 290 280 270 260 250 240 230 220 210 200 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 3-17 Polyamide-4,6/carbazole blend material from a variety of ampoules. Melting Peak Temperatures for first DSC heating ramp at 5 0C/min plotted against the weight percentage of polyamide derived by TGA.

At this stage it is difficult to definitively assign some of the melting peaks in Figure 3-17 to carbazole. There appear to be several peaks due to polyamide-4,6 having crystallised together with carbazole. These overlap the region where carbazole melts. It is possible to assign some of the melting peaks in cases where the known percentage of carbazole in the sample (from TGA) would otherwise have to be greater than 100% carbazole crystallinity. That can be determined by the melting enthalpy. The mass of carbazole in the sample is known from TGA and the enthalpy of fusion of carbazole is known from the literature so it is possible to exclude those peaks from melting (nearly) pure carbazole. This problem of assignment to one material does not occur with crystallisation because of the very distinctive crystallisation curve for carbazole (vide infra). The difficulties of peak assignment make Figure 3-17 more difficult to analyse but the peaks appear to be grouped into several sets. The first one is the group for the first melting peaks that are just below the melting temperature of carbazole. The above melting behaviour is consistent with increased melt temperature depression of the direct carbazole phases by polyamide as the polyamide percentage is made higher. 122

The last grouping is those of the last peaks lying above the line from near 50%, 280 0C to 100%, 290 0C.

This was increasing melting temperature

depression by up to 20 0C for nearly pure polyamide phase as the polyamide percentage was decreased.

Polyamide lamellae were embedded in nearly

pure amorphous polyamide, most likely as a result of phase separation. That was evident by the small Flory-Huggins interaction between the materials at those high temperatures. Phase separation had not taken place in every case because often there were a number of cases where the last melting temperature was as low as 236 0C, a drop of nearly 60 0C below that of the polyamide. Between the two groups of slightly depressed melting temperature peaks there were often other peaks at concentrations between 40 and 70% polyamide, mostly above the melting temperature of carbazole. It was inferred from this that these were highly depressed polyamide-4,6 peaks caused by interaction between the materials occurring in these cases. This was most likely with polyamide lamellae incorporating substantial amounts of carbazole in the inter-lamellar space. Alternatively it could, in principle, be due to some re-crystallisation of unstable forms taking place. The existence of a number of cases where the last melting peak was more than 30 0C below the melting temperature of the pure polyamide militates against the latter interpretation. A final melt closer to the melting temperature of polyamide would otherwise have been observed. It can be seen from Figure 3-8 for pan blended samples and Figure 3-17 for the ampoule blends that the melting temperature peaks from the first remelting after the formation of the blends are, as expected, displaying approximately similar information. The major difference is the absence in the latter of peaks for the conversion of metastable polyamide into stable forms just prior to the main melting peak.

3.4.4.2

Overall Crystallinity

The method for determining crystallinity is by using the endothermic peaks obtained during the reheating in a DSC.

123

100

Overall sample crystallinity (%)

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 3-18 Overall Percent Crystallinity vs weight percent polyamide for polyamide-4,6/carbazole ampoule material. Crystallinity was determined by comparing the enthalpy of melting during the first heating ramp in the DSC at 5 0C/min with the percentage polyamide weighted average of the Heat of Fusion for 100% crystallinity of each material.

The percentage of polyamide from the High Resolution, High Sensitivity (R4S4) TGA runs was used in the calculation of the overall crystallinity of the sample. This was combined with the known enthalpies of fusion of both the polyamide-4,6 and the carbazole to give the theoretical enthalpy expected on a proportional basis for 100% crystallinity. The actual total enthalpy of the first DSC heating ramp was compared with this to give the overall percentage of crystallinity in the sample. The results are plotted below in Figure 3-18. It can be seen that there is a very wide scatter of results in Figure 3-18. There is no discernible correlation between the percentage of polyamide and the level of overall crystallinity in the samples apart from a reduction from the high crystallinity of the carbazole sample. It is inferred from this that the prior crystallisation was from an unstable situation.

It should be

remembered here that the samples were all subjected to the same crystallisation regime in the ampoules. The cooling process was under close control so there appears to be a large statistical variation in the solidification process indicating an unstable crystallisation situation. The largest variation is in the range 45-75% polyamide and it is in just that region that the near concurrent crystallisation of the two materials was evident with samples 124

diverging

from

being

mainly

melting-depressed

polyamide

or

melting-depressed carbazole. We will see much more stable crystallinity in the next three chapters dealing with polyamide/carbazole blends.

Crystallisation Temperatures from first crystallisation in DSC at 2 0C/min

3.4.4.3

PA46Car material from a variety of ampoules Figure 3-19 displays the scattergram of first, last and other crystallisation temperatures against weight percent of polyamide. First DSC crystallisation peaks were usually encountered above 260 0C for 35-100% polyamide. Crystallisation depression from the pure polyamide crystallisation at 274 0C was generally greater at lower polyamide levels. Other first crystallisation peaks were regularly encountered at temperatures in the range 221-250 0C at 45-75% polyamide. These are too low to be simply a depression of polyamide lamellae crystallising and are considered to be due to the close crystallisation of the two materials at the same time. Others’ work shows

Crystallisation peak temperatures ( 0C)

that lamellae will not incorporate “foreign” materials within them. 290 polyamide-4,6 carbazole Peak1 Peak2 Peak3

280 270 260 250 240 230 220 210 200 190 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 3-19 First DSC crystallisation at 2 0C/minof polyamide-4,6/carbazole blends from ampoules. Temperatures of peaks for polyamide, carbazole and other peaks vs. weight percentage polyamide.

The distinctive carbazole crystallisation peak allowed definitive assignments of many of the peaks encountered.

It can be seen that carbazole

crystallisation peaks were depressed more as the level of polyamide increased, as would be expected. We will see that this is the case with other 125

material combinations with the melting temperature of the polyamide greater than that of the diluent. It can be seen, however, that there is much scatter below the general trend line, indicating that carbazole molecules are often trapped in the partially crystallised structure before being able to form separate crystallites. Some very small carbazole crystallisation peaks were not included in crystallisation peaks graphed in Figure 3-19 as they were very minor. They were below the major carbazole crystallisation peak and would only have confused the overall analysis of what was happening with the polyamide, the combined and the major carbazole peaks. The existence of these very small peaks is due to small groups of carbazole molecules that come into proximity with one another in the amorphous portion. These reach a state of lowered solubility and are able to crystallise together before being trapped by lower mobility at the continually reducing temperatures in the experiment. It can be seen from Figure 3-10 for pan blended samples and Figure 3-19 for that done in ampoules that the information from both graphs is not very different. The number of concentrations in the first one is more limited so the graph is sparser.

3.4.4.4

Crystallinity from first cooling

The high crystallinity of the carbazole portion in Figure 3-20 falls off with increasing polyamide in the blend, reaching zero above 75% polyamide. The percentage is extremely scattered in the range 30-75% polyamide.

The

polyamide-4,6 crystallinity is less scattered than the carbazole crystallinity. It drops slightly in the range 85% down to 50% polyamide but rises to 100% as the polyamide level is lowered further. This very high value near 20% polyamide is in the region where there is crystallisation at the normal polyamide crystallisation temperature. The high crystallinity is possibly due to phase separation taking place. Olmsted and co-workers [30] pointed to the common feature being found by themselves and others with many polymers that SAXS was indicating a reorganisation of polymer chains on a reasonable scale in the liquid before WAXD was showing crystal formation into lamellae. Olmsted et al. proposed that there was phase separation taking place within the pure molten polymer due to statistical density fluctuations into a denser, aligned, liquid phase 126

and a less dense phase. The denser “phase” would collapse further with the appropriate driving force of supercooling into the lamellar crystalline state. It has been mentioned in Chapter 1 with polymers that the application of pressure during crystallisation leads to higher crystallinity and higher melting temperatures.

Crystallinity of each material (%)

100 carbazole polyamide-4,6

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 3-20 Crystallinity of the Carbazole and the non-carbazole portions determined from Heat of Crystallisation at 2 0C/min during the first DSC cooling of ampoule blends of polyamide-4,6/carbazole.

There is a situation here where the polyamide is the minor phase that could result in forming droplets of polyamide in the carbazole due to phase separation. It is proposed here that interfacial tension between the phase separating materials should act on the polyamide droplets in a manner similar to the high pressure situation as described in Chapter 1.

This

should tend to produce a higher crystallinity if those droplets are at a suitable temperature to crystallise at that time. It will be seen in later chapters that the scatter is not evident in any of the three other polyamide combinations with carbazole but is to varying degrees with the polyamide/phenothiazine combinations. The common factor with the other three combinations with carbazole is that the polyamide melting temperature is lower than the diluent melting temperature.

127

3.4.4.5

Phase Diagrams for first time heating, and first time cooling ampoule material in DSC with later ampoules

At this stage it is possible to construct non-equilibrium phase diagrams for both heating and cooling. 300 TmCarDepr TmPA46Depr TmPA46Pure

280

Liquid Liquid with some PA46 crystallites

0

Melting temperatures ( C)

290

270 260 250

Solid & liquid

240

Solid & liquid

230 220 Solid

210 200 0

10

20


80

90

100

for

heating

Figure 3-21 Non-equilibrium phase diagrams for heating to the liquid state raw materials and blends of polyamide-4,6 (PA46) and carbazole (Car) demonstrating Flory-Huggins melt depression. Blends having peaks depressed are denoted in the legend by Depr.

Figure

3-21

shows

the

non-equilibrium

phase

diagram

polyamide-4,6, carbazole and their blends to the liquid state. It shows the Flory-Huggins depression of carbazole melting by polyamide-4,6 and the depression of polyamide-4,6 melting by carbazole plus later melting of some almost pure polyamide-4,6 crystallites near the polyamide-4,6 melting temperature. It is also of interest that in this case the polyamide-4,6 melting at polyamide concentrations below 35% wt are not depressed as far, showing non-linearity below 35 % wt polyamide for the interaction parameter. Figure 3-22 shows a very similar non-equilibrium phase diagram to that for heating in Figure 3-21 above.

Again there is Flory-Huggins style

crystallisation depression of each material by the other depression of the crossover point and non-linearity of the interaction parameter below 35% wt polyamide. There is a slight depression of transition temperatures at 35% polyamide where both materials melt or crystallise together. 128

0

Crystallisation temperatures ( C)

290 TcPA46Pure TcCarDepr TcPA46Depr

280 270

Liquid

Liquid with some PA46 crystallites

260 250 240

Solid & liquid

230

Solid & liquid

220 210 200

Solid

190 0

10

20


80

90

100

Figure 3-22 Non-equilibrium phase diagrams for cooling from the liquid state of raw materials and blends of polyamide-4,6 (PA46) and carbazole (Car) similar to the phase diagram for heating. Blends having peaks depressed are denoted in the legend by Depr.

3.5 Fourier Transform Infra Red Spectroscopy FTIR measurements were carried out to investigate whether there were any complexes being formed due to hydrogen bond interactions between the polyamide-4,6 and the carbazole. It will be shown in this section that there was no evidence for hydrogen bond interactions driving the formation of blends in the ampoules. The absorption frequencies of polyamide-4,6 and carbazole in both the mid-range and near infra-red are given in Chapter 1 and the experiments were carried out as described in Chapter 2.

3.5.1

Mid Infra-Red – Photoacoustic

Several polyamide-4,6/carbazole blends were made in ampoules. these, 62PA46Car and 79PA46Car are described in more detail.

Two of Both

display depressed peaks followed by TLS peaks during melting and two depressed peaks for the depressed peaks for the crystallisations of polyamide and carbazole. The results portrayed are typical of the spectra seen. A broad selection of spectra covering the concentration range are reproduced in Appendix D on CD.

3.5.1.1

N-H Stretches

The spectra in Figure 3-23 mathematical

model

where

compare the actual 62PA46Car curve with a the

spectra

129

of

the

starting

materials

polyamide-4,6 and carbazole were “added/combined” mathematically in different proportions to simulate the expected spectra from the combinations of materials. They were then scaled to give the same peak heights for the N-H stretches. This is described in Chapter 1. The solid lines in the set of figures below are the experimentally determined ones and the dashed lines are the models. The modelling was necessary, ss mentioned in Chapter 1, to handle the numerous

steep

overlapping

peaks

of

the

two

starting

materials.

Mathematical additions imply that no interactions were expected between the two materials.

It was mentioned in Chapter 1 that photoacoustic

spectrometry was used to look with high spectral resolution specifically for shifts due to hydrogen bond changes. Photoacoustic experiments could be carried out without modifying the morphology of the samples at all (unlike transmission spectroscopy or ATR). Spectral Addition 50 62PA46Car 45

Photoacoustic

40 35 30 25 20

3450

3400

3350 Wavenumbers (cm-1)

3450

3300

Figure 3-23 Comparison of a mathematical addition of starting material spectra with the actual 62PA4Car Photoacoustic IR spectrum from ampoule material for both N-H stretches in the mid IR region

It has been explained earlier that this does have the disadvantage that strong peaks will cause signal saturation resulting in some non-linearity of peak heights.

That is shown in Chapter 1 where the models for the

non-interacting composite give compressed peak tops.

These peaks are

expanded vertically in the overlay graph using the “Full Scale” option to give 130

peaks that can be easily compared with the measured ones for the polyamide-4,6/carbazole material from the ampoules. The area just above 3400 cm-1 is the N-H stretch of carbazole. The stretch (“free” N-H) in the mathematical model from the raw materials should be near 3420 cm-1. The rounding of the peak tips is an artefact of expanding the models based on non-linear spectra to give the same peak heights for comparison. The polyamide-4,6/carbazole samples have lesser signal saturation because the absorbing bonds are physically more dilute in the blend. The positions of peaks are the critical issues rather than the rounding. The peak measured with the 62PA46Car material was within 2 cm-1 of the 3420 cm-1 peak. The measured peak should have moved by 15 to 30 cm-1 towards the right because it was now “bound” to the carbonyl oxygen of an amide group if there was significant hydrogen bonding occurring between the carbazole and the polyamide-4,6 along the lines of the work on poly(vinylidene fluoride)/polyamide-11 blends and Joussot-Dubien, Engel and Vert [120] on polyamide-6/Calcium chloride complexes. The region near 3300 cm-1 is the N-H stretch of the polyamide. This is assumed to be “bound” (hydrogen bonded state) as hydrogen bonds in polyamide are almost always fully consummated for crystalline regions and are very high for the amorphous regions. The measured N-H stretch due to the polyamide-4,6 in the blend is within 3 cm-1 of the mathematical model. There should have been a noticeable shift to the

left if

there

was

hydrogen

bonding taking

place

between

the

polyamide-4,6 carbonyl oxygen and carbazole N-H because there should be more “free” polyamide N-H bonds than normally experienced. An example of the expected type of shift is from work

where a change from 3300 to

3390 cm-1 was found when destroying polyamide-6,6 hydrogen bonds with KI or iodine.

Spectra from the solidified material from each of the

polyamide-4,6/carbazole blends gave peaks at 3341 and 3421 cm-1, identical to those expected from the constituent materials. Spectra can be seen in Appendix D on CD. There was no evidence of shifting of major N-H stretches for “unbound” carbazole and “bound” polyamide to “bound” carbazole and “unbound” polyamide.

The fact that there was no movement of the two

131

peaks toward each other (compared to the model) is very important in showing a lack of hydrogen bond interaction between the two materials. Similar results were found for all other polyamide-4,6/carbazole blends. Spectra can be seen in Appendix D on CD. NIR studies below were carried out to confirm this situation. Figure 3-24 shows an example in the comparison between the actual FTIR spectrum (solid line) in the detailed 1700 to 1100 cm-1 region for material from ampoule sample 48PA46Car and the result of a mathematically scaled combination of the spectra (dashed line) of the starting materials.

Again,

there are no clear shifts of peaks to the left or right evident but there may be some differences in the relative heights of the peaks. This lack of significant shifts, as distinct from changes in relative heights, has been found with each spectrum examined.

It would be imprudent, without other evidence, to

attribute any of the peak height differences to anything other than artefacts caused by the mathematical modelling of peaks that may have been highly absorbing, giving signal saturation. It has been surprising that the actual spectra have been able to be modelled so well.

60

Photoacoustic

55 50 45 40 35 30 25 1700

Spectral Addition 62PA46Car 1600

1500 1400 1300 Wavenumbers (cm-1)

1200

Figure 3-24 Photoacoustic FTIR spectrum in the “fingerprint” region of the Mid IR range for ampoule material 22PA46Car and the model made by mathematically adding the PAS spectra for the constituent materials.

132

There are some differences between “models” based on spectra from the raw materials as processed in ampoules and measurements on blends, particularly in the “fingerprint” region. It must be remembered that instead of having a solid of highly crystalline small molecules or a predominantly

α-form of polyamide-4,6 lamellae within an amorphous polyamide-4,6 solid that we have a far more complex morphology in the blend and that will affect the spectrum of a blend. With a blend, there will most likely be a mixture of amorphous polyamide and carbazole in the inter-lamellar space. There will possibly also be nodules of crystalline carbazole embedded in a matrix of combined

amorphous

polyamide

and

carbazole.

The

percentage

of

amorphous polyamide will most likely have increased (based on Figure 3-20) and the crystallisation of the polyamide with carbazole between the lamellae has possibly modified the lamellar structure of the polyamide. All of these morphological changes in both materials will lead to alterations in some of the peaks. It is well documented that differences in crystallographic form [119, 153], lamellar perfection [154], amorphous content [82] and similar factors with single materials cause quite conspicuous changes in absorption bands. The intimate proximity of carbazole to polyamide-4,6 in the interlamellar space and in the amorphous regions would therefore be expected to modify the various absorption resonances in addition to those associated with changes to crystallographic form, amorphous content and crystal fold perfection This means that the above spectral differences do not necessarily require specific polyamide-diluent interactions to explain them.

3.5.2

Near Infra-Red – Diffuse Reflectance Infra-red Fourier Transform

Near Infra-Red measurements were carried out with the DRIFT technique in order to obtain spectra without altering the morphology of the solidified samples. Several experimental approaches had been tried and DRIFT gave the best signals in the NIR range. The spectra of the constituents and the sample being investigated were examined in the ranges 6920 to 5700 cm-1 and 5000 to 4200 cm-1 as indicated by Wu and Seisler in their NIR work on polyamide-11 [141]. No overtone bands in the NIR region have specifically been found for polyamide-4,6 in the literature. The bands for polyamide-11 are expected to be close to those of polyamide-4,6.

133

0.18 carbazole polyamide-4,6 0.17 23PA46Car

Absorbance

0.16 0.15 0.14 0.13 0.12 0.11 0.10 0.09 7000


6000

Figure 3-25 7000 to 5700 cm-1 on 23PA46Car blend, polyamide-4,6 and carbazole samples from ampoules, demonstrating that the blend takes on the polyamide hydrogen bonding peaks without modification.

It was not necessary to form mathematical models of theoretical mixtures because of the broad nature of the peaks. No measurable changes in peak positions were found. Examples of this are shown in Figure 3-25 & Figure 3-26 which display the DRIFT spectra in the NIR ranges of interest for the 23PA46Car and the ampoule materials from the constituent materials. In both cases, the solid line for the combined material has taken on some aspects of each of the curves for the constituent materials. Those peaks for the melt blended sample can be clearly seen to be unmoved in frequency from the corresponding peaks for polyamide-4,6 and carbazole. The regions are specifically for hydrogen bonding in the polyamide so it can be seen that polyamide hydrogen bond interactions are unchanged by the melt blending process with carbazole. Similar outcomes were found for all other samples investigated. Appendix D on CD shows spectra to avoid large numbers of graphs in the chapter. DRIFT is a different FTIR experimental technique to the photoacoustic one discussed earlier. It also has been used to cover a different spectral region where hydrogen bond interactions can be probed. It should therefore be a compelling argument that the interactions between the materials during crystallisation from the melt has not been driven by a preferential hydrogen 134

bond interaction between the two materials displacing the virtually fully consummated intra- and inter-molecular polyamide-polyamide ones. 0.28 0.27

carbazole polyamide-4,6 23PA46Car

Absorbance

0.26 0.25 0.24 0.23 0.22 0.21 0.20 5000

4750

4500 -1 Wavenumbers (cm )

4250

Figure 3-26 NIR Diffuse Reflectance Infra-Red measurements in the region 5000 to 4200 cm-1 on 23PA46Car blend, polyamide-4,6 and carbazole samples from ampoules, demonstrating that the blend takes on the polyamide hydrogen bonding peaks without modification.

3.6 Small Angle X-ray Scattering 160

Scattering Intensity

polyamide-4,6 140

64PA46Car

120

62PA46Car 31PA46Car

100

carbazole 80 60 40 20 0 0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

2Θ Figure 3-27 SAXS spectra at different scattering angles for polyamide-4,6, carbazole and polyamide-4,6/carbazole material from ampoules (with background removed and graphs smoothed).

135

1.2

The Small Angle X-ray Scattering (SAXS) experiments were carried out as described in Chapter 2 in order to gain an understanding of the lamellar spacings in the composite material compared to the starting materials. Figure 3-27

displays the SAXS spectra of polyamide-4,6, carbazole

and

several of the samples from various polyamide-4,6/carbazole blends. They are the smoothed results (after removal of background scattering) of intensity vs. the scattering angle 2θ for the samples run. Carbazole appears to have faint peaks at angles corresponding to 11.3 and 14.7 nm. These are so weak as to be insignificant.

Polyamide-4,6 has a

main peak at 2θ = 0.71 equivalent to average Bragg spacing of 12.3 nm. The spacing in blend samples varies in the range 0.75 < 2θ < 0.85, which gives average Bragg spacings between 10.5 to 11.3 nm, slightly smaller than that for the unblended polyamide-4,6. The interpretation of these limited results is unclear but could well be associated with the diluent molecules being small molecules and changing the characteristics of the interlamellar space.

3.7 Nuclear Magnetic Resonance Spectroscopy Nuclear Magnetic Resonance spectroscopy (NMR) can give information about molecular configurations in materials.

Solid state

13C

CP/MAS NMR was

run on only a few samples of material due to limited equipment availability and limited samples available at the time the experimental work could be carried out. The experiments were performed by Dr Andrew Whittaker of the University of Queensland, as described in Chapter 2.

The experiments

described here, however, came up with somewhat ambiguous results. that could be interpreted as either molecular interactions between the materials or as morphological differences. The samples that could be run with NMR were the polyamide-4,6, carbazole and “Red” material from the first ampoule. This last was quite crumbly. The “white” material from the same ampoule was a very hard material.

Dr.

Whittaker was unable to crush it to a fine enough powder for the NMR experiments because of the hardness and the available timeslot for the experimental work precluded running the trials with the white material at a later date. Assignments for NMR were made assisted by the work of de Vries {De Vries, 1989 #313} on polyamide-4,6 and oligomers. Small but significant effects were found with the “Red” material from Ampoule 1. 136

4000000 polyamide-4,6 35PA46Car carbazole

Intensity

3000000

2000000

1000000

0 250

200

150

100

50

0

Chemical shift (δ δ) Figure 3-28 13C Solid State NMR spectrum for Ampoule 1 "Red" material compared with polyamide-4,6 and carbazole.

Figure 3-28, the overall

13C

Ampoule 1, polyamide-4,6

Solid State NMR spectra of “Red” material from and carbazole shows similarities between the

“Red” material and both the spectra of the originating materials, as would be expected.

The two figures Figure 3-29 & Figure 3-30 below show

important sections in more detail where some differences may be seen. 2500000 polyamide-4,6 35PA46Car

Intensity

2000000

carbazole

1500000

1000000

500000

0 180

178

176

174

172

170

Chemical shift (δ δ) Figure 3-29 Solid State NMR carbonyl shift for polyamide-4,6, carbazole and red material from the Ampoule 1 blending.

137

168

Figure 3-29

covers the region where the Carbonyl shift in polyamide-4,6

occurs. It can be seen from the figure that the peak from the polyamide-4,6 has become less intense

and broadened slightly and that there is a very

marginal shift in the peak to the left. Figure 3-30 displays several changes in the spectrum of the “Red” material compared to the polyamide-4,6. The polyamide-4,6 peak for the C next to N, near a chemical shift of 42, shows a broadening, a small downfield shift and a slight reduction in intensity for the “Red” material from Ampoule 1. The C next to the Carbonyl C, near a shift of 36, exhibits broadening and intensity reduction. The higher peak at a shift of 27 of the double peak and the other peak at a shift of 25.5 both show broadening and intensity reductions.

The first of

these is for the C that is two from N and the peak at 25.2 is for the C two from Carbonyl C. These assignments are from the work of de Vries, Linssen and van der Velden, [155] from their work which was able to make assignments on polyamide-4,6, polyamide-6,6 and polyamide-6 for the main chain and for both the end groups. 4000000 polyamide-4,6 35PA46Car carbazole

Intensity

3000000

2000000

1000000

0 50

45

40

35 30 Chemical shift (δ δ)

25

20

Figure 3-30 Solid State NMR of polyamide-4,6, carbazole and “Red” material from Ampoule 1

The results were originally interpreted by Dr. Whittaker as more likely to be molecular interactions than morphological but they were ambiguous. 138

We

know from the FTIR results that hydrogen bond interactions are excluded and that with SAXS we are seeing lamellar spacings that are different. The interpretation is thus weighted towards the morphological differences being the cause of the slight shift with the C next to N and general broadening of all peaks. The work of Howe et al. [156, 157] was on

13C

NMR of poly(L-lactic

acid)/urea of inclusion compounds where there are mechanical constraints between intimately co-located molecules of differing types.

Their work is,

perhaps, a parallel to the types of complex morphological environments encountered in this work. They found strong effects on solid-state spectra

from

differences

in

the

conformations

and

chain

13C

NMR

packing

environments of their inclusion compounds.

3.8 Summary This chapter has shown

evidence for the

development of complex

morphologies being developed when polyamide-4,6 and carbazole are taken to the melt and solidified. This has been done on a small scale with DSC monitoring of the powders being melted, the melt being cooled quickly and being remelted giving information about the crystallographic development and phase separation during fast cooling.

Work on larger amounts melt

blended in ampoules and characterised afterwards in the DSC gave information about the morphology with slow cooling. FTIR measurements with Photoacoustic in the Mid-range IR and DRIFT in the Near Infra-Red were found to give null results for hydrogen bonding interactions between the materials. The outcome of this investigation into the melt blending of polyamide-4,6 and carbazole gives a picture of the two materials partially interacting in a non-hydrogen bonding manner.

There is also a tendency for the high

temperature

limited

solution

to

have

polyamide

solubility

at

high

temperatures and low polyamide concentrations with evidence of phase separation. Fast cooling of blends at mid concentrations leads to the two materials crystallising at temperatures 50 0C below the normal polyamide crystallisation temperature and 25 temperature.

Metastable

forms

0C

of

below the carbazole crystallisation

polyamide-4,6

crystallise

situations with carbazole most likely in the inter-lamellar space. 139

in

those

The thermograms enlighten us to the phase diagrams involved with these melt blends although the dynamic mode of the heating and cooling ramps introduces kinetic effects that have to be considered. We have found evidence in this chapter for polyamide-4,6 and carbazole crystallising together from the melt but that this has not involved hydrogen bond interactions between the two materials. It was possible to observe the dissolution of polyamide-4,6 particles by carbazole powder at temperatures below the melting temperature of the carbazole by exploring the process with small-scale blending in the DSC.

Remelting and recrystallising these

samples gave more understanding of the initial process and how changes occurred once material had solidified from the melt.

Larger amounts of

material for use with other techniques required making samples in ampoules. The picture that emerges is of the formation of crystallites of the polyamide and/or carbazole and/or a combination of the two in a matrix of amorphous material on solidification from the melt. The outcome depends on polyamide concentration. and the conditions of solidification.

The

formation of combined material occurred when the percentage of polyamide was 45-75%. It was possible to observe the dissolution of the polyamide-4,6 particles by carbazole powder just melting at approximately 230 0C when carrying out the blending in the DSC.

Polyamide-4,6 normally melts at approximately

290 0C. Carbazole melts at 246 0C but the heating of the two powders together allowed the very initial melting of carbazole to dissolve polyamide. The dissolution of polyamide-4,6 occurred fully at temperatures below the normal completion of carbazole melting if there was sufficient carbazole. There was some polyamide left to melt into the solution at near-normal polyamide-4,6 melting temperatures if the percentage of polyamide was high. The solubility of polyamide-4,6 in the carbazole was low and not all the polyamide dissolved when the weight percentage of polyamide was low. This left some to melt into the solution at near-normal polyamide-4,6 melting temperatures. Depression of the melting temperature of carbazole was seen when the level of polyamide was high and depression of the residual polyamide melting was also observed.

140

The remelting of melt blended solid mixes of polyamide-4,6 with carbazole occurred at lower temperatures than the melting of the powders. The melts occurred over a shorter range of temperatures and often exhibited double melting peaks, indicating that material crystallised quickly had been in a metastable form. The first DSC melting of materials made in ampoules was consistent with the remelting of blended material formed in small amounts in the DSC itself. An exception to the above temperature reductions on remelting was observed in the case where the polyamide-4,6 level was high and the melting temperature depression was reduced because of carbazole evaporation. This led to the sample more closely approximating pure polyamide-4,6. Crystallisation of material blended in the DSC pans showed very distinctive crystallisation peaks for forming carbazole crystallites, allowing definitive assignment to that material.

Some peaks were seen close to the normal

crystallisation temperature of the polyamide and this was due to lamellae of polyamide forming at slightly freezing-depressed temperatures due to the presence of carbazole. A wide range of other peaks caused by combined crystallisation was observed when the level of polyamide was between 45 and 75%. The non-carbazole

crystallinity

had

a

great

scatter

showing

that

the

crystallisation process was unstable or marginal. A sudden drop in crystallinity of carbazole at levels above 45% polyamide indicated the incorporation of carbazole into the interlamellar structure. This observation was

consistent

with

carbazole

being

consumed

in

the

combined

crystallisation. FTIR was carried out in the Mid-IR with photoacoustic spectroscopy and also with a different technique, DRIFT, in the Near Infra Red. Both techniques showed unambiguously that there was no hydrogen bond interaction between the N-H of the Carbazole and the O=C of the amide groups in any of the samples of polyamide-4,6/carbazole blends. For that to have happened the N-H stretches for carbazole in the Mid IR would have shifted from a “free” condition to a

“bound” condition and some of the polyamide N-H bonds

would have changed from a “bound” condition to a “free” one. The expected shifts of 15 to 30 cm-1 in opposing directions did not occur. In the sensitive 141

NIR region, there were no changes in bands associated with hydrogen bonding in polyamides, also requiring a conclusion that the combined crystallisation observed with the DSC did not involve inter-material hydrogen bonds. No shifts were seen with photoacoustic (PAS) measurements in the Mid IR fingerprint region although the relative heights of some narrow adjoining peaks were different from that expected from adding the PAS curves of the starting. These discrepancies would most likely be due to the tops of highly absorbing peaks being compressed, as often happens with PAS or due to the stereo-chemical proximity of the molecules in where combined crystallisation has taken place. The broader peak tops seen in the mathematical models for the N-H stretches of the polyamide and the carbazole are caused by significant truncation of peaks for the starting materials.

The signal saturation was

less in the case of the polyamide/carbazole mixes because of physical dilution of IR absorbing centres within the sample volume being probed by the photoacoustic experiments. The NMR results, while ambiguous, are consistent with morphological differences in the polyamide of the blends being different from normal polyamide-4,6 and these may well be due to twisted polyamide-polyamide hydrogen bonds for the metastable configuration. The picture that emerges from this work is of the polyamide-4,6 and carbazole

being able to crystallise within the same lamellar/interlamellar

structure without hydrogen bond interactions between the constituent materials but involving a crystallisation into a metastable form during the crystallisation. That appears possible in the range 45-75% polyamide-4,6 in the blend. The overall picture for the combination polyamide-4,6 melt blended with carbazole is that there is eutectic melting of solidified blend material with excess of one or the other material only being dissolved in the liquid at much higher temperatures.

Fast prior cooling leads to metastable lamellar

formation if there is a high level of carbazole in the mixture and initial melting of the powders is much more complicated than if the materials have 142

been melted together beforehand. High concentrations of carbazole lead to phase separation of polyamide. Each material depresses the crystallisation of some of the other in proportional to its concentration showing a Flory-Huggins like interaction. Fast crystallisation leads generally to suppression of crystallisation up to 50 0C below the normal polyamide crystallisation temperature but this leads to the metastable lamellar formation mentioned above. The crystallinity of the carbazole is widely variable with this material combination.

143

Chapter 4

POLYAMIDE-6 WITH CARBAZOLE CONTENTS 4.1

Introduction

144

4.2


145

4.2.1

Sample Moisture Content

145

4.2.2

TGA Reproducibility

146

4.2.3

Ramp method comparisons

147

4.2.3.1

4.3

147


4.3.1

Melt Blending in Pans

4.3.1.1 4.3.1.2 4.3.1.3 4.3.1.4

4.3.2

148 0

Melting Temperatures for first heating ramp of the powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min 0 Melting Peak Temperatures for second heating ramp at 5 C/min 0 Crystallisation Peak Temperatures for second cooling ramp at 25 C/min

Ampoule Material

4.3.2.1 4.3.2.2 4.3.2.3 4.3.2.4 4.3.2.5 4.3.2.6 4.3.2.7 4.3.2.8

4.4

Measured concentration of polyamide in samples

149 149 151 153 155

155

Accuracy of Melting & Crystallisation Enthalpy Measurements 0 Melting Temperatures (First melt in DSC) from heating at 5 C/min of material blended earlier in ampoules Overall Crystallinity 0 Crystallisation Temperatures at 2 C/min cooling for ampoule samples crystallising for the first time in the DSC. Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC 0 Third Melting of materials from ampoules/Second DSC Melt at 5 C/min Third Crystallisation of Materials/Second DSC Crystallisation


155 156 158 159 161 163 163 164

165

4.4.1

Mid Infra-Red - Photoacoustic

165

4.4.2

Near Infra-Red - DRIFT

168

4.5


170

4.6

Summary

171

4.1 Introduction This chapter extends the work done on polyamide-4,6 melt blended with carbazole with further work concentrating on melt blending polyamide-6 with carbazole. It will be seen in this chapter that many of the general characteristics seen in Chapter 3 are found again here. The similarities are evident despite the melting temperature of polyamide-6 being lower than 144

that of carbazole. There are also a number of significant differences caused by the much lower melting temperature of polyamide-6. In this case, the melting temperature of carbazole is above that of the polyamide. There are also other differences caused by polyamide-6 easily crystallising in a metastable form. Some of the characteristics are slightly less easy to distinguish here because of the closeness of melting temperatures of polyamide-6 and carbazole.

4.2 Thermogravimetric Analysis TGA provided a gross mechanism to monitor potential moisture absorption in ampoule samples as well as the normal determination of carbazole levels. 4.2.1 Sample Moisture Content TGA evaluation was made of evaporation taking place at 150 0C as a measure of the potential inclusion of moisture in samples. The level was difficult to determine precisely because carbazole will also begin evaporating at relatively low temperatures in the nitrogen gas stream of the TGA furnace.

Figure 4-1 Thermogravimetric analysis of wet and dry polyamide-6, carbazole and a blend to determine the effect of moisture.

Comparisons can only be made between evaporation rates with known low and high percentages of moisture in polyamide and the “normal” situation 145

with carbazole that has been equilibrating for a long time in a dry glovebox. Figure 4-1 demonstrates the evaporation for some materials in a TGA instrument. Polyamide-6 was used as it can absorb relatively high amounts of moisture.

Carbazole, freshly vacuum-dried polyamide-6, a sample of

polyamide-6 that had been kept in a humidity cabinet at 52 0C and 100% relative humidity for 2 months, and a sample from an ampoule with 85% polyamide-6 with carbazole underwent TGA trials. It can be determined from this work that the percentage weight loss at 150 0C for “wet” polyamide-6 with 12% moisture content is a loss of 4% at 150 0C at 10 0C/min. The inflection at 220 0C is due to the reverse reaction for polymer condensation because of the amount of water present. Vacuum dried polyamide-6 had a weight loss of 0.07% by that temperature. The 85PA6Car ampoule sample and carbazole from the glovebox have weight losses of 0.7 and 1.7% respectively These results could partly be evaporation of the carbazole and, potentially, partly be residual moisture or other contaminants. The carbazole purchased from Sigma-Aldrich was only available at 97% purity. It could indicate significant problems in sample handling if there were to be a high loss (greater than 1.7%) of material by 150 0C. The 85PA6Car sample did not lose as much material as the carbazole sample by 150 0C because there was less carbazole in the sample. No

problem

samples

were

encountered

with

moisture

in

the

TGA

measurements on ampoule samples from any of the combinations of polyamide-4,6,

polyamide-6,

polyamide-6,9

and

polyamide-6,12

with

carbazole or phenothiazine. The check did provide a simple form of quality control in a gross sense. 4.2.2 TGA Reproducibility Some samples taken from near each other in bulk pieces from ampoules were measured in duplicate runs with the R4S4 ramp method, as had been done with polyamide-4,6/carbazole blends. The number of samples of meltblended material made with polyamide-6 and carbazole was much less than for polyamide-4,6 and carbazole. Those, where the samples were taken from similarly coloured sections close to one another, gave concentrations of polyamide within 5% of each other, as with the polyamide-4,6/carbazole blends. 146

4.2.3 Ramp method comparisons The results of a comparison of polyamide percentages in samples taken from near each other and of a similar colour for the two TGA ramp methods (R4S4 and “straight”) are shown in Figure 4-2 below. It can be seen that the measured concentrations of polyamide are generally within five percent of one another, approximately the same as for any two samples measured with the same method. The results confirm those for polyamide-4,6 blends with carbazole.

Polyamide concentration by "straight" method (%wt)

100 PA6Car 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration by "R4S4" method (%wt) Figure 4-2 Comparison of TGA plateau levels at 300 0C between R4S4 and "straight" methods for polyamide-6/carbazole blends, demonstrating the similarity in measurement outcomes.

4.2.3.1 Measured concentration of polyamide in samples Figure 4-3 below shows a comparison between the theoretical percentage of polyamide based on the weights blended in ampoules and the actual concentrations of polyamide in different samples taken from the ampoule material.

Some of the samples taken from the same ampoules were of

differing colours. The

samples

often

deviate

by

more

than

10% from

the

expected

concentration based on the weights of raw materials used. This indicates that they are real variations in material composition encountered within ampoule samples due to uneven mixing as they are well above the maximum 5% variation expected.

That had also occurred with the first ampoule 147

sample (polyamide-4,6/carbazole) which separated into red, white and fawn


parts of radically different polyamide concentrations. 100 PA6Car 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 4-3 Actual weight percentage polyamide (from TGA plateau values at 300 0C) against. expected weight percentage polyamide from weights of materials added to the ampoule in polyamide-6/carbazole blend samples.

4.3 Differential Scanning Calorimetry Polyamide-6 has a double melting endotherm with a main peak at 224 0C and usually a peak preceding that in the range 207-217 0C. Peaks at the lower end of this range can be almost invisible but peaks near 217 0C can be quite conspicuous. The size and position of the first peak depends on the prior thermal history of the material. The lower peak is due to the polyamide having had the opportunity to crystallise into thin lamellae with a metastable form. The metastable form melts when the sample is reheated and recrystallises into the stable form that melts at the higher temperature of 224 0C. The metastable form is encountered more strongly when the prior cooling rates had been faster. It should be remembered that the original pan blended samples reported here were usually made at the higher cooling rate of 25 0C/min in order to minimise carbazole loss by minimising the time at high temperature. The ampoules were cooled at 2 0C/min to allow crystallisation or phase separation to occur at a relatively leisurely pace. The subsequent DSC evaluation applied to the ampoule samples also utilised the slower cooling rate despite the associated loss of carbazole. The effects of 148

that loss are noted below on occasions. The polyamide crystallises at 198 0C for a cooling rate of 2 0C/min and at 173 0C for a cooling rate of 25 0C/min. 4.3.1 Melt Blending in Pans 4.3.1.1 Melting Temperatures for first heating ramp of the powders at 5 0C/min The thermograms of the melting of polyamide-6, carbazole and the melt blending of some mixtures of polyamide-6 and carbazole powders are shown in Figure 4-4.

Figure 4-4 DSC thermograms during the first melting at 5 0C/min of polyamide-6, carbazole and powder mixtures melt-blended in the DSC.

The individual thermograms are described in detail below: a) The polyamide-6 thermogram has an almost imperceptible first melting near 208 0C that can only be detected by analysing temperature derivatives of the heat flow. This most likely means that the original pellets were cooled very quickly during manufacture. Based upon experience, the grinding of pellets into powder would have introduced some stress into the surface regions of the sample and the vacuum drying at 110 0C would have relaxed some of that stress. The effect on thermograms is usually evident to a minor degree below 149

150 0C. The material is identical to that used to make the blends seen in the thermograms. b) Introducing some carbazole into the 71PA6Car sample has a strong effect, causing a broad melt at substantially lower temperatures than for polyamide-6 and comprising three peaks in the range 180-215 0C. It is probable that, early on, a very small amount of amorphous polyamide-6 has started to soften and dissolve carbazole forming a lower viscosity eutectic. That liquid is then in a position to dissolve polyamide-6

lamellae

at

temperatures

lower

than

the

normal

polyamide-6 melting temperature. c) The case of 50PA6Car with equal weights of polyamide-6 and carbazole extends this further, resulting in a broad single melting peak having a maximum near the lowest of the three peaks of the previous case. The melting process is now more or less complete at a slightly lower temperature than previously but there is some slight endothermic activity between 210 0C and the normal melting temperature of carbazole. This composition could possibly be just beyond the situation of maximum solubility of carbazole in polyamide-6 at those temperatures, leaving some carbazole to dissolve later in the eutectic liquid only at higher temperatures. There does seem to be a faint peak in the tail of the thermogram for an extremely small amount of carbazole melting at the carbazole melting temperature. d) The case with 38% polyamide-6 in the mixed powders clearly leads to carbazole dissolving in polyamide-6 at similar temperature to the 50PA6Car. There is, however, more carbazole unable to dissolve in the high temperature solution. That remaining carbazole melts/dissolves fully

only

at

temperatures

approaching

the

normal

melting

temperature of carbazole. The broad, slightly skewed, peak for the remaining carbazole is most likely due to solubility of carbazole in the solution increasing slowly as the temperature increases during the experiment. There is a slight inflection in the peak near 243 0C implying that the last of the remaining carbazole is only melting at the peak melting temperature of carbazole due some sort of a solubility restraint. 150

Overall, we see here a eutectic liquid forming at 20 0C below the polyamide melting temperature as carbazole is being dissolved into a solution of carbazole and polyamide-6. There is later dissolution of residual polyamide-6 lamellae if there is substantially over 50% polyamide in the powder mix or later dissolution of excess carbazole if the concentration is 50% or more carbazole.

In each case, the excess is only dissolving with the increased

solubility at elevated temperatures giving rise to a TLS peak. There is some evidence in the blends with the two lower polyamide concentrations that the normal carbazole melting temperature is necessary to melt the last of the carbazole. 4.3.1.2 Crystallisation for first cooling ramp of the molten blend at 25 0C/min A set of thermograms is shown in Figure 4-5 for the first crystallisation at 25 0C/min for powders melt-blended in pans.

They are displayed with

thermograms for the molten raw materials subjected to the same cooling ramps.

Figure 4-5 DSC thermograms for the first crystallisation of pan blended polyamide-6/carbazole and the constituent materials cooling from the melt at 25 0C/min.

151

a) The 71PA6Car sample shows no separate crystallisation of carbazole and only an exotherm lower than the normal crystallisation for polyamide-6, indicating that carbazole is being locked into the polyamide-6 lamellar structure. The polyamide-6 crystallisation is being hampered to a large extent by the presence of the carbazole, resulting in a drop in peak crystallisation by 15 0C. The peak is much broader due to the hampering of the crystallisation, taking place later and occurring over a longer period of time. There is also a small residual crystallisation peak evident near 150 0C in the tail of the main peak. b) The trace for 50PA6Car shows 2 peaks, an initial one at 205 0C due to the hampered crystallisation of carbazole and also, at a lower temperature, a narrower

peak

than

for

71PA6Car.

This

second

peak

is

approximately.10 0C below that of pure polyamide-6 and slightly higher than for 71PA6Car. The interpretation of this is that each material is largely crystallising independently, as polyamide poor and polyamide rich crystallites, hampered by the presence of the other material. The shape of the carbazole crystallisation deviates from those seen in Chapter 3 in that the “spikiness” is absent. It is probable that there is polyamide-6 being incorporated at a molecular level in the carbazole and these less “spiky” peaks will be encountered in other chapters as well. The crystallisation depression of carbazole is reasonably large at 25 0C. c) The thermogram for the highest level of carbazole in the blends, 38PA6Car, showed crystallisation depression for carbazole but not so much as for the 50PACar sample and the peak was also not a “spiky” one. The depression of the carbazole crystallisation was just under 10 0C. There was also a second peak of the same form and near the same temperature as for 50PA6Car. The differences in carbazole crystallisation temperature between these last two thermograms are most likely merely due to a reduced level of polyamide hindering the crystallisation of carbazole for the 38PA6Car sample. We see from this set of thermograms that there seems to be a tendency for hindered crystallisation of both carbazole and polyamide-6 when the concentration of polyamide is the same as or lower than that of carbazole. Some form of phase separation is possibly occurring although the non-spiky 152

form of carbazole crystallisation means there is certainly some inclusion of polyamide-6 in the carbazole structure. There is a single, broad, crystallisation peak at even further depressed temperatures when the concentration of polyamide in the melt is appreciably higher than that of the carbazole.

The constancy of the second

crystallisation temperature is indicative of eutectic crystallisation. It will be seen in later chapters that there is often a slightly lower crystallisation temperature at the concentration where both materials crystallise together at the same time. A large difference exists compared to Chapter 3. 4.3.1.3 Melting Peak Temperatures for second heating ramp at 5 0C/min The set of thermograms in Figure 4-6 differ slightly from the initial melting of the powders in Section 4.3.1.1.

This occurs similarly to Chapter 3 where

there was a difference because the remelting is starting from a more intimate mix of the materials. There are also differences evident compared to the first melting due to the way the polyamide had previously crystallised with a fast cooling rate.

Figure 4-6 DSC thermograms from heating at 5 0C/min for the second melting of materials previously crystallised in DSC pans for polyamide-6/carbazole blends and the constituent materials.

153

a) The blend sample 71PA6Car with the highest level of polyamide shows a clean, single melting peak set towards the higher end of the original broad multiple melting endotherm after a faint dip beforehand. b) We find a double peak now for the (substantially depressed) polyamide-6 melt in the 50PA6Car blend, indicating that the polyamide was melting and recrystallising from the metastable form to a stable lamellar form before melting the stable form. The double peak of the metastable form melting, recrystallising and melting was not evident in the above remelting of 71PA6Car. The melting endotherm at higher temperatures is similar to that seen with the first melting of 38PA6Car.

There is,

however, much less residual carbazole to be dissolved. The process is complete with the smaller amount of carbazole well before the normal melting temperature of carbazole is reached. There is, perhaps, a trace of this thermal activity evident at 220 0C in the 71PA6Car sample. The first time the powders of this 50PA6Car sample were melted (see Figure 4-4) there was indeterminate thermal activity in the region between the depressed melting of polyamide-6/dissolution of carbazole because the carbazole was still a powder. This time, the materials were more intimately entwined and the melting was more straightforward. c) The polyamide-6 double melt in 38PA6Car is almost identical in form and temperatures to that seen in the 50PA6Car sample. The melting endotherm at higher temperatures is similar to that for the same sample during the first melting of the powders seen in Figure 4-4. The melting curves are cleaner in the second melting of blends. The prior fast crystallisation has now lead in the cases with lower polyamide concentration to polyamide crystallising first in the metastable form often encountered with this polyamide at faster cooling rates. The sample with higher polyamide-6 concentration is different in that the melting is at a higher temperature. It does not have a double melting peak, indicating that it is in the more stable form. The 50PA6Car (and possibly the 71PA6Car) show similar carbazole dissolution processes to the 38PA6Car sample but are cut off earlier due to less excess of carbazole. 154

4.3.1.4 Crystallisation Peak Temperatures for second cooling ramp at 25 0C/min

Figure 4-7 DSC thermograms of the second crystallisation at 25 0C/min of pan blended polyamide-6/carbazole samples and the constituent materials.

The thermograms in Figure 4-7 for the second crystallisation in the DSC at a cooling rate of 25 0C/min are very similar to those of the first crystallisation of these pan blended samples in Figure 4-5, as would be expected and as was found in a similar situation in Chapter 3 with polyamide-4,6/carbazole. The only deviation, apart from slightly lower crystallisation temperatures (also found in Chapter 3), is the small, broad peak between the carbazole and polyamide-6 peaks of 38PA6Car. That peak indicates a small amount of complexation with carbazole crystallising in the inter-lamellar space. 4.3.2 Ampoule Material 4.3.2.1 Accuracy of Melting & Crystallisation Enthalpy Measurements The total enthalpy of crystallisation from the melt should be the same as the enthalpy of melting when taking the sample back into the melt. The total areas of crystallisation peaks during cooling were plotted against the total area of melt peaks in Figure 4-8,

along with the line for a one-to-one

correspondence. It can be seen that there is generally a good correlation, as occurred also for the polyamide-4,6 blends with carbazole in Chapter 3. 155

120 PA6Car 1:1 line

∆ Hf Second melt (J/g)

100 80 60 40 20 0 0

20

40

60

80

100

120

∆Hf First crystallisation (J/g) Figure 4-8 Comparison of the total heat of crystallisation (J/g) for the first crystallisation of polyamide-6/carbazole blends to the heat of fusion (J/g) for the second melts.

This forms a good verification that the integrals are accurately measured because the crystallisation curve shapes are quite different from the melting curves and often have complexities in form. 4.3.2.2 Melting Temperatures (First melt in DSC) from heating at 5 0C/min of material blended earlier in ampoules The thermograms in Figure 4-9 show the melting profiles in the first DSC heating ramp for polyamide-6, carbazole and their blends that had previously been formed in ampoules. They should be close to those of the second melting of the pan-blended materials. There will be differences due to the pan-blended samples having previously been crystallised at 25 0C/min rather than the 2 0C/min used to crystallise materials in ampoules. a) The mild double peak for the polyamide-6 sample shows that some material had crystallised in the metastable form within the ampoule. The thermogram form is typical of the cooling rate used. b) The 89PA6Car sample has peaks at 195 and 222 0C.

These are

interpreted as a eutectic melt followed by the dissolution of the polyamide into the eutectic liquid giving a TLS peak in the manner described in Chapter 1.

156

Figure 4-9 DSC thermograms of polyamide-6/carbazole ampoule samples for the first melt in the DSC at 5 0C/min.

c) The 75PA6Car thermogram has only one peak at 193 0C, slightly lower than that of the lower peaks for 89PA6Car and 68PA6Car. That peak is from carbazole that has co-crystallised with polyamide-6 in the ampoule. At this polyamide level there is no evidence of crystalline pure polyamide-6 or pure carbazole domains in the sample. With a single rather than a double peak, this means that the polyamide-6 must either be in a stable form or in the metastable form but with no opportunity to melt and recrystallise to the stable form. The answer to this is not clear because the single eutectic peaks here lie between those of the metastable and the stable melting temperatures in the remelting of the fast cooled pan blended samples. It is also to be noted that the main peak here is at a slightly lower temperature than the eutectic temperatures of the other concentrations. That occurs with melting when the concentration is close to the eutectic concentration for a number of the other polyamide/diluent combinations.

157

d) The 68PA6Car sample is similar to that of Figure 4-6 50PA6Car with a peak temperature of 197 0C except that there is only a hint of a double melt for the main peak. This infers that the polyamide-6 is mainly in the stable form. The endotherm at higher temperatures than the main peak is weak showing that there is little excess carbazole to dissolve once the solution has become saturated near 200 0C. e) The thermogram for 54PA6Car is very similar to that of 50PA6Car in Figure 4-6 with the difference that the main peak is a single peak rather than a double one. It is also similar to the 68PA6Car in Figure 4-9 except that it has a strong endotherm above the main peak for the dissolution of excess carbazole in a TLS peak. The explanation is the same except that the polyamide-6 appears to be in the more stable form. That is probably due to the differing crystallisation rates in the prior thermal history. f) The same comments apply to a comparison between 25PA6Car and the 38PA6Car in Figure 4-6 with the difference that the lower polyamide level in the sample means there is a smaller carbazole melting peak here and a larger TLS peak. Summarising these results for the higher levels of polyamide-6, there is evidence

for

carbazole

having

previously

crystallised

together

with

polyamide-6 in the inter-lamellar spaces at close to 75% polyamide by weight. Excess polyamide or carbazole has crystallised within the remaining amorphous solid. The polyamide-6 within the co-crystallised part is most likely in the metastable form but does not have the opportunity to transform into the stable form with increasing temperatures due to its intimate physical involvement with the carbazole. The polyamide-6 would have naturally tended to crystallise mainly in the stable form with the slow cooling rates used. 4.3.2.3 Overall Crystallinity The percentages of polyamide from TGA were used with the total enthalpy of the first melting heating ramp to calculate the overall crystallinity of ampoule samples in the same manner as in Chapter 3.

The results are

plotted below in Figure 4-10. of the overall crystallinity percentage against the percentage polyamide. 158


100 90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 4-10 Overall crystallinity versus weight percentage polyamide determined from the total first DSC melting enthalpy for ampoule blended polyamide-6/carbazole samples and TGA plateau levels at 300 0C.

It can be seen that, on average, there is an overall decrease in crystallinity with decreasing carbazole in the samples.

The results show a consistent

decrease unlike the situation in Chapter 3 with polyamide-4,6 and carbazole where results were more scattered. The decrease is, however, not a linear decrease between the endpoints and shows that the blending process has influenced the morphology to give a lesser level of crystallinity for the blends.

4.3.2.4 Crystallisation Temperatures at 2 0C/min cooling for ampoule samples crystallising for the first time in the DSC. Figure 4-11 shows the thermograms of the crystallisation of material that had been melt-blended in ampoules, taken to the melt in DSC and then is being crystallised at 2 0C/min in the DSC. a) There is a double peak for the crystallisation of polyamide-6. This is also visible to a lesser degree in the pan blended sample of Section 4.3.1 and has occurred regularly in the course of this project. Varun Ratta, in his thesis on semicrystalline polyimides, [158 p. 92] refers to this being generally explained by secondary crystallisation processes. b) The thermogram of 89PA6Car is very similar to that of polyamide-6. Taking the sample to 265 0C and only cooling at the slow rate of 2 0C/min has allowed some of the small amount of carbazole in the sample to 159

evaporate. This has raised the relative level of polyamide in the sample closer to pure polyamide-6.

Figure 4-11 DSC thermograms of the first ampoule material crystallisation in the DSC at 2 0C/min.

c) The 75PA6Car sample shows compatibility between the two materials in solution until nearly 60 0C below the crystallisation temperature of carbazole and 25 0C below that of polyamide-6.

The sharp vertical

leading side of the peak and the double peak point to the materials starting to crystallise separately.

The initial part of this peak deviates

from the normal start to a carbazole crystallisation peak. It means that the two materials, or perhaps only polyamide-6, have just started to crystallise when local variations in this strongly supercooled solution have lead to the carbazole crystallising partly independently. The change in compositional balance has likely prompted the polyamide-6 to crystallise at that later point, the temperature being very much lower than in the polyamide-6 crystallisation case and for the other depressed crystallisations in the other thermograms. d) The 68PA6Car sample has a ‘spiky’ carbazole crystallisation peak depressed to nearly 40 0C below the normal carbazole crystallisation 160

temperature. Carbazole is crystallising out in domains within the polyamide-6/carbazole solution. We have the depressed crystallisation of polyamide-6, presumably with some carbazole, at a temperature 15 0C below the normal crystallisation temperature for polyamide-6. The materials have crystallised out separately. e) The 54PA6Car sample has the carbazole crystallising at a depressed temperature but not as low as the 75PA6Car case. The crystallisation is not quite as “spiky” as normal for carbazole and is tending towards the shape found in Figure 4-7. That implies more inclusion of polyamide in the crystallising material. The polyamide-6 crystallises at a temperature similar to that observed with the 68PA6Car but with a carbazole spike near the peak. f) The 25PA6Car thermogram, with its high level of carbazole, crystallises much of that out at very close to the normal carbazole crystallisation temperature. The polyamide-6 crystallises at the same temperature as seen with 68PA6Car and 64PA6Car. The main feature of this series of thermograms is the high compatibility of the materials in the melt for 75PA6Car at temperatures down to 175 0C where a eutectic solidification takes place.

Polyamide-6/carbazole will

crystallise at approximately 177 0C over a wide range of compositions once excess carbazole has crystallised. There is a large scatter in the temperature where the excess carbazole crystallises but no reason for that is evident. No phase separation of the liquid appears to take place with the crystallisation in the DSC at 2 0C/min of blends previously formed in ampoules. There is spiky crystallisation here for the first DSC crystallisation of ampoule material, just as we had seen with the polyamide-4,6/carbazole combination in Chapter 4.

4.3.2.5 Crystallinity from first crystallisation in the DSC Figure 4-12 below plots the crystallinity of the phases that are close to pure carbazole in the first crystallisation cooling and the crystallising peaks that are not carbazole domains. It was mentioned previously that there is no mistaking the crystallisation of phases that are basically pure carbazole (see Chapter 1 Figure 1-20 ). It is quite clear from Figure 4-12 that the crystallinity of the carbazole falls off in an almost linear manner to zero with 161

increasing polyamide content above 75%. Carbazole in these samples with high polyamide levels is being incorporated closely in the polyamide structure without the ability to crystallise. This is happening for a material that has a very high propensity to crystallise, normally being above 90% crystalline

for

the

basic

raw

material.

Van

der

Heijden

studied

1-dodecanol/atactic poly(styrene) phase separation for membrane formation using temperature induced phase separation (TIPS) in his thesis [7]. In his case he is working with an atactic polymer that cannot crystallise so the situation here with a crystallisable polymer adds a level of complexity. He shows in his Chapter 5 that there is an observed linear fall of in enthalpy of crystallisation against polymer content to zero at the Berghmans Point. He argues that this is associated with the diluent causing a glass transition depression but not crystallising in the (amorphous) polymer. In the study here we also have the polymer melting below the melting temperature of the diluent unlike the scattered graph of polyamide-4,6/carbazole.


100 carbazole polyamide-6

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 4-12 Crystallinity of carbazole and non-carbazole parts versus the percentage polyamide determined from the first crystallisation in the DSC at 2 0C/min of polyamide-6/carbazole material formed originally in ampoules.

The crystallinity of the non-carbazole domains appears to be slightly higher than that of pure polyamide solidified in an ampoule. That may be due to the carbazole being incorporated between the crystalline polyamide lamellae.

162

4.3.2.6 Phase Diagrams for first time heating and cooling ampoule material in DSC

260 250 Liquid

240 Temperature ( 0 C)

230 220 210 200

TmPA6Pure TmPA6Depr TmCarDepr TmEut TcPA6Pure TcPA6Depr TcCarDepr TcEut

190 180 170 160 150


Liquid &



Solid Liquid Solid Liquid & crystallites Solid

140 0

10

20


80

90

100

Figure 4-13 Non-equilibrium phase diagrams for heating and cooling polyamide-6, carbazole, and their blends made in ampoules.

The eutectic phase diagram seen in Figure 4-13 differs strongly from the phase diagrams seen with polyamide-4,6/carbazole blends in Chapter 3. They are more similar to those of Cha et al’s work [8] and that of van der Heijden [7] although both have non-crystallising diluents and van der Heijden’s deals with non-crystallising polymer, looking at Kelley-Bueche depression [50] of Tg at high polymer concentrations. There is depression here

of

the

eutectic

temperature

at

the

eutectic

point.

No

melting/crystallisation at eutectic temperatures can be seen to the right of the eutectic point. The points near 54% wt polyamide should actually be between 68 and 75 % wt of polyamide due to substantial loss of diluent during the thermal cycling in the DSC.

As expected, heating and cooling

phase diagrams differ mainly in a vertical displacement by 10 to 20 0C.

4.3.2.7 Third Melting of materials from ampoules/Second DSC Melt at 5 0C/min The samples in pans from the first DSC runs were passed through a repeat melt/crystallisation cycle in the DSC described in this and the following subsection. This is the third time, overall, that the materials are undergoing a melt/crystallisation, cycle, having been through one originally in the ampoule. 163

It should be recognised that evaporation of some of the carbazole has taken place as with the second DSC run of samples melt blended in DSC pans. This is more the case with DSC runs having the slow 2 0C/min cooling rate than with the faster cooling used here with pan blending because the sample remains at elevated temperatures for much longer. Figure 4-14 below shows the DSC thermograms of the melt portions of the repeat DSC runs.

Figure 4-14 DSC thermograms of the second melting at 5 0C/min in the DSC of polyamide-6/carbazole ampoule material.

The thermograms are very similar to those of the first melt in the DSC, as expected. The differences are that the very broad peak of 54PA6Car has disappeared and that the extent of the 68PA6Car broad peak has reduced. Both of these are due to loss of some of the excess carbazole.

4.3.2.8 Third Crystallisation of Materials/Second DSC Crystallisation Thermograms in Figure 4-15 of the second crystallisation in the DSC are also generally similar to the first time. The exceptions are: a) 54PA6Car

where

carbazole

does

not

crystallise

as

previously

at

approximately 187 0C but begins in earnest mid way through the 164

(depressed) polyamide-6 crystallisation. There is also an extremely small “spiky” carbazole crystallisation near 173 0C. This crystallisation must be a residual concentration of carbazole that delays crystallisation until that temperature due to hindrance from the polyamide. The mass of that carbazole crystallising can be calculated from the area of the spike to be a domain of 18 µg.

Figure 4-15 DSC thermograms of the second crystallisation at 2 0C/min in a DSC of polyamide-6/carbazole ampoule material.

b) A very minor peak for 25PA6Car at close to 200 0C between those of the (depressed) carbazole and the (depressed) polyamide-6. This is possibly due to a small amount of polyamide-6 phase separating.

4.4 Fourier Transform Infra Red Spectroscopy 4.4.1 Mid Infra-Red - Photoacoustic The FTIR spectra of polyamide-6 and carbazole from ampoule samples were “added/combined” mathematically in different proportions as described in Chapter 1 to simulate the expected spectra from the combinations of materials. The relevant peaks can be seen in Figure 4-16.

165

The region near 3304 cm-1 is the N-H stretch of the polyamide. This is assumed to be “bound” (hydrogen bonded state) as hydrogen bonds in polyamide are almost always fully consummated in the crystalline region and very high in the amorphous state. The area just above that is for the N-H stretch of carbazole. Peak for PA-6 where Carbazole Confirmatory has low absorbance Carbazole peak

80 - - - polyamide-6 carbazole 70

--

Photoacoustic

60 50 40 30 20 Carbazole peak where PA6 absorbance is low

10 3500

3000

2500 2000 1500 Wavenumbers (cm-1)

1000

Figure 4-16 FTIR peaks used for modelling polyamide-6/carbazole ampoule blend spectra with the polyamide-6 and carbazole spectra

Spectral Addition 130 68PA6Car

Photoacoustic

120 110 100 90 80 70 3400

3350 Wavenumbers ( cm-1)

3300

Figure 4-17 Measured and modelled FTIR spectra for ampoule sample 68PA6Car in the N-H regions between 3450 and 3250 cm-1.

166

500

The rounding of the model compared with the actual measured values is due to saturation of the photoacoustic signals for these major peaks in the constituent materials whereas the blends had absorbing centres physically more diluted in the samples. The carbazole N-H stretch (“free” N-H) in the model from the raw materials should be near 3420 cm-1 in Figure 4-17. In fact, the peak measured with the 68PA6Car sample

material was within 2 cm-1 and to the left.

The

measured peak should have moved by 15 to 30 cm-1 towards the right because it was now “bound” if there was hydrogen bonding occurring between the carbazole and the polyamide-6. Similarly, the actual N-H stretch due to the polyamide-6 is within 5 cm-1 to the right of the model. Hydrogen bonding in this case should have led to a noticeable shift to the left because there were more “free” N-H bonds than normally experienced. This situation with all of the polyamide-6/carbazole blended ampoule samples is identical to that found with blends formed of polyamide-4,6 and carbazole in Chapter 3.

The arguments here are the same as those in

Chapter 3 and lead us to believe that there is no hydrogen bond interaction between the polyamide-6 and carbazole. Spectra of the N-H stretches of blends compared with the models based on the photoacoustic spectra of the polyamide-6 and of carbazole can be seen in Appendix D on CD demonstrating that a simple weighted summation of the spectra gives virtually identical results to the measured spectra of the blend. The Near Infra-Red studies below in section 4.4.2 were carried out to confirm this situation. Figure 4-18 shows an example of a comparison between the actual FTIR spectrum in the complex 1700 to 1100 cm-1 region for material from ampoule sample 54PA6Car

and the result of a mathematically scaled

combination of the constituent material FTIR spectra.

There are no clear

shifts of peaks to the left or right evident but there may be some indication of differences in the relative heights of the peaks, as explained in Chapter 3. This lack of significant shifts, as distinct from changes in relative heights, has been found with each spectrum from ampoule material with polyamide-6 and carbazole where the spectrum has individually been modelled. 167

The

partial saturation of the peaks from the raw materials used to create the model will cause some differences in the “fingerprint” region between model and blend. 70

Spectral Addition 54PA6Car

65 60

Photoacoustic

55 50 45 40 35 30 25 20 1600

1400

1200 1000 Wavenumbers (cm-1)

800

600

Figure 4-18 Measured and modelled FTIR spectra in the 1700 cm-1 to 400 cm-1 range for the 54PA6Car ampoule sample.

The inclusion of carbazole in the polyamide-6 inter-lamellar space or inter-spherulitic regions, alterations to crystallinity levels and changes to crystallographic form or morphology of the polyamide could all cause some changes to the “fingerprint” region of the mid range IR spectrum. It is not necessary to infer hydrogen bond interactions to explain the differences seen in this region of the spectrum.

4.4.2 Near Infra-Red - DRIFT The spectra of the constituents and the sample being investigated were examined in the ranges 6920 to 5700 cm-1 and 5000 to 4200 cm-1, as with the polyamide-4,6/carbazole blended ampoule samples in Chapter 3. Examples of this are shown in Figure 4-19 and Figure 4-20 which display the DRIFT spectra in the NIR ranges of interest for ampoule sample 54PA6Car and the ampoule materials from the constituent materials.

As

with the polyamide-4,6/carbazole blends described in Chapter 3, there was no evidence in any of the cases that shifts occur in the Near Infra-Red region. It also provides confirmation to the Mid Range Infra-Red findings of no hydrogen bond interactions between the materials blended in ampoules. 168

carbazole 0.180 polyamide-6 54PA6Car 0.17 0.16

Absorbance

0.15 0.14 0.13 0.12 0.11 0.10 0.09 7000

6800

6600


6000

5800

Figure 4-19 Near Infra-Red DRIFT spectra of ampoule sample 54PA6Car and constituent materials in the hydrogen bonding region 7200 to 5600 cm-1.

It can clearly be seen in the figures that the spectrum peaks of the 54PA6Car ampoule material have come directly from the peaks of the constituent materials without any shifts. Spectra can be seen in Appendix D on CD.

0.27

carbazole polyamide-6 54PA6Car

0.26

Absorbance

0.25 0.24 0.23 0.22 0.21 0.20 5000

4800


4400

Figure 4-20 Near Infra-Red DRIFT measurements on ampoule sample 54PA6Car and constituent materials in the hydrogen bonding range 5000 to 4200 cm-1.

169

4.5 Small Angle X-ray Scattering 160 57PA6Car 44PA6Car carbazole

140

Scattering

120

polyamide-4,6 62PA46Car 31PA46Car

100 80 60 40 20 0 0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

2Θ

Figure 4-21 SAXS spectra for two polyamide-6/carbazole ampoule samples, PA46, Carbazole, & material from typical PA46:Car ampoule samples (with background removed and graphs smoothed).

SAXS measurements were carried out to learn about the lamellar structure encountered with polyamide-6/carbazole blends.

Figure 4-21

plots the

smoothed results (after removal of background scattering) of intensity versus scattering angle 2θ for some of the samples run. Those displayed in the figure do not include polyamide–6 because that particular sample was not available from an ampoule crystallisation process at the time ampoule samples were sent to Connecticut.

The curves for polyamide-4,6 and a

typical polyamide-4,6/carbazole material are also displayed for comparison. It can be seen from the curves that: a) The two polyamide-6/carbazole curves have major peaks that are to the left

of

those

for

polyamide-4,6

and

for

the

typical

polyamide-4,6/carbazole material. This means that the spacing is greater for the polyamide-6 blends than the polyamide-4,6 and its blends. b) The main peaks for the Bragg spacings vary between 13.3 and 13.8 nm for the polyamide-6/carbazole samples. c) The peak near 2θ = 0.38 is reasonably constant at close to 23 nm for all samples. It is difficult to distinguish between the samples because of the limited measurement resolution at small values of 2θ and because the peaks sit on the steep downward slope from 2θ = 0. 170

The conclusions that can be drawn from this work are obviously very limited due to the absence of a polyamide-6 sample from an ampoule being measured with SAXS.

4.6

Summary

The polyamide-4,6/carbazole case discussed in Chapter 3 was strongly influenced by the higher-melting polyamide-4,6 being dissolved by the lowermelting carbazole as the carbazole started to melt. The work described here on melt blending polyamide-6 with carbazole has been characterised by the same principle but with the higher-melting carbazole being dissolved by the lower-melting polyamide-6. In both the polyamide-4,6 and polyamide-6 situations, there has been a limit to the amount of the higher-melting material that can be dissolved in the lower one. In this case, the limit appears to be 50% carbazole in the mix before there is an excess of carbazole. In both cases, the higher-melting material requires much higher temperatures to achieve dissolution in the saturated solution. Almost the normal melting temperature of the highermelting material is required to dissolve all of that material if there is a great deal of excess. The melting/dissolution process is shortened when there is only a small amount of excess. The melting and crystallisation DSC results for both pan-blended and ampoule-blended samples are consistent with a eutectic phase diagram having a eutectic temperature some 25 0C below the polyamide-6 melting temperature.

The experience in Chapter 3 of the

eutectic temperature being reduced by 2 – 4 0C for the eutectic composition is repeated here and will be seen in some of the other chapters.

That

reduction in temperature for the eutectic composition is also found for crystallisation.

The eutectic composition appears to be in the vicinity of

70-75% polyamide-6. There has also been more visibility in this chapter of the ability of polyamide-6 to crystallise in a metastable form even if cooled slowly. That metastable form generally displays itself with a double melting peak as thin lamellae melt and recrystallise to the more stable form during the melting process. The different melting temperatures of the double melt peak have allowed some understanding of the polyamide-6 crystalline form when single or double peaks have been encountered. The difference in previous thermal 171

history between samples melt blended in pans and those crystallised much more slowly in ampoules was used to display the ramifications on the lamellar form in the blends. This has obviously been complicated to a degree by melting temperature depressions due to carbazole being incorporated at an intimate level within the lamellar space. These melting temperature depressions have generally been in the range 20-30 0C below that of the (lower-melting) polyamide-6 and 40-50 0C below that of the carbazole. The remelting of previously crystallised samples in pans has led to cleaner melts due to the more intimate mixing of the molecules. This is the same as was found with polyamide-4,6/carbazole powder blends formed in DSC pans. The crystallisation from the melt tends to result in separate crystallisation of the carbazole and polyamide-6 when the concentration of polyamide-6 is lower than or equal to that of the carbazole. Both of these crystallisations are generally at depressed temperatures although the carbazole crystallisation appears to be at very variable temperatures. The phase separation is not complete because with this combination of materials we find the carbazole does not fully have its characteristic “spiky” crystallisation. The two materials almost crystallise completely together in the same narrow peak at even lower temperatures in the slowly cooled ampoule when there is approximately 75% polyamide-6.

This is at 20 0C below the normal

polyamide-6 crystallisation temperature.

No separate crystallisation of

carbazole has taken place until this point, a full 60 0C below the normal carbazole crystallisation temperature.

This case appears to lead to the

polyamide-6 crystallising in the metastable form but is unable to change into the stable form as it is reheated. Overall crystallinity of ampoule samples as monitored during their first foray into the DSC is shown to decrease with increasing polyamide level from the very high levels found with just carbazole. Analysis of the subsequent crystallisation in the DSC shows that separate carbazole crystallinity decreases in an almost linear fashion to near zero at

172

(very) approximately 65% polyamide. On the other hand there is a peak in the non-carbazole crystallinity over the broad range 50-75% polyamide. It can be concluded from the FTIR measurements in the Mid Range Infra-Red and in the Near Infra-Red that there is no hydrogen bond interaction between the carbazole and the polyamide-6 in the experiments undertaken with melt blending.

These results are the same as those for

polyamide-4,6/carbazole blends. The photoacoustic results in the 1700 to 400 cm-1 region do not give rise to any definitive conclusions because the measurements are not likely to be linearly additive.

That is due to photoacoustic signal saturation in the

spectra of the constituent materials. Unfortunately the absence of an ampoule sample of polyamide-6 measured with SAXS means the only conclusion that can be drawn is that lamellae in polyamide-6/carbazole blends are thicker than those in polyamide-4,6 and polyamide-4,6/carbazole blends. That conclusion is not surprising because of the number of carbon atoms involved on average per amide group. Some other general measurement and sample issues arose in the course of these experiments. No problems were encountered with moisture in ampoule samples during the TGA measurements as determined from the simple check that the weight loss by 150 0C be less than 1.7%. This gross quality control check had been done on all ampoule samples of all carbazole or phenothiazine blends of polyamide using the TGA results. It was found that TGA gave concentrations of polyamide within 5% of each other when repeat runs were made on similarly coloured samples from near each other in an ampoule. This is similar to the Chapter 3 experience.

173

Chapter 5


Introduction

174

5.2


175

5.3


176

5.3.1

Pan Melt Blending

5.3.1.1 5.3.1.2 5.3.1.3 5.3.1.4

5.3.2

Ampoule Material

5.3.2.1 5.3.2.2 5.3.2.3 5.3.2.4 5.3.2.5 5.3.2.6 5.3.2.7

176

0

Melting Temperatures for first heating ramp of the powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min 0 Melting Peak Temperatures for second heating ramp at 5 C/min for pan blended samples. 0 Crystallisation Peak Temperatures for second cooling ramp at 25 C/min.

176 178 179 180

181

0

Melting Temperatures (First melt in DSC) at 5 C/min for ampoule material. Overall Crystallinity 0 DSC Crystallisation Temperatures at 2 C/min for remelted ampoule material. Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC 0 Third Melting of ampoule materials/Second DSC Melt at 5 C/min. 0 Third Crystallisation of Materials/Second DSC Crystallisation at 2 C/min.

181 183 184 186 187 188 189

5.4


190

5.5

Summary

190

5.1 Introduction This chapter extends the work done on polyamide-4,6/carbazole and polyamide-6/carbazole

melt

blends

with

further

work

on

polyamide-6,9/carbazole melt blends. Polyamide-6,9 differs from polyamide-4,6 in that both the diamine and diacid moieties of the repeat unit are longer and also it is an “even-odd” polyamide-m,n rather than an “even-even” polyamide.

It differs from

polyamide-6 in that it is a polyamide-m,n rather than a polyamide-n type. These factors influence the ways in which the polyamide can crystallise. They also affect the flexibility of the polymer chains by having a lower density of amide groups and affect other properties such as the melting temperature. 174

Some of the themes seen in the earlier chapters will be shown to recur here. The situation is more like that of the polyamide-6/carbazole blends in Chapter 4 because the polyamide-6,9, like the polyamide-6, melts below the carbazole melting temperature.

The polyamide-6,9 also has a stronger

tendency than polyamide-6 to crystallise in the high temperature stable form. The melt is a double melt as the material melts/recrystallises and melts again [48 p. 46].

5.2 Thermogravimetric Analysis Thermogravimetric Analysis (TGA) was carried out in order to determine the


percentage of carbazole in ampoule samples. 100 PA69Car 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 5-1 Actual versus expected weight percentage polyamide in polyamide-6,9/carbazole samples from ampoules

TGA

reproducibility

had

been

found

in

the

polyamide-4,6/

and

polyamide-6/carbazole blends to be within 5% from sample to sample. This was taken to be the expected reproducibility for these trials also. Figure 5-1 below shows a comparison between the expected percentage of polyamide (based on the weights blended in ampoules) and the actual concentrations of polyamide in different samples taken from the ampoule material. Some of the samples are different by more than 10% from the expected concentrations.

They are well above the maximum 5% variation expected 175

based upon the work covered on Chapters 3 and 4. It indicates that they are real variations in material composition encountered within ampoule samples due to uneven mixing or the development of specific compositions.

5.3 Differential Scanning Calorimetry Polyamide-6,9 has a double melting endotherm with a weak endotherm in the range 150-175 0C combined with an exotherm at 185 0C

and a main

endothermic peak in the range 208-211 0C. The endotherm/exotherm pair is due to the polyamide having crystallised preferentially into a metastable lamellar state. The metastable lamellae melt near 173 0C and recrystallise near 185 0C into the stable form that melts at 211 0C as the sample is heated further.

The exotherm between the two endotherms is very

noticeable in the case of polyamide-6,9. There is a very small thermal activity at temperatures just above the main peak. The reason for that has not been determined.

The polyamide crystallises near 193 0C for a cooling rate of

2 0C/min and near 175 0C for a cooling rate of 25 0C/min. 5.3.1 5.3.1.1

Pan Melt Blending Melting Temperatures for first heating ramp of the powders at 5 0C/min

The DSC thermograms of the melting at 5 0C/min of polyamide-6,9, carbazole and the melt blending of some mixtures of polyamide-6,9 and carbazole powders are shown Figure 5-2. The individual thermograms will be described in more detail below: a) The 63PA69Car thermogram has a very weak, broad endotherm over the range 180-200 0C after a faint endotherm near 160 0C. It is most likely that this behaviour is initially the dissolution of some carbazole powder into the amorphous portion of polyamide grains that it is in contact with. There is a broad endotherm of the melt recrystallising at a temperature slightly lower than normally for polyamide-6,9 followed by a very broad and weak melting peak that finishes near the end of the normal polyamide-6,9 main melting peak.

There is no separate melting

endotherm for the carbazole in the powder mix indicating that all the carbazole has been taken up in the polyamide-6,9 between 160 and 210 0C. The very broad peaks are typical of those seen in Chapters 3 and 4 for the first melts of powder mixes in the DSC pans and, similarly, are at slightly lower temperatures than for the polyamide. 176

Figure 5-2 DSC thermograms during the first melting at 5 0C/min of polyamide-6,9/carbazole and powder mixtures.

b) The 38PA69Car thermogram begins with the melting of metastable lamellae near 150 0C and is followed by the depressed melting of polyamide-6,9.

Some of the available carbazole surrounding the

polyamide has apparently had the opportunity by this stage to absorb/dissolve into the polyamide or the polyamide melting temperature would not have been depressed by nearly 30 0C into a eutectic melt. There is a TLS peak above that peaking near 230 0C for the consumption of excess carbazole.

A minor peak exists above that at the carbazole

melting temperature.

This infers that by 235 0C the solubility of the

carbazole has been restricted. A possible binodal near that part of the phase diagram or kinetic effects have come into play slowing the last dissolution of carbazole into the saturated solution.

The pattern seen

here is broadly similar overall to the 38PA6Car case in Chapter 4 and, with reverse roles of the materials, to pan-blended 83PA46Car of Chapter 3. The lower melting material dissolves as much of the higher melting material as possible until there is a saturated eutectic solution and more can only be dissolved with increasing temperature giving a TLS peak. 177

5.3.1.2


Figure 5-3 DSC thermograms for the first crystallisation of pan blended polyamide-6,9/carbazole cooled from the melt at 25 0C/min.

A set of thermograms for the first crystallisation of powders melt-blended in pans is seen in Figure 5-3. They are shown with thermograms for molten raw materials subjected to the same 25 0C/min cooling ramps. a) The 63PA69Car thermogram has no separate peaks for carbazole crystallisation and only a double peak more than 20 0C below the normal crystallisation temperature of polyamide-6,9.

This means that the

carbazole has not phase-separated out within the high temperature solution into carbazole domains that could crystallise near the normal temperature for carbazole. There has also been no phase separation at lower temperatures for the polyamide-6,9.

The main peak is a double

peak so there is a eutectic crystallisation of polyamide-6,9/carbazole after the crystallisation of a small amount of similar material with a different composition. b) The 38PA69Car thermogram has the crystallisation of a large amount of carbazole-rich

material

near

the 178

normal

carbazole

crystallisation

temperature.

It also has a eutectic crystallisation peak for polyamide-

6,9/carbazole. The results seen here are consistent with those for polyamide-6/carbazole blends in Chapter 4. 5.3.1.3

Melting Peak Temperatures for second heating ramp at 5 0C/min for pan blended samples.

Figure 5-4 DSC thermogram at 5 0C/min of the second melting of polyamide-6,9/carbazole materials previously melt blended and crystallised in DSC pans.

The thermograms in Figure 5-4 differ slightly from the first melting of the powders in Section 5.3.1.1. This is similar to Chapters 3 and 4 where there were differences because the second melting ramp is starting from a more intimate molecular mix of the materials. a) Polyamide-6,9 displays a different thermogram for the second melting with sharp melting/re-crystallisation peaks for the metastable lamellae. This happens because of the fast cooling rate used on this sample compared with the unknown prior thermal history. b) The 63PA69Car thermogram has changed from the first time melting in that it displays sharper melting and re-crystallisation curves than 179

previously. Crystallisation of polyamide-6,9 had taken place at a fast rate into the metastable lamellae preferred under those conditions.

The

temperatures involved are 1 – 3 0C lower now, giving 30 0C depressions to both the melt/re-crystallisation and the main melting. There is some fine detail in the thermogram at the higher end of the main melting peak that is a TLS peak for polyamide-6,9 or for carbazole but it is not possible to say from that evidence which material is in excess. c) The 38PA69Car thermogram is also a refinement of the previous melting. There is a eutectic melt at 180 0C followed by a TLS peak for the excess carbazole.

The overall form is similar to the repeat melting ramp of

38PA6Car and 63PA46Car (reversed melting relationship between the polyamide and carbazole. The remelting of the previously melt blended pan samples shows similar changes to those from the ill-defined first melts of polyamide-4,6/ and polyamide-6/carbazole

blends to the sharper melting profiles when the

materials were passed through a second heating/cooling cycle. Reductions in melting temperatures with eutectic melting are also seen, as is the dissolution only at high temperatures of excess of the higher melting material into the saturated solution. 5.3.1.4

Crystallisation Peak Temperatures for second cooling ramp at 25 0C/min.

The thermograms in Figure 5-5 are very similar to those of the first crystallisation of pan blended samples in Figure 5-3, as expected, and as found in a similar situation in Chapters 3 and 4 with polyamide-4,6 or polyamide-6 and carbazole.

The only change was the slightly lower

crystallisation temperatures (also found in Chapters 3 and 4).

180

Figure 5-5 DSC thermograms of the second crystallisation of polyamide-6,9/carbazole pan blended samples at 25 0C/min.

5.3.2 5.3.2.1

Ampoule Material Melting Temperatures (First melt in DSC) at 5 0C/min for ampoule material.

The thermograms in Figure 5-6 show the melting profiles in the first DSC heating ramp for polyamide-6,9, carbazole and their blends that had previously been formed in ampoules. They should approximate those of the second melting of the pan blended materials. There will be differences due to the 2 0C/min used to crystallise materials in ampoules rather than the pan blended samples having previously been crystallised at 25 0C/min. a) The polyamide-6,9 sample from the ampoule differs from the second melt of polyamide-6,9 powder in that the double peak is not so sharp in the lead up to the main peak. This is due to the differing thermal histories of the two samples. The sample from the pan blending work had previously been cooled at 25 0C/min whereas this sample had been cooled at 2 0C/min in the ampoule. The slower prior cooling here has given more time for the sample to crystallise in a more favourable manner. situation is analogous to the polyamide-6 case. 181

The

Figure 5-6 DSC thermograms of polyamide-6,9/carbazole ampoule samples during the first DSC melting at 5 0C/min.

b) The 64PA69Car sample shows a similar thermal behaviour to the equivalent situation with 68PA6Car in Chapter 4. There are differences with the second melting of the pan blended 63PA69Car sample in the main peak not being a double peak due to the different thermal histories of the two samples (vide infra). The main peak has also been depressed by 30 0C compared to the polyamide-6,9 sample.

This is close to the

depression observed in the pan blended samples. There is another minor difference in that there is now a minor TLS peak for carbazole and a very minor melt peak just under the normal carbazole melting temperature. This last small peak shows that there is a limitation in dissolving the last carbazole into the solution. This could be either due to kinetic effects or to the high temperature part of the phase diagram including a binodal. c) The 61PA69Car thermogram differs little from 64PA69Car described just above.

Again, there is a depression of 30 0C compared to the

polyamide-6,9 melting temperature. The only difference is the absence of the very small peak just under the carbazole melting temperature. Apparently there are subtle differences between the two samples in the 182

way the carbazole is distributed in the sample. Little difference would be expected because of the small difference in polyamide levels between the samples. d) The 28PA69Car sample shows the typical behaviour seen numerous times previously with carbazole dissolving in the melting polyamide up to a saturated level followed by rapidly increasing dissolution as the temperature is increased. In this case, 72% of the material is carbazole. The end of the second peak is virtually at the carbazole melting temperature.

The peak size is large because of the amount of excess

material requiring that high temperature to melt. The 30 0C depression of polyamide-6,9 melt/dissolution temperature is close to that observed in the two samples immediately above and those in the second melting of pan blended samples. The picture seen for the second melting of polyamide-6,9/carbazole incorporates a number of aspects seen in other trials described earlier in this and in other chapters. There are the more clearly defined melting peaks, the eutectic and TLS peaks, the reduced melting/re-crystallisation of metastable lamellae because of the previous slow crystallisation and some difficulties in melting all of the higher melting material. 5.3.2.2


The percentages of polyamide were used with the total enthalpy of the first melting heating ramp to calculate the overall crystallinity of ampoule samples in the same manner as in Chapters 3 and 4. The results are plotted below in Figure 5-7. It can be seen that, on average, there is an overall decrease in crystallinity with decreasing carbazole in the samples. The few results show a monotonic decrease, the same as in Chapter 4 with polyamide-6/carbazole but unlike the situation in Chapter 3 with polyamide-4,6/carbazole where results were more scattered.

The locus of the measured points lies below the linear

relationship between the crystallinity of the pure materials and so shows that the blending process has led to some overall suppression of crystallinity of the blends.

183


100 90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 5-7 Overall percentage crystallinity of polyamide-6,9,/carbazole versus percentage polyamide determined from total first DSC melting enthalpy and TGA plateau level at 300 0C.

Thus the Gibbs free energy of mixing is non-zero. It should be noted here that the density of carbazole is close to that of a weighted average of amorphous and crystalline polyamide-6,9 and so the linear relationship on a molar volume basis will be within a few percent for a weight percentage basis.

The density of phenothiazine is over 20% higher so a linear

relationship on a molar volume basis will require a slight curve downwards between 0 and 100% on a weight basis to represent a colligative relationship. 5.3.2.3

DSC Crystallisation Temperatures at 2 0C/min for remelted ampoule material.

Figure 5-8 shows the thermograms of the crystallisation of material meltblended in ampoules, taken to the melt in DSC and then crystallised at 2 0C/min. a) The polyamide-6,9 crystallisation curve at 2 0C/min cooling rate can be seen to be relatively sharp compared to the second crystallisation of the polyamide-6,9 powder sample and at a temperature 18 0C above that for the 25 0C/min cooling rate.

The narrowness is seen here on a

temperature axis rather than against time. The time taken for (90% of) the crystallisation was more than four times greater because of the slower cooling rate, allowing a more favourable development of lamellar 184

thickness. The increase in crystallisation temperature is typical of polymers crystallising at slower rates and requiring less temperature supercooling to crystallise.

Figure 5-8 DSC thermograms of the first crystallisation of polyamide-6,9/carbazole ampoule material at 2 0C/min in the DSC.

b) The 64PA69Car and 61PA69Car thermograms are very similar to each other in having two peaks, a “spiky” carbazole peak depressed by 40-50 0C compared to carbazole followed by a polyamide-6,9 peak depressed

25 0C

below

the

normal

polyamide-6,9

peak.

The

supersaturated solution has had excess carbazole crystallise first. The same situation is expected to have occurred previously in the ampoule samples although temperature control would have been poorer because of the size of the sample (1.5 g.) and the physical cooling conditions in the furnace. Some of the carbazole remains in solution and crystallises with the polyamide at lower temperatures.

The ampoule sample 75PA6Car

of Chapter 4, in a similar situation, had a total absence of carbazole crystallisation down to 60 0C below the normal carbazole crystallisation temperature compared to the 68PA6Car having a depression of 35 0C. The samples described in this section lie in concentrations between those 185

two from Chapter 4 and have carbazole crystallisation depressions broadly consistent with the polyamide-6/carbazole blends.

There is

another very minor difference between the two in that the 61PA69Car sample has a small peak for the crystallisation of a minuscule amount of pure polyamide just prior to the carbazole “spike”. c) The crystallisation of 28PA69Car takes place in a similar manner to the 25PA6Car ampoule sample of Chapter 4. Carbazole crystallises from the highly

loaded

solution

at

just

under

the

normal

crystallisation

temperature for carbazole followed by the polyamide depressed to the same temperature as for the 64PA69Car and 61PA69Car samples. In this section we have seen similar behaviour to that in the first crystallisation in the DSC of polyamide-6/carbazole blends in that there is a linear relationship between enthalpy of crystallisation and percentage polyamide.

Again, the polymer/diluent system has the polyamide melting

temperature lower than that of the diluent. The point where the enthalpy has fallen to zero is close to 90% polyamide. 5.3.2.4

Crystallinity from first crystallisation in the DSC



90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 5-9 Crystallinity of Carbazole and Non-Carbazole parts from crystallisation from a polyamide-6,9/carbazole melt at 2 0C/min in the DSC.

Figure 5-9 plots the crystallinity of the phase domains that are virtually pure carbazole in the first crystallisation cooling and the crystallising peaks that 186

are not mainly carbazole domains. It was mentioned previously that there is no mistaking the crystallisation of phases that are almost pure carbazole (see Chapter 1, Fig. 1-20).

It is can be seen from Figure 5-9 that the

crystallinity of the carbazole falls off towards zero with increasing polyamide content. This linear fall off without scatter is occurring in the same manner as that of polyamide-6/carbazole.

The relationship appears linear but

unfortunately there are few points available to make a stronger statement on this one set of data. The carbazole in these samples with high polyamide levels is being incorporated in the inter-lamellar or inter-spherulitic regions without being able to crystallise. The crystallinity of the non-carbazole phase appears to be higher than that of the pure polyamide from the ampoule. That may be due to the mass of carbazole incorporated in the crystalline polyamide part. The results are different from those of polyamide-6/carbazole in Chapter 4.in that the level of carbazole crystallinity near 60% polyamide has not dropped to as low a value. The carbazole still has a strong inclination to phase separate and crystallise out of solution at these moderately high polyamide levels. This is consistent with the previous melting behaviour in Figure 5-6 with the same samples. There, the higher TLS peak had been seen of carbazole dissolving only at elevated temperatures. This is even more evident in Figure 5-11 below. That saturation at lower levels of carbazole is the reason the excess is crystallising here. 5.3.2.5

Phase Diagrams for first time heating and cooling ampoule material in DSC

Figure 5-10 is very similar to Figure 4-13 for the phase diagrams with polyamide-6/carbazole blends in that it is a eutectic style.

Heating and

cooling non-equilibrium phase diagrams differ mainly in a 10 to 15 0C vertical displacement.

187

260 250 240

Solid & liquid

Liquid

Liquid

Temperature (0C)

230 220 210 200 Solid & liquid

190 180


170 160 150 140

Solid Liquid

Solid Solid & liquid Solid

130 0

10

20


80

90

100

Figure 5-10 Non-equilibrium phase diagrams for polyamide-6,9, carbazole and their blends showing eutectic-like behaviour

5.3.2.6

Third Melting of ampoule materials/Second DSC Melt at 5 0C/min.

The ampoule samples in pans from the first DSC runs were passed through a repeat melt/crystallisation cycle in the DSC similar to chapters 3 and 4.

Figure 5-11 DSC thermograms of the second melt in the DSC at 5 0C/min of polyamide-6,9/carbazole ampoule material.

188

Figure 5-11 shows the DSC thermograms of the melt portions of the repeat DSC runs. The thermograms are very similar to those of the first melt in the DSC, as would be expected. a) The minor differences are that the very broad TLS peaks of 64PA69Car and 61PA69Car have become slightly sharper with the repeat run. That is due uneven cooling in the ampoule having caused a less than ideal crystallisation that was detected in the DSC melting run afterwards. The cooling in the DSC pan was under very tightly controlled conditions and was for a small mass of material, much easier to keep all at (nearly) the same temperature. It is not obviously clear just from these thermograms whether the polyamide or the carbazole is in excess. b) The size of the carbazole melt/dissolution in 28PA69Car at high temperatures has reduced somewhat due to carbazole evaporation as only 45% of the original amount of was left after the second heating/cooling cycle. 5.3.2.7

Third Crystallisation of Materials/Second DSC Crystallisation at 2 0C/min.

Figure 5-12 DSC thermograms of the second crystallisation at 2 0C/min in a DSC of polyamide-6,9/carbazole ampoule material.

189

Thermograms in Figure 5-12 of the second crystallisation in the DSC are also generally similar to the first time. The exceptions are: a) Evidence in 61PA69Car and perhaps in 28PA69Car for a minimal amount of nearly pure polyamide-6,9 crystallising at 190 0C. This can be seen for 28PA69Car in Appendix A. b) A reduction in the carbazole crystallisation temperature of 3-4 0C, probably due to loss of carbazole.

5.4 Fourier Transform Infra Red Spectroscopy Photoacoustic FTIR measurements were carried out in the Mid Infra-Red and DRIFT FTIR in the Near Infra-Red that, in all cases, resulted no evidence for hydrogen bond interactions being found between polyamide-6,9 and carbazole in polyamide-6,9/carbazole blend samples. This null result is the same as found in earlier chapters on other polyamide/carbazole blends. Detailed spectra are provided in Appendix D on CD.

5.5 Summary As with the polyamide-6/carbazole blends, there are reasonable similarities with melt blending polyamide-4,6 with carbazole once the roles of the materials are interchanged. Again, this is because polyamide-4,6 melts at temperatures above carbazole rather than the other way around with polyamide-6/ and polyamide-6,9/carbazole blends. The remelting of previously crystallised samples in pans from powders has led to sharper melts due to the more intimate mixing of the molecules in agreement with the other blends. The effect on the polyamide-6,9 melting temperature is a depression of approximately 30 0C. There is approximately 25 0C depression of the polyamide-6,9 crystallisation during cooling. The depression of carbazole crystallisation in the experiments was seen to be 40-50 0C. Polyamide-6,9 has a greater propensity than polyamide-6 to crystallise in the metastable form rather than the stable form.

We have seen here similar

effects to polyamide-6 in the polyamide crystallising in different forms depending upon previous thermal history.

190

Another similarity with both polyamide-4,6/ and polyamide-6/carbazole blends is the lower melting material starting to melt at reduced temperatures in a eutectic melt and dissolving the higher melting material to a certain limit.

At that stage any excess of the higher melting material requires

substantial increases in temperature to dissolve all the remainder. We also find here that the overall crystallinity of blends from ampoules decreases with increasing polyamide level. The crystallinity of the carbazole decreases rapidly with increasing polyamide level. There is a maximum in non-carbazole crystallinity between approximately 50 and 80% polyamide. Here again we have the experience found with polyamide-4,6/ and polyamide-6/carbazole blends that there is no evidence for hydrogen bond interactions taking place between polyamide-6,9 and carbazole when they are melt blended together in ampoules. The interactions are driven by the (limited) compatibility of the two materials and their ability to coexist sterically at a molecular level rather than due to any hydrogen bonding. There are differences, however, compared to carbazole blends with the other two polyamides. One difference is the carbazole not crystallising out at 63% polyamide for fast cooling but crystallising (at the same concentration range) for slow cooling. At that concentration the solution becomes saturated earlier during remelting requiring higher temperatures to dissolve the excess carbazole. Another difference was the observation of two peaks for dissolution of excess carbazole above the melting of polyamide for both pan blended and ampoule sample monitoring in the DSC. The crystallinity of carbazole in recrystallising ampoule samples does not drop to zero as quickly as for the other two polyamides with higher densities of amide groups. The reasons for these differences in blend solubility/melting/crystallinity properties lie in this polyamide having longer carbon chains in the repeat unit and/or in this polyamide being an m,n type of polyamide.

191

Chapter 6


Introduction

192

6.2


193

6.3


194

6.3.1

Pan Melt Blending

6.3.1.1 6.3.1.2 6.3.1.3 6.3.1.4

6.3.2

0

Melting Temperatures for first heating ramp of the powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min 0 Melting Peak Temperatures for second heating ramp at 5 C/min 0 Crystallisation Peak Temperatures for second cooling ramp at 25 C/min

Ampoule Material

6.3.2.1 6.3.2.2 6.3.2.3 6.3.2.4 6.3.2.5 6.3.2.6 6.3.2.7

0

Melting Temperatures (First melt in DSC) at 5 C/min of ampoule material Overall Crystallinity 0 DSC Crystallisation Temperatures at 2 C/min for remelted ampoule material. Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC Third Melting of materials/Second DSC Melt 0 Third Crystallisation of Materials/Second DSC Crystallisation at 2 C/min.

194 194 195 197 198

200 200 201 202 204 205 206 207

6.4


208

6.5

Summary

208

6.1 Introduction This chapter extends the work done on melt blends of polyamide-4,6, polyamide-6 and polyamide-6,9 with carbazole with further work on polyamide-6,12/carbazole melt blends. The polyamide-6,12 differs from the polyamide-4,6 in that the diamine and/or diacid moieties of the repeat unit are both longer than for the other polyamides.

It is an “even-even” polyamide-m,n like polyamide-4,6.

It

differs from polyamide-6 in that it is a polyamide-m,n rather than a polyamide-n type and from polyamide-6,9 which is an “even-odd” polyamide. These factors influence the ways in which the polyamide can crystallise and its melting temperature. They also affect the flexibility of the polymer chains by having a lower density of amide groups and affect other properties such 192

as the melting temperature which is lower than polyamide-4,6 and polyamide-6 but higher than polyamide-6,9. Some of the themes seen in the earlier chapters will be shown to recur here. The

situation

is

most

like

that

of

the

polyamide-6/

and

polyamide-6,9/carbazole blends because the polyamide-6,12, like the other two, melts below the carbazole melting temperature. The melt is usually a double melt as the material melts/recrystallises and melts again [48 p. 46]. This combination of materials will be shown, like the polyamide-6 and polyamide-6,9 combinations with carbazole, to lead to a linear reduction in carbazole enthalpy of crystallisation with increasing polyamide content. In this case the concentration for zero carbazole crystallinity will be shown to be near 70% by weight of the polyamide.

6.2 Thermogravimetric Analysis Thermogravimetric Analysis (TGA) was carried out in order to determine the percentage of carbazole in ampoule samples.

Figure 6-1 shows a

comparison between the theoretical percentage of polyamide and the actual polyamide concentrations in samples taken from ampoule material.


100 PA612Car 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 6-1 Actual versus expected weight percentage polyamide in polyamide-6,12/carbazole samples from ampoules.

Deviations of more than 5% from the expected concentrations of polyamide, as seen in Figure 6-1, are real differences in concentration for the samples 193

compared with the expected. These are variations due to either imperfect mixing in the ampoules or to the way the samples have crystallised in the ampoules.

6.3 Differential Scanning Calorimetry Polyamide-6,12 has a weak endotherm in the range 170-190 0C combined with an exotherm at 201 0C and then a main endothermic peak at 216 0C The endotherm/exotherm pair is due to the polyamide having crystallised in a metastable form. The metastable form melts above 160 0C when the sample is heated and recrystallises at 201 0C into the stable form that melts at 216 0C. The exotherm between the two endotherms is very conspicuous for this polyamide. The polyamide crystallises at 197 0C for a cooling rate of 2 0C/min and at 185 0C for a cooling rate of 25 0C/min. 6.3.1 6.3.1.1


The individual thermograms for polyamide-6,12, carbazole and two blends shown in Figure 6-2 are described in more detail below:

Figure 6-2 DSC thermograms from the first melting at 5 0C/min of polyamide-6,12 carbazole and powder mixtures.

194

a) The sample 63PA612Car was formed when polyamide-6,12 began to melt as it dissolved carbazole at the same time. The main endothermic peak is just under 190 0C and extends to 205 0C. There is some indeterminate thermal activity above this temperature, with residual carbazole being dissolved at higher temperatures in the saturated solution. b) The thermogram for 43PA612Car begins with a similar endotherm to 63PA612Car but changes into a TLS peak for carbazole and has a substantial additional peak near the carbazole melting temperature. The last peak is showing that the solution is either entering a high temperature region where a single phase is unfavourable or there are strong kinetic effects slowing down the dissolution of carbazole into the solution. c) 38PA612Car encountered experimental difficulties with the data at the higher end of the ramp and has been cut short.

It also had quite a

noticeable loss of carbazole through evaporation. It has, however, been included because it provides supplementary information to the other thermograms and the data from crystallisation and the repeat cycle were good. The thermogram follows a similar path to the previous two samples except that there is more carbazole to dissolve above the main dissolution/melt. The data where the thermogram had to be terminated was clearly taking a path akin to that of the 43PA612Car sample that had a final melting of residual carbazole only at the carbazole melting temperature. The initial melting of the metastable form of polyamide-6,12 is coupled with the dissolution of carbazole before the polyamide has the chance to recrystallise properly into the metastable form. The relatively large peak for the

melting

of

carbazole

at

the

carbazole

melting

temperature

for

43PA612Car and indications of similar behaviour for 38PA612Car show that there is either a miscibility problem at high temperatures and carbazole concentrations or the kinetics of the dissolution are quite slow. 6.3.1.2


A set of thermograms is seen in Figure 6-3 for the first crystallisation at 25 0C/min with powders melt-blended in pans. 195

Figure 6-3 DSC thermograms for first crystallisation of pan blended polyamide-6,12/carbazole from the melt at 25 0C/min.

a) The cooling of the 63PA612Car sample is showing no evidence of separate crystallisation of nearly pure carbazole domains within the liquid but there is crystallisation of, what appears to be, a compound/blend of polyamide-6,12 and carbazole at temperatures well above the normal crystallising temperature of the polyamide.

This is followed by the

crystallising of polyamide-6,12 with substantial amounts of carbazole depressed more than 20 0C below the normal crystallising temperature of polyamide-6,12, in the manner seen with other polyamides. b) The thermogram for 43PA612Car is similar to the above but with the first peak at a higher temperature. c) The cooling of 38PA612Car begins with a small crystallisation of some carbazole followed by the crystallising of a polyamide-6,12/carbazole blend at higher temperatures than seen above with the 63PA612Car and 43PA612Car samples.

The cooling process then results in the

crystallising of the remaining polyamide-6,12/carbazole in the sample at an identical temperature and peak shape to the other two blends. There is some phase separation occurring here with the first peak at virtually 196

the same temperature as for pure carbazole. The crystallisation coming from phase separation substantiates the comments made in the previous section on the first melting. The cooling part of the cycle for the first heat/cool process of polyamide-6,12 and carbazole powders shows behaviour typical of a eutectic phase diagram as will be shown later in the chapter.

There is a large peak depressed a

constant amount from polyamide-6,12 crystallisation and above that there is a peak which decreases in peak temperature as the amount of carbazole is reduced. The eutectic composition must lie above 63% polyamide because the upper peak for 63PA612Car is above the polyamide-6,12 crystallisation temperature.

Additionally, at high temperatures and high carbazole

concentrations, there is phase separation taking place. 6.3.1.3

Melting Peak Temperatures for second heating ramp at 5 0C/min

Figure 6-4 DSC thermograms of second melt at 5 0C/min for materials previously crystallised in pans for polyamide-6,12/carbazole.

The thermograms in Figure 6-4 for the second melting of pan blended samples at 5 0C/min are slightly different from the first powder melting. This is the same `as in Chapters 3 to 5 where there were differences because the melt is now starting from a closer molecular mixing of the materials. 197

a) The

remelting

of

polyamide-6,12,

known

to

have

crystallised

at

25 0C/min, displays the typical slow melting of the metastable form, the sharp exothermic re-crystallisation at 203 0C into the stable lamellar form and the main melting of that stable form of the polyamide at 216 0C. b) The 63PA612Car sample, on remelting, shows the melting of the carbazole/polyamide depressed by 27 0C. That melting is displaying only a little evidence of a metastable-to-stable conversion. The lamellae are either in the stable form or are hindered by the carbazole from making the conversion. The main melting peak is followed by the carbazole TLS peak. c) The thermogram for 43PA612Car is similar to the 63PA612Car sample except that the TLS peak is at higher temperatures. d) The melting of the previously formed 38PA612Car sample is almost identical for the main peak to that of the 63PA612Car and 43PA612Car samples above. A TLS peak for the remaining carbazole in the saturated solution can be seen.

The general fall-off in the thermogram as the

temperature raises above the main peak is due to significant carbazole evaporation from the sample. This sample only had 58% of the original carbazole level after the two heating/cooling cycles applied to it as a postDSC weighing found.

Another set of samples near this concentration

could possibly have been run on TGA and DSC with another attempt to get good sealing of the hermetic DSC pans. The common experience of repeat melting of pan blended powders has given rise again to sharper melting curves. The main melting peaks are depressed in eutectic melts from the normal polyamide-6,12 melting temperature and they do not have significant conversion of metastable to stable form. 6.3.1.4

Crystallisation Peak Temperatures for second cooling ramp at 25 0C/min

Figure 6-5 presents the DSC thermograms of the second crystallisation at 25 0C/min of materials originally melt blended in the DSC in from the constituent material powders. a) There has been a change in the crystallisation of the 63PA612Car sample from the first time it was crystallised from the melt.

The higher

temperature peak is now showing more characteristics of a sharper 198

crystallisation, although it does not yet have the “spiky” characteristic of a near-pure carbazole crystallisation. b) The thermogram for 43PA612Car is similar to 63PA612Car except that the upper peak is at a higher temperature.

Figure 6-5 DSC thermograms of second crystallisation of pan blended polyamide-6,12/carbazole samples at 25 0C/min.

c) The

38PA612Car sample

has approximately

the

same

stages

in

crystallisation as previously, viz. a small spiky form crystallisation, a crystallisation

slightly

below

that

and

a

depressed

polyamide

crystallisation which is at the same temperature as with the 63PA612Car sample. The main difference is that the first peak has become smaller, less spiky and dropped from 228 0C to 220 0C. In addition, the peak at 217 0C has become slightly sharper. The move to lower temperatures for the spiky peak and reduction in size are due to the evaporation of carbazole noted for the thermogram of the second melting. The picture for the second crystallisation is much the same as for the first crystallisation with eutectic formation, although modified by evaporation of carbazole for the sample with higher carbazole levels. 199

The evidence for

polyamide-6,12 with carbazole in the inter-lamellar space is repeated, but with a slightly sharper crystallisation peak. 6.3.2 6.3.2.1

Ampoule Material Melting Temperatures (First melt in DSC) at 5 0C/min of ampoule material

The thermograms in Figure 6-6 show the melting profiles in the first DSC heating ramp for ampoule samples of polyamide-6,12, carbazole and their blends. They should approximate the second melting of the pan blended materials. There will be differences due to different prior cooling rates.

Figure 6-6 DSC thermograms of polyamide-6,12/carbazole ampoule samples first melting ramp at 5 0C/min in the DSC

a) The polyamide-6,12 sample has a main melting peak at 217 0C for the stable form with only a faint shoulder for the conversion of metastable lamellae

to

the

stable

form.

The

lack

of

a

significant

endotherm/exotherm pair prior to the main melting is due to the slow prior crystallisation in the ampoule. b) 70PA612Car, 65PA612Car and 52PA612Car all have single melting peaks very nearly 30 0C below the polyamide-6,12 ampoule sample One aspect that distinguishes the three is the lead-in to the melting peak at 180 0C with there being a slight pre-melting for the 52PA612Car sample and 200

even less of an effect on the other two. The other is the more rounded higher temperature side of the main peaks for 65PA612Car and 52PA612Car. These two appear to have slight vestiges of TLS peaks. The only indication that it is for an excess of carbazole rather than polyamide is that the 70PA612Car has none, meaning that 70% polyamide is close to the eutectic concentration. c) The 39PA612Car sample has a main melting peak at almost the same temperature as for the other three blends but preceded by a slightly stronger deviation in the lead-in to the main melting peak.

All the

samples are showing similar behaviour to the start of the polyamide-6,12 ampoule sample. The region above the main melting peak is the same TLS peak form as other polyamide/carbazole blend combinations at high carbazole levels where excess carbazole is only dissolved up in the saturated solution at high temperatures. These samples from ampoules show that the high temperature solution of carbazole in polyamide-6,12 is saturated near 70% polyamide at the depressed eutectic melting temperature near 190 0C.

This provides

confirmation of the results of the pan blended samples that this combination of materials differs noticeably from that of polyamide-6,9/carbazole blends in particular. In all cases we have seen single main melting peaks with the polyamide-6,12 materials from ampoules where the slower cooling

has

resulted in a different form from the fast cooled samples made in pans. 6.3.2.2


The percentages of polyamide were used with the total enthalpy of the first melting heating ramp to calculate the overall crystallinity of ampoule samples in the same manner as in Chapters 3 and 4. The results are plotted below in Figure 6-7. It can be seen that, on average, there is an overall decrease in crystallinity with decreasing carbazole in the samples. The few results show a consistent decrease, the same as with polyamide-6/ and polyamide-6,9/carbazole but unlike the situation with polyamide-4,6/carbazole where results were more scattered. The lack of a linear relationship in crystallinity with concentration of polymer shows that the crystallinity is being depressed by the blending showing a non-zero free energy of mixing. 201


100 90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 6-7 Overall crystallinity against weight percentage polyamide determined from TGA and total first DSC melting enthalpy of ampoule samples for polyamide-6,12/carbazole.

6.3.2.3


Figure 6-8 shows the thermograms of the crystallisation of material meltblended in ampoules, taken to the melt in DSC and then crystallised at 2 0C/min. a) The 70PA612Car sample has a single crystallisation peak depressed by 25 0C below the normal polyamide-6,12 crystallisation peak. That peak does appear to have a slight, almost-vertical, section just before the top of the peak as the temperature is lowered.

The peak rises after a slow

beginning to the crystallisation. That implies the polyamide is slightly in excess under those conditions. The solution is almost 65 0C below the carbazole crystallisation temperature at that stage. Until that point, the third of the sample that is carbazole is being hindered from crystallising in the normal way by the presence of the polyamide.

The polyamide

would normally be increasing in viscosity due to the lower temperature but the solvent (carbazole) is lowering the viscosity. At that point, the supercooling of both materials that is the driving force for crystallisation has taken over, initiating the beginning of the polyamide crystallisation. That has altered the balance and the carbazole has then begun to rapidly crystallise. This again has altered the balance in this thermodynamically 202

unstable solution and the rest of the crystallisable polyamide has crystallised. That curve can be seen in expanded form in Appendix A.

Figure 6-8 DSC thermograms of the first crystallisation at 2 0C/min in the DSC of polyamide-6,12/carbazole ampoule material.

b) The 65PA612Car has a near-vertical section for the single crystallisation peak, implying that the carbazole is driving the crystallisation (at almost exactly 60 0C below the normal carbazole crystallisation temperature). This is almost reaching the stage of 75PA6Car in Chapter 4 where there was a more pronounced break between the end of the carbazole crystallising and the rest of the peak. It is, however, at a noticeably lower polyamide level in the blend. c) The 52PA612Car thermogram with a double peak 55 0C below the carbazole crystallisation temperature is showing more separation of the peaks with the carbazole being seen to start crystallising first and being overtaken by polyamide-6,12/carbazole crystallising at slightly lower temperatures after the tip of the spike. d) The 39PA612Car is at such a high carbazole level that it can only sustain the whole sample in liquid form until 25 0C below the normal carbazole crystallisation temperature. At that point, a large amount of carbazole 203

crystallises out quickly leaving a saturated solution at that temperature. The depressed polyamide-6,12/carbazole solution finally crystallises at exactly the same temperature as in the samples above. All samples show themselves as a consistent series in their behaviour. We lave a single peak at 70% polyamide-6,12. We see the carbazole unable to remain stably in solution at such depressed temperatures, triggering earlier and earlier crystallisation of carbazole domains within the sample as the level of carbazole increases from sample to sample. There is a tendency to crystallise the carbazole earlier as the level of carbazole increases until by 39PA612Car the crystallisation of the two phases is separated by 40 0C. The similar forms of melting thermograms over a wide concentration range for the first three belie the subtleties more visible here.

The lowest

crystallisation (double) peak is for the sample with 70% polyamide but the first material to start crystallising is polyamide.

In the 65% sample with

polyamide it is the carbazole that begins to crystallise. It is clear that the optimal concentration is in this narrow range. 6.3.2.4




90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 6-9 Crystallinity of carbazole and non-carbazole parts versus percentage polyamide, obtained from polyamide-6,12/carbazole crystallisation from the melt in the DSC.

Figure 6-9 below plots the crystallinity of the phase domains that are virtually pure carbazole in the first crystallisation cooling in the DSC. The 204

crystallising peaks that are polyamide-rich domains are also plotted in the figure. It is quite clear from Figure 6-9 that the crystallinity of the carbazole falls off in an almost linear fashion with increasing polyamide content, reaching zero near 75% polyamide. That is in the region where the lowest crystallisation peak was seen and where the melting samples had a eutectic concentration. The carbazole in these samples with high polyamide levels is being incorporated closely in the solid without having the ability to crystallise separately.

This would be with individual carbazole molecules

between polyamide-6,12 chains in the general amorphous regions or in the amorphous interlamellar space. The polyamides being studied here are semicrystalline which gives further complexity to the polymer/diluent situation because the diluent can solidify in the amorphous phase, causing depression of the glass transition temperature, or lie in the interlamellar space. We see in the figure below that the crystallinity of the diluent is reducing in a nearly linear fashion to zero at just the concentration identified above where both materials crystallise at the same time. That is a similar situation to the one found by van der Heijden in his work, although he was dealing with a diluent/amorphous polymer system rather than a semicrystalline one. The crystallinity of the non-carbazole phase is very much higher than that of pure polyamide from the ampoule.

The results are similar to those of

polyamide-6/ and polyamide-6,9/carbazole blends in earlier chapters. 6.3.2.5


Figure 6-10 shows similar non-equilibrium eutectic phase diagrams to those of Chapters 4 and 5.

Here again the crystallisation temperature of the

diluent is higher than that of the polyamide.

The depressed melting and

crystallisation peaks for 52PA612Car should actually be to the right of their positions in the figure as there was significant evaporation of the diluent during the DSC runs.

205

260 250 240


0

Temperature ( C)

230 220

Liquid

Solid & liquid

210 200

Solid Liquid

190


180 170 160 150

Solid Solid & liquid Solid

140 0

10

20


80

Figure 6-10 Non-equilibrium phase diagrams for polyamide-6,12, carbazole and their blends showing eutectic behaviour during heating and cooling.

6.3.2.6

Third Melting of materials/Second DSC Melt

Figure 6-11 DSC thermograms of the second melt at 5 0C/min in the DSC of polyamide-6,12/carbazole ampoule material.

206

90

100

The ampoule samples in pans from the first DSC runs were passed through a repeat melt/crystallisation cycle in the DSC similar to chapters 3 to 5. Figure 6-11 below shows the DSC thermograms of the melt portions of the repeat DSC runs. There is virtually no difference between the first and second melts in the DSC

of

these

polyamide-6,12/ccarbazole

samples

apart

from

peak

temperatures differing by 1 – 2 0C and the second (TLS) peak for excess carbazole dissolution

with 39PA612Car being shortened due to carbazole

evaporation. 6.3.2.7

Third Crystallisation of Materials/Second DSC Crystallisation at 2 0C/min.

There are only small differences in Figure 6-12, compared with the first crystallisation in the DSC of these ampoule samples. a) There is no significant change with the 70PA612Car sample. b) The peak for 65PA612Car has dropped by 2 0C due to loss of carbazole. There are hints of insignificant thermal activity at 212 0C and 188 0C.

Figure 6-12 DSC thermograms of the second crystallisation in a DSC at 2 0C/min of polyamide-6,12/carbazole ampoule material.

52PA612Car has carbazole-rich crystallisation taking place 2 0C lower and the peak for the remainder is 1 0C lower, both due to carbazole 207

evaporation.

There is a small disturbance in the thermogram near

192 0C because of slight crystallisation of near pure polyamide-6,12. c) The spiky carbazole crystallisation has dropped by 4 0C for 39PA612Car but the second peak is unchanged, again consistent in the series with the loss of carbazole.

There is a smaller polyamide-6,12 crystallisation at

190 0C than with 52PA612Car. The small differences seen are associated with the evaporative loss of some carbazole during the extended period at elevated temperatures for the slow DSC cooling rate.

There is a very small amount of polyamide-6,12

crystallisation at the polyamide-6,12 crystallisation temperature due to phase separation for the two highest level carbazole samples.

6.4 Fourier Transform Infra Red Spectroscopy No hydrogen bond interactions were found between polyamide-6,12 and carbazole in polyamide-6,12/carbazole blend samples using Photoacoustic FTIR measurements in the Mid Infra-Red and DRIFT Infra-Red.

FTIR in the Near

This result is the same as found in earlier chapters on other

polyamide/carbazole blends. Detailed spectra are provided in Appendix D on CD.

6.5 Summary This polyamide is different from the other polyamides previously studied in that it is an even-even polyamide with long carbon chains between the amide groups. These points affect the way in which the polyamide can crystallise. They are affecting the opportunities for polyamide-6,12 and carbazole to coexist in liquid form at high temperature and for the polyamide-6,12 to dissolve the carbazole as the blends are heated. It was found in the pan blending that there was much less excess carbazole to dissolve at the higher carbazole level powder combinations and that the dissolution of excess only occurred at the highest level of carbazole.

The

carbazole is able to dissolve in melting polyamide-6,12 to a higher degree before the solution becomes saturated. That experience was also evident with the ampoule samples even though the slower prior cooling of the ampoule samples resulted in the polyamide-6,12 starting to dissolve from a stable rather than a metastable form. 208

The melting of powder mixes in pans at high levels of carbazole results in virtually all the carbazole being dissolved before the high temperature solution becomes saturated. The crystallisation of pan blended samples having high levels of carbazole leads to much less carbazole crystallising out in the initial crystallising stages. The carbazole that is in the high temperature solution is more easily able to remain in solution with fast cooling rates. Both of these are telling us that the solubility of carbazole in high temperature polyamide-6,12/carbazole solutions is noticeably higher on remelting than seen with the polyamide-6/ and polyamide-6,9/carbazole high temperature solutions from samples cooled at fast rates. The ampoule samples, originally cooled at 2 0C/min, have showed similar increased solubility of carbazole in the blend upon remelting in the DSC. We find differing crystallisation behaviour at the slower cooling rates compared to that of the other polyamides blended with carbazole in ampoules. The sample is far more inclined to remain completely as a liquid until

much

lower

temperature,

even

at

relatively

high

carbazole

concentrations near 50%. The solutions are able to drop up to 65 0C below the normal carbazole crystallisation temperature before there is any crystallisation at all.

This is approximately 25 0C below the normal

polyamide-6,12 crystallisation temperature at those cooling rates. The crystallinity of the carbazole drops linearly to zero with increasing polyamide content in this combination of materials, as for the other polyamides where the polyamide melting temperature is lower than that of the diluent. It does drop to zero here below 70% polyamide, lower than for the other two polyamides. The non-carbazole crystallinity does show the same increase in crystallinity in the range 50-80% polyamide that was seen with other polyamides but the level

of

crystallinity

has

increased

more

than

for

the

other

second

DSC

polyamide/carbazole combinations. The

small

differences

between

first

DSC

and

melt/crystallisation cycles for ampoule samples are due to loss of carbazole.

209

This is similar to the observations for carbazole combined with other polyamides in ampoules. A careful investigation with Mid Range and Near Infra-Red FTIR for interactions between polyamide-6,12 and carbazole in ampoule melt blended samples has delivered a similar result to the earlier investigations on polyamide-4,6, polyamide-6 or polyamide-6,9 blends with carbazole in that nothing could be found of any hydrogen bond interactions. The main conclusions that can be drawn from this set of trials is that the compatibility between polyamide-6,12 and carbazole is better than that of the

polyamide-6/

and

polyamide-6,9/carbazole

combinations.

The

behaviour of polyamide-6 in blends lies between that of polyamide-6,12/ and polyamide-6,9/carbazole blends.

The latter two both have longer repeat

units, showing that the even-even polyamide is much more compatible with carbazole than the even-odd polyamide.

210

Chapter 7

POLYAMIDE-4,6 WITH PHENOTHIAZINE CONTENTS 7.1

Introduction

211

7.2


212

7.3


213

7.3.1

Pan Melt Blending

7.3.1.1 7.3.1.2 7.3.1.3 7.3.1.4

7.3.2

Ampoule Material

7.3.2.1 7.3.2.2 7.3.2.3 7.3.2.4 7.3.2.5 7.3.2.6 7.3.2.7

214

0

Melting Temperatures for first heating ramp of the dry powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min. 0 Melting Peak Temperatures for second heating ramp at 5 C/min 0 Crystallisation Peak Temperatures for second cooling ramp at 25 C/min 0

214 216 218 220

221

Melting Temperatures (First melt in DSC) at 5 C/min of ampoule material Overall Crystallinity 0 DSC Crystallisation Temperatures at 2 C/min for remelted ampoule material. Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC Third Melting of materials/Second DSC Melt Third Crystallisation of Materials/Second DSC Crystallisation

221 223 224 227 228 229 230

7.4


231

7.5

Summary

231

7.1 Introduction We have looked at four polyamides combined with carbazole in the previous four chapters.

The first polyamide, polyamide-4,6, had a melting

temperature higher than the carbazole but the others had melting temperatures lower than the carbazole. In those chapters, we saw some quite large differences in the way the four polyamides melt blended with carbazole.

We also saw behaviour that was common between the

polyamides. Chapters 7 to 10 will allow us to look at how the four polyamides behave when melt blended with phenothiazine. This is a small aromatic molecule, similar in shape to carbazole, but it includes a bulky sulphur atom on the opposite side of the structure to the N-H group. This material has a melting temperature of 186 0C compared with carbazole at 246 0C. 211

All four

polyamides have melting temperatures above the temperature at which phenothiazine melts. We will see differences due to the lower temperature at which phenothiazine melts despite polyamide-4,6 having a higher melting temperature than both of the blending materials. Phenothiazine and carbazole are slightly differently shaped molecules with differing electron density distributions. The polyamide being investigated in this chapter, polyamide-4,6, is an even-even polyamide with the chain lengths of the 2N, 2*(N + 1) type. The diamine repeat sub-units are reasonably short and the diacid units are of medium length. These factors influence the way in which the polyamide can crystallise and give rise to the high polyamide-4,6 melting temperature just under 300 0C. This will be briefly alluded to in this chapter’s summary. The repeat unit characteristics will become an important factor in the discussions of the General Conclusions chapter where all the results for the different polyamides and materials blended with them are brought together.


7.2 Thermogravimetric Analysis 100 PA46PTh 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 7-1 Actual versus expected weight percentage polyamide in polyamide-4,6/phenothiazine blend samples from ampoules as determined by TGA plateau levels at 300 0C and compared with that expected from weights of materials used in the ampoules.

TGA was carried out on ampoule samples in order to determine the percentages of polyamide in the samples for crystallinity calculations. The 212

results are shown in Figure 7-1

in comparing the actual percentages of

polyamide to those expected from the weights used in the ampoules. Several of the points are much more than 5% away from the expected concentrations

of

polyamide.

The

large

differences

in

polyamide

concentration from the expected values in Figure 7-1, are real ones. These are variations due to imperfect mixing in the ampoules or to some molecular reorganisation during crystallisation in the ampoule. It is most likely that the variations are because a compromise was followed on the maximum temperature. On the one hand there was an aim to protect this polyamide from degradation leading to scission of some chains and also to further polymerisation resulting in increases in molecular weight of other chains at high temperatures. On the other hand the temperature had to be high enough that all polyamide lamellae were melted and the viscosity in the melt lowered sufficiently by elevated temperatures to ensure rapid diffusion of the two components of the liquid. The temperature used was either a little on the low side or phase separation took place resulting in less even distribution of the two components in various samples. We will see evidence in several parts of this chapter for phase separation taking place. Statistically there will be less predictable concentrations of polyamide in individual samples where phase separation has taken place, particularly if there has been substantial phase ripening.

7.3 Differential Scanning Calorimetry Polyamide-4,6 normally has a single melting peak at close to 295 0C with crystallisation peaks in the range 259-275 0C, depending on crystallisation conditions, as was mentioned in Chapter 3.

Phenothiazine melts close to

186 0C with crystallisation between 158 and 162 0C, depending on the cooling rate. The high polyamide and low diluent melting temperatures did cause some problems in that the phenothiazine was being taken far above its melting temperature in order to reach the maximum working temperatures of the trials. Very minor deviations from perfection in sealing the hermetic DSC pans rapidly led to significant phenothiazine losses due to evaporation. These losses were confirmed by the practice of weighing pans after trials to check for phenothiazine loss.

The combination of materials used for the 213

work in this chapter resulted in the biggest problems of evaporative loss encountered across all the work on the project and the problems for ampoule samples were the highest. 7.3.1 7.3.1.1

Pan Melt Blending Melting Temperatures for first heating ramp of the dry powders at 5 0C/min

Individual thermograms for polyamide-4,6, phenothiazine and three blends shown in Figure 7-2 are described in detail below for heating powder mixes of polyamide-4,6 and phenothiazine to the melt at 5 0C/min in DSC pans.

Figure 7-2 DSC thermograms during the first DSC melting at 5 0C/min of polyamide-4,6, phenothiazine powders and powder mixtures.

a) The 63PA46PTh thermogram first has a phenothiazine melting peak at the temperature normally found for phenothiazine, but then displays a broad, flat peak over the range 230-280 0C with the polyamide-4,6 dissolving.

The form of the thermogram is different from that

encountered with polyamide/carbazole blends. With those, we encountered dissolution of the higher melting material at temperatures lower than the normal melting temperature of the lower 214

material until saturation was reached.

Further dissolution only

accelerated with carbazole as the temperature increased. Eventually all of the excess higher melting material had been dissolved in a solution that was saturated at that temperature. Here, we have melting of the phenothiazine at exactly the normal temperature as if the polyamide was not there. It will be seen later in this chapter that polyamide-4,6 tends to phase separate at high phenothiazine concentrations having a very small solubility under those conditions.

There is only slight evidence of

crystalline polyamide dissolution in the 30 0C above the phenothiazine melting peak. The initial melting of phenothiazine is obviously at a high concentration

of

phenothiazine

relative

to

the

polyamide-4,6

concentration and it is occurring at temperatures well below the normal melting temperature of polyamide-4,6. It is not surprising that there is difficulty

in

dissolving

phenothiazine-rich solution.

the

polyamide-4,6

powder

into

the

The uneven, and slightly endothermic,

thermogram height from 180-270 0C may point to some interaction between the molten phenothiazine and the amorphous part of the polyamide.

At those temperatures the amorphous polyamide will be

moving from being a rubbery material towards being a liquid. It is only at nearly 40 0C above the phenothiazine melting temperature that there is significant indication of the polyamide beginning to dissolve. There is some evidence of phenothiazine evaporation at high temperatures with a slight fall-off in the signal above 280 0C. b) 50PA46PTh is very similar except that there is a sharp double peak at the melting temperature of phenothiazine. The reason for the double peak is not clear. There is a slight depression of the thermogram, similar to the previous sample, just above the phenothiazine melting and before the polyamide melts/dissolves in the range 230-270 0C. These multiple weak exotherms are indicative of changing solubility of the various phases as the temperature is increased dissolving the polyamide powder for the first time. The temperature is being ramped at 5 0C/min and it is likely that there are also kinetic effects in the dissolution of polyamide powder in phenothiazine at temperatures below the normal polyamide melting temperature. 215

c) The 38PA46PTh pan blended sample is similar except that some of the small amount of polyamide-4,6 is in the sample refuses to dissolve in the solution, undergoing a separate and sharp endotherm just under the normal polyamide-4,6 melting temperature. There is reduced solubility for polyamide-4,6 in the high temperature solution at high concentrations of molten phenothiazine.

A similar situation existed for polyamide-

4,6/carbazole blends with low levels of polyamide. The picture emerging from this set of thermograms is of a phenothiazine melt at the normal phenothiazine melting temperature followed by a broad, flat double peak covering the temperature range 230-280 0C. limited

solubility

of

polyamide

in

molten

It indicates

phenothiazine

at

high

temperatures. The additional variant to this is for the sample with highest level of phenothiazine. There, the smaller percentage of polyamide was only partly soluble in phenothiazine, requiring virtually the normal melting temperature of polyamide-4,6 to melt the polyamide lamellae in the sample. The

thermograms

display

the

reasonably

indeterminate

first

melts/dissolutions of the higher melting materials as encountered in earlier chapters. 7.3.1.2

Crystallisation for first cooling ramp of the molten blend at 25 0C/min.

The thermograms in Figure 7-3 show the DSC results of cooling the high temperature solutions made from melting powder mixtures at 25 0C/min and also show those of the raw materials under the same conditions. a) 63PA46PTh, the sample with the lowest level of phenothiazine, has a single crystallisation peak at a temperature depressed by 12 0C from the normal polyamide-4,6 crystallisation temperature after a faint peak at the polyamide crystallisation temperature.

There appears to be a small

amount of phase separation resulting in some nearly pure polyamide-4,6 crystallising before the main crystallisation of polyamide-4,6 together with some phenothiazine.

That main peak is the Flory-Huggins style

behaviour with depression of polyamide-4,6 crystallising with some phenothiazine in the interlamellar space. b) 50PA46PTh,

with

more

phenothiazine

in

the

melt,

has

three

crystallisation peaks, a peak for the crystallisation of phase separated, nearly pure polyamide-4,6 depressed by only 6 0C below the normal 216

polyamide-4,6 crystallisation temperature, an intermediate peak and the “spiky” peak for the crystallisation of excess phenothiazine slightly depressed from the normal

phenothiazine crystallisation temperature.

The

is

intermediate

peak

showing

some

formation

of

a

polyamide-4,6/phenothiazine compound with polyamide-4,6 lamellae having some phenothiazine crystallising in the inter-lamellar space but the majority of the crystallisation is by either very polyamide-4,6-rich or phenothiazine-rich material.

These peaks are showing both phase

separation and the partial compatibility of the two materials at high temperatures. There is some polyamide-4,6 affecting the crystallisation of the phenothiazine because the crystallisation temperature is depressed further than with the pure phenothiazine and the 38PA46PTh samples.

Figure 7-3 DSC thermograms for the first crystallisation of pan blended polyamide-4,6/phenothiazine, cooled from the melt at 25 0C/min.

c) The thermogram for 38PA46PTh is similar to that of 50PA46PTh except that

the

crystallisation

depressions

of

both

polyamide-4,6

and

phenothiazine are less than for the sample described above and the peak area of the middle peak is also reduced relative to the other two. The reduced depression of the polyamide-4,6 crystallisation is showing that 217

the solubility of polyamide-4,6 in the high temperature solution is reduced at high phenothiazine levels.

This is a substantiation of the

comments made on the first melting of the sample. depression

for

crystallisation

the of

phenothiazine

the

remaining

crystallisation phenothiazine

The lack of

peak is is

showing

virtually

pure

phenothiazine. The change in the relative sizes of the three peaks is showing less compatibility between the materials at high phenothiazine levels. The small size of the phenothiazine peak near 160 0C is partly due to the evaporative loss of phenothiazine at the high temperatures required to melt the polyamide-4,6. The main reason is suppression of phenothiazine crystallinity. These

thermograms

are

supporting the

notion

of

low

solubility of

polyamide-4,6 in phenothiazine at high temperatures, particularly when the level of phenothiazine in the solution is increased. They are also showing that at 50% phenothiazine and above it is possible to have polyamide-4,6 and phenothiazine crystallise at the same time.

The small area in total

under the peaks for those solutions with 50% phenothiazine and greater mean that much of the material becomes amorphous upon solidification under these conditions. 7.3.1.3


The thermograms in Figure 7-4 are for the remelting at 5 0C/min of materials from the melt/crystallisation cycles described above. a) Sample 63PA46PTh differs substantially from that of the first melting. This whole set of samples had been taken to approximately 120 0C above the melting temperature of phenothiazine. In doing this, the 63PA46PTh sample had lost approximately 50% of the phenothiazine in the first melt/crystallisation cycle. That value has been determined from weighing the sample after the DSC run and assumes, based on experience, that the weight of the polyamide component of the sample is virtually unchanged. The sample could perhaps better be named 82PA46PTh or similar but the original naming has been kept for clarity in following individual samples through

the

repeat

cycles.

This

sample

has

undergone

two

transformations between the two melt cycles. Firstly there has been the loss in phenothiazine just mentioned and secondly there has been a 218

better mixing of the two materials. A single, skewed peak can be seen that covers the melting of the crystalline polyamide-4,6/phenothiazine of the previous crystallisation integrated with the melting of polyamide-4,6. This differs strongly from the situation described in 7.3.1.1 where amounts of the powders had been put in the DSC pan with phenothiazine resting on top of the polyamide-4,6 grains and the resulting thermogram reflected the slow dissolution of those grains in the liquid.

Figure 7-4 DSC thermogram at 5 0C/min of second melt of materials previously crystallised in DSC pans for polyamide-4,6/phenothiazine. .

b) The 50PA46PTh sample has lost 55% of the phenothiazine by taking the sample through the first heating/crystallisation cycle to 310 0C and back to room temperature. Alternatively, it could have been called 67PA46PTh for this DSC run. The three peaks in this thermogram are the melting of phenothiazine, the melting of polyamide-4,6/phenothiazine compound and the melting of residual polyamide-4,6. This corresponds closely with the reverse of the crystallisation process described earlier in section 7.3.1.2.

That previous cooling of this sample had resulted in the

crystallisation

of

polyamide-4,6,

polyamide-4,6/phenothiazine

phenothiazine within the remaining amorphous material. 219

and

c)

The 38PA46PTh sample had lost well over half of the phenothiazine by evaporation

in

the

first

heating/cooling

cycle.

The

remaining

phenothiazine was incorporated in the sample in very much the same manner as the (reduced) 50PA46PTh sample described above. The repeat DSC runs of the three polyamide/phenothiazine samples here are dominated in their thermal responses by the prior loss of phenothiazine from the first heating/cooling cycle and the loss during this heating ramp. The sample with the highest polyamide content has changed to a (well mixed) slightly phenothiazine contaminated polyamide whilst the other two are providing confirmation, in reverse order, of the previous cooling results. Those

showed

the

crystallisation

polyamide/phenothiazine

compound

of

polyamide-4,6,

and,

lastly,

then

of

crystallisation

a of

phenothiazine. The three thermograms differed markedly from those for the melting/dissolution of the original powders in 7.3.1.1. 7.3.1.4


Figure 7-5 shows the thermograms of the pan blended samples during their second cycle, cooling from the melt at 25 0C/min.

Figure 7-5 DSC thermograms of the second DSC crystallisation of pan blended polyamide-4,6/phenothiazine samples at 25 0C/min.

220

a) The 63PA46PTh sample in re-crystallisation is showing the thermal response of further loss of phenothiazine, by now becoming virtually pure polyamide-4,6. b) The 50PA46PTh thermogram comprises the three peaks seen before for the crystallisation of polyamide-4,6, polyamide/phenothiazine compound and phenothiazine, very much at the previous temperatures observed. c) 38PA46PTh crystallises much as previously in the first DSC cycle except that the peak for polyamide-4,6 has increased at the expense of the peak for

the

polyamide/phenothiazine

compound

due

to

the

loss

of

phenothiazine by evaporation. The thermograms for the polyamide-4,6/phenothiazine samples have crystallised this time in the manner of the first crystallisation ramp except that curves have been modified because of the evaporative loss of phenothiazine. 7.3.2

Ampoule Material

It should be noted here that the ampoule sample 56PA46PTh was made at a time the ampoule heating ramps were still being refined. It is possible that the thermograms of the first melt in a DSC for this sample are not representative of the final process developed later. The effects of this should be zero after the first melt in the DSC. 7.3.2.1

Melting Temperatures (First melt in DSC) at 5 0C/min of ampoule material

The thermograms in Figure 7-6 show the melting profiles in the first DSC heating ramp for samples of polyamide-4,6, phenothiazine and their blends for materials previously crystallised in ampoules. They should be similar to those of the second melting of the pan-blended materials.

There will be

differences due to different prior cooling rates and the loss of phenothiazine in the first heating/cooling cycle for the pan blended samples. There had been no loss in the ampoule during crystallisation in the ampoule because it had been completely sealed during the process in the furnace. This series of samples are probably best treated together rather than individually because they form a reasonably clear progression in their thermograms.

221

Figure 7-6 DSC thermograms of polyamide-4,6/phenothiazine ampoule samples first melting in the DSC at 5 0C/min.

All samples (except for 56PA46PTh) have initial melting peaks 2 – 3 0C below that of the ampoule sample of phenothiazine.

That 56PA46PTh sample

which was an exception had been, as previously mentioned, crystallised in an ampoule at the time when the cooling regime in the furnace was still being refined.

Some of the past crystallisation history for that sample is

playing itself out in this subsequent melting ramp. It has, nonetheless, been included here as part of the overall series made at differing polyamide concentrations. There is a general trend for increasing size of the phenothiazine melting peak as a progression is made from the 75PA46PTh to the 21PA46PTh sample. Single or double peaks are found above the phenothiazine peak and the temperature of the peak (or peaks) decreases with decreasing polyamide content. The double peaks may result from some phase separation during the cooling in the ampoule.

The double peaks are occurring only for the

higher phenothiazine concentration samples, where more phase separation has been found in past chapters. They have been seen earlier in this chapter and will be seen more in some of the future chapters. Double peaks had 222

been seen in Figure 7-4 for the second melting of the pan blended samples after fast cooling. These, however, had the upper peak consistently at the polyamide melting temperature with the lower peak dependent upon polyamide concentration. explanation

in

this

Phase separation is, thus, not a convincing

case.

Another

explanation

of

melting/re-

crystallisation/melting of metastable crystallographic forms into more stable ones is also not entirely satisfactory as the temperature difference between the two peaks for a sample is over 15 0C.

The difference between the

metastable and stable peaks melting for rapid cooling of polyamide4,6/carbazole in Chapter 3 was only 7 0C. The double peaks are to be seen again in Figure 7-11 for some of the thermograms from the second melting in the DSC. They are not related to any uneven cooling in the ampoules themselves.

The reason for the double peaks in some of these samples

therefore does not have a clear explanation. The increasing size of the phenothiazine melting peaks is to be expected for an increasing amount of residual phenothiazine not being included in the polyamide-4,6/phenothiazine compound or, alternatively, incorporated to a small degree with the polyamide during the crystallisation in the ampoule. The thermograms here can also be compared with those of the second melt of the pan blended samples in section 7.3.1.3. The major thermal difference in the creation of the ampoule samples is that they have previously been crystallised at 2 0C/min instead of 25 0C/min. The difference seen here at the melting stage is the absence of separate polyamide-4,6 melt peaks near 290 0C. The pan blended samples with high phenothiazine levels crystallised some of the polyamide-4,6 due to phase separation with the fast cooling. That did not happen with the ampoule samples because there was more time during the slower cooling ramp to crystallise the two materials together without having the polyamide-4,6 crystallise separately at an earlier stage. 7.3.2.2


The results of ampoule sample overall crystallinity calculations are plotted below in Figure 7-7 along with a curve for a linear relationship in crystallinity based on a molar volume percentage. The overall percentage of crystallinity in the samples has a broad minimum at approximately 60% polyamide

with higher crystallinity for the 223

pure phenothiazine and

polyamide-4,6.

The minimum for the samples near 60% polyamide

concentration is pointing to most of the sample mass being tied up in the amorphous phase as they crystallise in the ampoules. The two materials are adversely affecting the crystallinity of each other. 100 Overall sample crystallinity (%)

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 7-7 Overall crystallinity determined from TGA and total first DSC melting enthalpy of polyamide-4,6/phenothiazine samples from ampoules. This reflects the crystallinity at the time the samples are taken from ampoules.

A linear relationship for colligative properties is usually based on molar volumes. In this case the specific density of both crystalline and amorphous polyamide-4,6 lie between 1.0 and 1.1 and that of phenothiazine is 1.38. That means a dip of 6-8% in crystallinity between 42 and 43 wt% polyamide based on a linear relationship on a molar volume basis. The maximum dip found in crystallinity calculations with respect to that line is even 25% lower. 7.3.2.3


Figure 7-8 shows the thermograms of the crystallisation of material meltblended in ampoules, taken to the melt in the DSC and then crystallised at 2 0C/min cooling rate. a) The

thermogram

polyamide-4,6

for

75PA46PTh

depressed

by

20 0C

shows because

the of

crystallisation

of

the

of

presence

phenothiazine followed by the crystallisation of a very small amount of phenothiazine

some

12 0C

below the 224

phenothiazine

crystallisation

temperature. This is showing the ability of some of the two materials to crystallise at the same time in a Flory-Huggins style depression of the crystallisation temperatures.

Figure 7-8 DSC thermograms of the first crystallisation at 2 0C/min in DSC of polyamide-4,6/phenothiazine ampoule material and the weight loss encountered over two melt/crystallisation cycles.

b) The 68PA46PTh sample shows minor crystallisation of polyamide-4,6 depressed by about 8 0C followed by the crystallisation in two stages of polyamide-4,6/phenothiazine depressed in a double peak by a total of 25 0C. The double peak is most likely due to uneven consistency of the concentration in the sample.

Finally there is the crystallisation of

phenothiazine depressed by 12 0C. c) The

62PA46PTh

thermogram


shows

peak

only

depressed

two by

peaks,

30 0C

and

a a

phenothiazine crystallisation peak depressed by 4 0C below that for phenothiazine. The reasoning here on the magnitude of depressions is the same as in the 75PA46PTh sample but with a greater phenothiazine concentration.

225

d) The 56PA46PTh sample has a single peak depressed by 20 0C below the polyamide-4,6 crystallisation temperature.

It is also a sample where

there has been substantial loss of phenothiazine by evaporation, as indicated by post-DSC weighing.

It has been mentioned earlier that

having to take samples with phenothiazine to the very high melting temperature of polyamide-4,6 for the trials presented problems with ensuring that the hermetic pan seal was perfect. There was no way to easily determine how well the pan was sealed prior to the DSC run. The other alternative, that the different prior history of this sample to the rest may have been responsible does not have a strong case as the prior history will have been removed by the melting process. That sample had been taken 40 0C above the highest temperature of thermal transitions. e) 36PA46PTh

is

showing

a

peak

45 0C

below

the

polyamide-4,6

crystallisation temperature and a phenothiazine peak depressed 3-4 0C below the phenothiazine crystallisation.

The first peak is a very

substantial drop in crystallisation temperature and indicates some concurrent solidification of phenothiazine with polyamide-4,6. f) The thermogram of 21PA46PTh has its first peak during crystallisation at more than 70 0C below the crystallisation temperature of polyamide-4,6. This is an extremely large depression of the crystallisation temperature and points towards a combined crystallisation of polyamide-4,6 and phenothiazine. The second crystallisation peak is only 2-3 0C below the crystallisation temperature of phenothiazine.

That would be expected

with the relatively low concentration of polyamide in the sample. There is a definite pattern in the thermograms in Figure 7-8 of large depression of the first crystallisation peak proportional to the phenothiazine concentration in the sample. The reduction in crystallisation temperature by more than 70 0C for one sample shows that there can be a strong Flory-Huggins style interaction between the two materials. The depression of crystallisation temperatures for each materials appears to be linear in concentration of the “contaminant” once evaporative loss of phenothiazine has been accounted for.

The two materials are able to solidify together,

resulting in the extremely large crystallisation depression.

226

A comparison with the (comparable) second crystallisation of the pan blended samples shows that the first crystallisation peak, of polyamide-4,6, is missing before the crystallisation peak for polyamide/phenothiazine with samples having higher levels of phenothiazine. That single peak generally found here above the phenothiazine crystallisation occurs because the cooling sample has more time at the slower cooling rate here to crystallise with phenothiazine in the inter-lamellar space rather than polyamide-4,6 partly phase separating with rapid cooling. The depression of the phenothiazine crystallisation temperature is far more modest and appears to be capped at a maximum depression of 12 0C when the polyamide concentration has risen to 63%. The

thermograms

are

strongly

affected

by

the

evaporative

loss

of

phenothiazine from some samples, as evidenced by lesser crystallisation temperature depression.

The loss of phenothiazine results in the sample

being more concentrated in polyamide than the blend that was begun with. 7.3.2.4


The crystallinity of phenothiazine and polyamide, as displayed in Figure 7-9 from the first cooling ramp in the DSC, does have some limitations because of the evaporative loss of phenothiazine to varying degrees from the samples. 100 phenothiazine polyamide-4,6

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

Polyamide concentration (%wt) Figure 7-9 Crystallinity of phenothiazine and polyamide from the first crystallisation in the DSC at 2 0C/min of polyamide-4,6/phenothiazine blends made in ampoules.

227

90

100

There is a drop in the phenothiazine enthalpy of crystallinity as the polyamide level is increased.

The values are, however, scattered as was

found with the polyamide-4,6/carbazole combination in Chapter 3 and unlike the consistently more linear relationships found in Chapters 4 to 6. We will see evidence in the remaining chapters that they also have scatter in phenothiazine enthalpy of crystallisation. The polyamide crystallinity rapidly increases again from a substantial minimum as the polyamide concentration is reduced below 50%. 7.3.2.5

Phase Diagrams for first time heating and cooling ampoule material in

0

Temperature ( C)

DSC 300 290 280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130

Liquid & crystallites Liquid Liquid



Solid & liquid Liquid & crystallites

Solid & liquid

TmPA46Pu TmPA46De TmPThDe TcPA46Pu TcPA46De TcPThDe


Solid

0

10

20


80

90

100

Figure 7-10 Non-equilibrium phase diagrams for polyamide-4,6, phenothiazine and their blends showing Flory-Huggins style depressions of melting and crystallisation temperatures.

The Flory-Huggins style depression of the melting/crystallisation of blends of polyamide-4,6 and phenothiazine with each material depressing the transition temperature of the other can be seen clearly in Figure 7-10 without the distraction of anomalies that usually occur near the crossover points. This is similar to the case of polyamide-4,6 with carbazole where the polyamide crystallisation temperature is higher than that of the diluent. In this case we can also see some melting/crystallisation of small regions of relatively pure polyamide-4,6.

228

7.3.2.6


The ampoule samples were passed through a repeat melt/crystallisation cycle in the DSC, as was done in earlier chapters. Some evaporation of phenothiazine has taken place in between the first and second DSC runs of the ampoule samples. The effect will be greater than in the pan blended samples because of the protracted times spent at high temperatures caused by the slower cooling ramp here Figure 7-11 below shows the DSC thermograms of the melt portions of the repeat DSC runs on the ampoule samples.

Figure 7-11 DSC thermograms of the second melt at 5 0C/min in the DSC of polyamide-4,6/phenothiazine ampoule material.

The pattern of the thermograms in Figure 7-11 for the repeat DSC melting is very similar to that of the first crystallisation in the DCS of samples made in ampoules. The exceptions are the 56PA46PTh sample that had already been noted to have had evaporative phenothiazine losses and the 68PA46PTh sample that shows some melting of nearly pure polyamide-4,6 as was seen in the previous crystallisation of the sample.

229

7.3.2.7

Third Crystallisation of Materials/Second DSC Crystallisation

The differences in Figure 7-12 compared with the first crystallisation in the DSC of these ampoule samples are not large if phenothiazine evaporative loss is taken into account. There has been slightly more of a tendency to crystallise out some of the polyamide and phenothiazine separately. That occurred with polyamide-4,6 crystallisation

for

36PA46Car

and

68PA46PTh.

The

phenothiazine

crystallised out separately to a greater extent with 68PA46PTh and the small phenothiazine crystallisation of 75PA46PTh had become broader and ill defined.

Figure 7-12 DSC thermograms of the second crystallisation at 2 0C/min in a DSC of polyamide-4,6/phenothiazine ampoule material.

The only other noteworthy change was the increase in phenothiazine crystallisation depression by 1 – 2 0C. The thermogram for 21PA46PTh displays noticeable noise in what should be a straight line apart from the (real) peak near 200 0C and the phenothiazine crystallisation near 160 0C. There were occasional problems with noise that are continuing to be worked on with the instrument manufacturer. 230

7.4 Fourier Transform Infra Red Spectroscopy FTIR in the Mid-Range and Near Infra-Red was carried out on samples of polyamide-4,6/phenothiazine melt blended in ampoules. In neither region of the electromagnetic spectrum was there any evidence of effects due to hydrogen bond interactions between the materials. Spectra can be found in Appendix D on CD.

7.5 Summary The combination polyamide-4,6 with phenothiazine had been expected to perform thermally in a similar manner to the polyamide-4,6/carbazole blends in Chapter 3.

In both cases, the polyamide melts at higher

temperatures than the material with which it is being blended and the two diluent

molecules

have

reasonably

similar

molecular

shapes.

The

phenothiazine has a melting temperature at 186 0C, much lower than the 246 0C carbazole melting temperature.

That low melting temperature

complicated the observations because

working temperatures for the

polyamide-4,6 had to be as high as 307 or 312 0C to melt this polyamide. That caused much more loss in phenothiazine because of evaporation. The low melting temperature of the phenothiazine also meant that the polyamide was much further thermodynamically from its normal melting temperature, making dissolution of polyamide-4,6 lamellae into the liquid more difficult. That point affected the initial dissolution kinetics of pure polyamide-4,6 grains in the powder blends. The initial melting of phenothiazine produced a highly phenothiazine concentrated solution that sits unfavourably for dissolution of polyamide-4,6. The carbazole had dissolved as much polyamide as possible (up to nearly the same mass of polyamide-4,6) during the first heating of carbazole and polyamide-4,6 powders in Chapter 3. This was associated with a depression of the carbazole melting temperature by some 10-15 0C. Further polyamide could only be dissolved as the temperature was increased.

The rate of

dissolution increased with temperature until all polyamide had dissolved or the polyamide-4,6 melting temperature was reached, where any remaining polyamide would melt at the normal polyamide-4,6 temperature. The change seen for this chapter with polyamide-4,6/phenothiazine blends is that there is no depression of the phenothiazine melting peak near 186 0C. 231

There is also little thermal activity to be seen until approximately 230 0C where there is a broad flat melting/dissolution of the polyamide. This has a reasonably sharp onset and finishing near 270 0C, depending upon the proportion of polyamide in the sample.

Not all the polyamide could be

dissolved at low levels of polyamide in the mix. The temperature needed to be increased to just under the polyamide-4,6 melting temperature in order to melt all the polyamide.

It therefore appears very difficult to dissolve

polyamide-4,6 in phenothiazine until 230 0C is reached or, alternatively, the kinetics of the dissolution are slow. That situation was quite different with material already melted once, and where the two materials were more intimately mixed at a molecular level. Crystallisation of the high temperature solution is dependent upon both concentration and cooling rate.

There was generally a depression of

crystallisation, regardless of cooling rate and there was no separate phenothiazine crystallisation at concentrations significantly over 60% polyamide.

The depression appears to be proportional to the amount of

phenothiazine, based on some measurements of sample mass after cooling to room temperature. Crystallisation at high cooling rates (25 0C/min) with lower percentages of polyamide resulted in crystallisation of some of the polyamide-4,6 at temperatures close to the normal polyamide-4,6 crystallisation temperature followed by solidification together near 220 0C, and later crystallisation of some remaining phenothiazine. At fast cooling rates there are usually one or three crystallisation peaks. Crystallisation at slower cooling rates (2 0C/min) mostly results in two peaks.

The first one is the depressed concurrent crystallisation of

polyamide-4,6

with

phenothiazine

under

the

gentler

crystallisation

conditions followed by the crystallisation of residual phenothiazine just under the phenothiazine crystallisation temperature. The depression of the first peak with slow crystallisation is proportional to the percentage phenothiazine and can reach a depression of 70 0C for a sample that is 21% polyamide-4,6.

232

Reheating all these samples leads to thermograms that reverse the crystallisation steps.

The temperatures are slightly different as would be

expected when involving the crystallisation and remelting of polymers. There appears to be little interaction between the two materials melting at temperatures below approximately 220 0C although the phenothiazine is at much higher temperatures than its melting temperature and will be at very low viscosity. There are strong interactions at temperatures above 220 0C if the molecules are intimately mixed and not powders or crystallised into separate phases. IR shows no evidence of hydrogen bonding, though. The repeat melting of blends after cooling quickly at 25 0C/min does not lead to the double main melting peaks found with polyamide-4,6/carbazole. There, the polyamide had been trapped in a metastable form that melted and recrystallised before the melt of the stable form. At first sight it could be said, in comparison, that there is also little interaction between carbazole and polyamide-4,6 but the carbazole appears to dissolve powders of semi-crystalline polyamide-4,6 more easily than phenothiazine does. There appears to be a basic difference between the two materials in their interaction with polyamide-4,6.

We will see in further

chapters how the two materials react similarly or differently with the other polyamides being studied and this will be brought together in the final chapter.

233

Chapter 8

POLYAMIDE-6 WITH PHENOTHIAZINE CONTENTS 8.1

Introduction

234

8.2


235

8.3


236

8.3.1

Pan Melt Blending

8.3.1.1 8.3.1.2 8.3.1.3 8.3.1.4

8.3.2

Ampoule Material

8.3.2.1 8.3.2.2 8.3.2.3 8.3.2.4 8.3.2.5 8.3.2.6 8.3.2.7

236

0

Melting Temperatures for first heating ramp of the powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min. 0 Melting Peak Temperatures for second heating ramp at 5 C/min 0 Crystallisation Peak Temperatures for second cooling ramp at 25 C/min

236 237 239 241

242

0

Melting Temperatures (First melt in DSC at 5 C/min) Overall Crystallinity 0 DSC Crystallisation Temperatures at 2 C/min for remelted ampoule material. Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC Third Melting of materials/Second DSC Melt Third Crystallisation of Materials/Second DSC Crystallisation

242 245 246 247 248 249 250

8.4


251

8.5

Summary

251

8.1 Introduction This chapter covers the melt blending of polyamide-6 with phenothiazine and provides the opportunity to make comparisons with those of polyamide-4,6 with phenothiazine and polyamide-6 with carbazole. The combinations here differ from that with polyamide-4,6/phenothiazine because the polyamide repeat units are longer and differently constructed, polyamide-6 being a polyamide-n type rather than a polyamide-m,n.

The

melting temperature of phenothiazine is much closer to that of polyamide-6 than polyamide-4,6. It differs from polyamide-6/carbazole in that now the second material melts before the polyamide-6 rather than after polyamide-6 melting. There may also be differences because the molecular shape and electron densities of the carbazole and phenothiazine do differ. 234

It will be shown in this chapter that generally the melting behaviour of the melt blends is similar to that in earlier chapters with polyamide-6, polyamide-6,9 and polyamide-6,12 combined with carbazole, but having reversed roles for the two materials. A small amount of the polyamide will phase separate and crystallise separately in a phase uncontaminated with phenothiazine.

That happens

when the concentration of polyamide-6 is lower than 55% and it occurs whether the melt is being cooled quickly or slowly. This is to be seen also for polyamide-4,6 blended with either carbazole or phenothiazine.


8.2 Thermogravimetric Analysis 100 PA6PTh 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 8-1 Actual versus Theoretical weight percentage polyamide in samples from polyamide-6/phenothiazine blends from ampoules.

Actual percentages of polyamide in samples from ampoules were determined by TGA. They differed in some cases from the percentages calculated from the masses of the constituent materials put into the ampoules.

That

difference would be due to local variations caused by phase separation, reorganisation during crystallisation or incomplete mixing in the melt, as mentioned in previous chapters. A comparison of the actual percentages of polyamide to those expected from the weights used in the ampoules is shown in Figure 8-1.

235

Several of the points are much more than 5% from the expected concentrations of polyamide based on weights used in making the ampoule material.

The large differences from the expected seen in Figure 8-1 are

much greater than seen in Sections 3.3.1 and 4.2.2 on general variability between samples and are real ones in concentration. The largest differences are encountered with one of the earlier ampoules before the furnace cooling had been optimised for consistency.

8.3 Differential Scanning Calorimetry Polyamide-6 normally has a melting peak close to 223 0C with crystallisation peaks in the range 173-198 0C depending on crystallisation conditions, as mentioned in Chapter 4. 8.3.1 8.3.1.1


Figure 8-2 DSC thermograms during the first melting in the DSC at 5 0C/min of polyamide-6, phenothiazine and powder mixtures of the two.

Figure 8-2 shows the DSC thermograms of polyamide-6, phenothiazine and their mixtures taken as powder samples in DSC pans to the melt at 5 0C/min. These are described in detail below: 236

a) Pure polyamide 6 powder melts in the same way as in Chapter 4 ie dependent upon the previous thermal history of the sample. b) The 83PA6PTh sample undergoes weak thermal activity (seen more clearly in Appendix A) in the region near 160 0C and caused by the small amount of phenothiazine present being absorbed into the amorphous structure of polyamide grains. This plasticises the amorphous polymer lowering its viscosity.

The dissolution of polymer lamellae into the

plasticised amorphous polyamide takes place slowly at temperatures lower than the normal polyamide-6 melting peak in what is a TLS peak for polyamide. c) The 63PA6PTh thermogram displays a eutectic peak 8 0C below the normal phenothiazine melting temperature. followed by a small peak at the phenothiazine melting temperature and a weak TLS peak for polyamide-6 which has a maximum at 200 0C and is complete by 209 0C. The fact that there is a peak at the phenothiazine melting temperature and a TLS peak afterwards shows that these two materials had difficulty in forming a solution at these temperatures and timeframe of the experiment or that the materials were in a region of phase space where phase separation occurs. d) Sample 38PA6PTh shows a similar style of behaviour but modified in the eutectic and TLS peak peaks by the lesser percentage of polyamide-6 and in the middle peak by the larger amount of phenothiazine being melted at the phenothiazine melting temperature. The first melting of polyamide and diluent powders did not result in as easy dissolution of the materials into a eutectic melt as was seen in other material combinations.

Either there was phase separation or the kinetics of

dissolution were slow at the temperatures involved. The eutectic melts were encountered for polyamide concentrations less than or equal to 65% by weight and involved a depression of 8 0C below the normal diluent melting temperature. 8.3.1.2

Crystallisation for first cooling ramp of the molten blend at 25 0C/min.

Figure 8-3 below shows the DSC thermograms for the powder samples of Section 8.3.1.1 being cooled from the melt at 25 0C/min to room temperature. 237

a) The first crystallisation of 83PA6PTh is a skewed main peak. It can be shown on closer examination to be a double peak, as can be seen in Appendix A. The first peak is a small amount of almost pure polyamide-6 crystallising, followed by crystallisation of polyamide-6 together with phenothiazine. The separate initial crystallisation of polyamide-6 points towards phase separation under the conditions of cooling rate and concentration used.

Figure 8-3 DSC thermogram for first crystallisation of pan blended polyamide-6/phenothiazine from the melt at 25 0C/min.

b) The cooling ramp for 63PA6PTh shows some initial crystallisation of polyamide-6

followed

polyamide-6/phenothiazine

by

the

phase

crystallisation

between

the

of

a

crystallisation

temperatures of the constituent materials, and concluding with the crystallisation of phenothiazine.

The first peak is showing phase

separation at those high temperature/concentration conditions. There is apparently the possibility at lower temperatures for polyamide-6 and phenothiazine to crystallise at the same time with phenothiazine in the inter-lamellar space. The last peak of this thermogram is sharp with a steeply rising face but is decidedly not as “spiky” as pure phenothiazine. 238

That lesser spikiness points to solidification of some polyamide-6 combined

with

crystallising

phenothiazine

embedded

within

the

remaining amorphous mix. The lead-in to the last peak shows that there is an attempt to begin crystallisation of residual polyamide-6 that by this stage has been suppressed from crystallisation by nearly 30 0C.

The

polyamide-6 depression of the phenothiazine crystallisation is by approximately 15 0C. c) The cooling thermogram of 38PA6PTh is similar to the two samples immediately above in that there is an initial small crystallisation of polyamide-6 because of phase separation. The main peak is actually a double peak with (mainly) phenothiazine crystallising followed by the depressed

crystallisation

of

polyamide-6

together

with

some

phenothiazine. The three polyamide-6/phenothiazine pan blended samples all have crystallisation of a small amount of polyamide-6 at the normal polyamide-6 crystallisation temperature. These can indicate an incompatibility between the two materials at that temperature/cooling rate, pointing to phase separation occurring. That crystallisation of small amounts of the polyamide was seen in polyamide-4,6/carbazole and polyamide-4,6/phenothiazine combinations of previous chapters. All three PA6PTh samples also have solidification of the two materials together at temperatures below the polyamide-6 crystallisation temperature and are depressed more with higher levels of phenothiazine. The crystallisation of phenothiazine takes place independently and occurs with the crystallisation of some polyamide-6 included in it. This occurs in the same way as seen in Figure 4-5

with pan samples for the

polyamide-6/carbazole combination in Chapter 4 and leading to a sharp non-spiky peak. 8.3.1.3


The DSC thermograms for the remelting at 5 0C/min of samples originally melted in pans from the powders are displayed in Figure 8-4. a) The first faint melting at 160 0C in 83PA6PTh has become fainter in the repeat heating run of the sample. The main peak has become a single, 239

slightly rounder, peak and lies at marginally higher temperatures. These changes are due to the two materials being in more intimate molecular contact and to the slight loss of some phenothiazine.

Figure 8-4 DSC thermogram at 5 0C/min of second polyamide-6/phenothiazine materials previously melt powders then crystallised in pans.

melt of blended

b) It was mentioned in the preceding section that the 63PA6PTh sample had crystallised

into

three


phases, and

polyamide-6/phenothiazine portion.

a

a very

polyamide-6,

a

phenothiazine-rich

The previous double main melting

peak from 8.3.1.1 has dropped to a lower temperature on reheating and is now 15 0C below the phenothiazine melting temperature.

The TLS

dissolution of polyamide-6 evident in the first melting has now moved to slightly lower temperatures and is complete by just over 200 0C. This is most likely due to the better mixing of the materials from the previous crystallisation. There is a slight melting peak at the polyamide-6 melting temperature indicating a small amount of pure polyamide existed in the structure that is being remelted.

240

c) The 38PA6PTh sample has reduced the peak temperature of the original main melting (double) endotherm by over 10 0C upon remelting.

The

melting is a triple peak with the two main phases mentioned in the crystallisation description now melting again and followed by a small, TLS peak shoulder on the higher temperature side. There is a slight melting peak at the polyamide-6 melting temperature that represents the melting of the small near-pure polyamide-6 portion seen during crystallisation. The melting of some fractions at 8-10 0C lower seen in this set of thermograms is due to the better mixing of the two materials. The results are best seen in the light of the earlier difficulties in achieving the phenothiazine solution.

powder

dissolution

in

the


That was seen from the separate peaks at exactly the

phenothiazine melting temperatures in Figure 8-2. Very small melting peaks at the polyamide-6 melting temperature for all three samples shows that there is reluctance in dissolving the polyamide-6 phase into the high temperature solution at temperatures above 200 0C. 8.3.1.4


Figure 8-5 DSC thermograms of the second crystallisation of pan blended polyamide-6/phenothiazine samples in the DSC, cooling at 25 0C/min.

241

Figure 8-5 shows the thermograms of the pan blended samples during their second crystallisation cycle at 25 0C/min. a) Sample 83PA6PTh is virtually identical in re-crystallisation to the original crystallisation. b) The first two crystallisation endotherms for sample 63PA6PTh are quite similar to the original ones on first crystallisation of the pan blended samples. An observable difference is the slightly larger size of the two peaks, meaning that the sample is more crystalline although no explanation is available for this. The main phenothiazine-based peak is slightly smaller and is some 3 0C lower in temperature due to the evaporative loss of phenothiazine. c) The re-crystallisation of 38PA6PTh takes place, as before, with a barely visible crystallisation of some polyamide-6 due to phase separation. The peak is followed, during the cooling, by a single main peak for phenothiazine at lower temperatures. The peak is slightly more “spiky” showing that there is less polyamide-6 incorporated in that crystalline phase.

The crystallisation temperature is depressed a further 6 0C to

make it 11 0C below the pure phenothiazine crystallisation temperature. All three blends again show some crystallisation of polyamide-6 during the re-crystallisation from the melt. The indications, again, are that there is limited compatibility between the two materials, resulting in liquid-liquid phase separation of polyamide-6 from the diluent in the melt at higher temperatures. There does appear to be more compatibility between the two materials at lower temperatures resulting in some crystallisation of phenothiazine together with polyamide-6 at the 25 0C/min cooling rate. 8.3.2 8.3.2.1

Ampoule Material Melting Temperatures (First melt in DSC at 5 0C/min)

The thermograms in Figure 8-6 show the heating curves (at 5 0C/min) from melting samples formed originally in ampoules for the first time in the DSC. Thermograms from ampoule samples in Figure 8-6 were obtained in the DSC after the samples were crystallised in ampoules at a slow rate. The results are reasonably in agreement with the thermograms from the second melting of the pan blended samples despite the latter having been crystallised at the 242

much faster 25 0C/min. The descriptions of this set of data will be partially grouped because of the number of thermograms in total and the obvious broad similarity of some of them.

It should be remembered that the

percentages of polyamide in the actual sample measured in the DSC may differ by a small amount from those of the TGA measurements.

This is

because different samples had to be used for TGA and DSC, although taken from next to each other in the bulk ampoule sample.

That means, for

example, 61PA6PTh used in the DSC may have had slightly less polyamide-6 than 60PA6PTh.

Figure 8-6 DSC thermograms of polyamide-6/phenothiazine samples from ampoules in their first melt in the DSC at 5 0C/min.

Samples 69PA6PTh, 55PA6PTh and 29PA6PTh were from ampoules 17 and 18. These were made before the furnace process had been set up for best consistency. In fact, samples 69PA6PTh and 29PA6PTh are both from the same ampoule. The results have been included, but with the recognition that they are not entirely representative of the ampoule melt blending process generally used for ampoule data in the thesis. a) The polyamide-6 sample has a double peak with a shoulder for the melting/re-crystallisation of the metastable material into the stable form 243

before the main melting peak. The position of the shoulder is typical of the cooling rate used in the furnace for the ampoule. b) 74PA6PTh is similar to the pan blended 83PA6PTh except that, with a lower polyamide level, there is more depression of the main melting peak temperature. c) The three samples with polyamide concentration in the range 60-70% show similar thermal performance to the pan blended 63PA6PTh sample although

the

first

peak

for

69PA6PTh

begins

at

slightly

lower

temperatures than the other two of this group. An average of the three would be very close in peak temperatures and peak heights to the pan blended 63PA6PTh sample on its second melting ramp. This is despite the cooling ramps differing strongly between pan and ampoule samples in their previous thermal treatments. d) There are no concentration equivalents in the pan blended samples to the 56PA6PTh and 55PA6PTh samples from ampoules. The two samples together show a transition from the thermograms of the previous three and the sample 29PA6PTh to be described below.

The 56PA6PTh

thermogram is reasonably similar to the ampoule samples with concentrations between 60 and 70% polyamide.

The pan blended

38PA6PTh sample in Section 8.3.1.3 had a first melting peak at the same temperature as both these samples and had displayed a small TLS dissolution of the polyamide once the temperature of the saturated solution was increased. The smaller size of the first peak and larger size of the second peak for 55PA6PTh compared to 56PA6PTh reflect actual differences

in

polyamide

levels

more

than

the

1%

from

TGA

measurements. The overall slightly lower crystallinity of the 55PA6PTh sample is, most likely, because the sample was from Ampoule 17 made before the furnace cooling of the ampoule samples had been completely refined. e) The 29PA6PTh sample has a main peak with no further evidence of dissolution of polyamide-6.

A comparison with the pan blended

38PA6PTh shows that the polyamide-6 concentration at saturation for 180 0C must be between 29 and 38%. There is a very small peak at 145 0C near the temperature of a fainter peak barely visible with 244

55PA6PTh. The origin of this is unclear and there is no equivalent one with the 38PA6PTh pan blended sample at its second melting. It was mentioned earlier that these were from early ampoules where the cooling profile had not been fully refined, so it is possible that that may have contributed to the differences. These thermograms of ampoule-formed samples in their first pass through a DSC

cycle

are

quite

consistent

with

the

pan

blended

polyamide-6/phenothiazine samples at their second ramp. They show the polyamide-6/phenothiazine melt at depressed melting temperatures. This is followed, for concentrations of the polyamide above 35%, by dissolution of excess polyamide-6 in the saturated solution. In principle, the behaviour is similar to that seen with several polyamides and carbazole, but with reversed roles. One specific curiosity in the thermograms is the varying temperature of the first peaks with this series of thermograms. The 29PA6PTh and 69PA6PTh samples with the low first melting temperatures both come from ampoule 18. That ampoule had been fired in the furnace before the furnace profile had been completely refined. There could be an effect from that cooling ramp on re-melting temperatures.

That does not explain the melting temperature

differences between samples 56PA6PTh and 60PA6PTh. 8.3.2.2


The percentages of polyamide were used with the total enthalpy of the first melting heating ramp to calculate the overall crystallinity of ampoule samples in the same manner as in earlier chapters. The results are plotted below in Figure 8-7. This represents the crystallinity of the samples as they were taken from the ampoules. The overall crystallinity in Figure 8-7 obtained from the first melting in the DSC of ampoule material shows a rapid drop in crystallinity of the sample to approximately 40% at polyamide levels of 30% or higher. There is a relatively constant crystallinity for polyamide concentrations above that.

The non-

linear overall relationship means that the blending process is adversely affecting crystallinity and shows that there is some interaction between the materials leading to a non-zero free energy of mixing. 245


100 90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

Polyamide concentration (%wt) Figure 8-7 Overall crystallinity of polyamide-6/phenothiazine ampoule samples determined from TGA and total first DSC melting enthalpy.

8.3.2.3


Figure 8-8 DSC thermograms of the first crystallisation in the DSC at 2 0C/min of polyamide-6/phenothiazine ampoule samples.

246

100

Figure 8-8 shows the thermograms of the crystallisation of material meltblended in ampoules, taken to the melt in the DSC and now being crystallised at 2 0C/min. The two polyamide-6/phenothiazine samples with the highest percentages of phenothiazine have small peaks at exactly the same temperature as polyamide-6 crystallisation.

This shows phase separation of polyamide-6

from the rest of the solution at the cooling rate of 2 0C/min.

Similar

crystallisation of polyamide occurred with the pan blended samples during cooling at 25 0C/min during their second crystallisation ramp, as described in Section 8.3.1.4. The crystallisation of polyamide-6 at moderate to low polyamide levels is thus occurring regardless of whether the melt is cooled quickly or slowly.

It points to incompatibility of the polyamide-6 in the

solution at medium to low polyamide concentrations when temperatures are near 200 0C. There is a difference between fast and slow cooling at higher polyamide levels in that the polyamide does not phase separate and crystallise out at the slower cooling rate, unlike with the faster cooled pan blended samples. An overall trend of increasing depression of the polyamide-6 crystallisation can clearly be seen from the thermograms as the level of phenothiazine is made higher. This shows some interaction between the two materials and the fact that no polyamide-6 crystallises separately under this slower cooling rate. The bulk of the polyamide-6 crystallising in these later stages can be depressed by up to 35 0C. Phenothiazine crystallisation depression is quite variable between 12 and 25 0C. 8.3.2.4


The crystallinity of phenothiazine and polyamide can be calculated as in earlier chapters and is displayed in Figure 8-9. This is determined from the first cooling ramp in the DSC although it does have some limitations because of the evaporative loss of phenothiazine to varying degrees from the samples. The crystallinity of the phenothiazine, as with the crystallisation depression, is very variable but reduces to zero above approximately 75% polyamide-6. The variability has also been seen with several other polyamide/diluent 247

combinations, but only those with the melting temperature of the polyamide higher than that of the diluent. The crystallinity of the non-phenothiazine portion has a maximum between 60 and 70% polyamide that is higher than that of the pure polyamide.


100 phenothiazine polyamide-6

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 8-9 Crystallinity of phenothiazine and Non-phenothiazine parts from crystallisation of polyamide-6 blend from the melt in the DSC.

8.3.2.5


230 220

Liquid

Temperature (0C)

210 Solid & liquid Liquid

200 190


180

Solid & Liquid Liquid & crystallites Solid & Liquid

170 Solid Liquid & crystallites

160

Solid Solid & Liquid

150 140

Solid


130 0

10

20


80

90

Figure 8-10 Non-equilibrium phase diagrams for polyamide-6, phenothiazine and their blends showing Flory-Huggins style depression of the peak temperatures

248

100

The non-equilibrium phase diagrams in Figure 8-10 show Flory-Huggins style crystallisation depressions of the polyamide by the diluent and vice versa. The melting phase diagram appears to have a near horizontal first melting transition above 30% polyamide concentration which could be eutectic melting. There is also some crystallisation of small amounts of nearly pure polyamide at temperatures very close to the normal polyamide-6 crystallisation temperature. That was also found with both polyamide-4,6/diluent blends but not with the polyamide-6, polyamide-6,9, or polyamide-6,12/carbazole blends where the polyamide crystallisation temperature is lower than the diluent one.

8.3.2.6


The ampoule samples in pans from the first DSC runs were passed through a repeat melt/crystallisation cycle in the DSC as was done in earlier chapters.

Figure 8-11 below shows the DSC thermograms of the melt

portions of the repeat DSC runs.

Figure 8-11 DSC thermograms of the second melt at 5 0C/min in DSC of polyamide-6/phenothiazine ampoule material.

249

Some evaporation of phenothiazine will have taken place in between the first and second DSC runs of the ampoule samples. The effect will be greater than in the pan blended samples because of the protracted times spent at high temperatures due to the slower cooling ramp used with these samples. Small differences are seen between the first melt in the DSC and the second one. The first melting peaks with this DSC run are smaller and the second peaks are larger than with the first DSC run. These factors are due to the evaporation of some phenothiazine. Less of the polyamide can be dissolved before a saturated solution is reached. There is more left to dissolve into the solution only when higher temperatures have been reached. There are melting peaks also evident at the normal polyamide-6 melting temperature in the cases of the 56PA6PTh and 29PA6PTh samples. This is the small amount of polyamide-6 seen to crystallise as a pure polyamide-6 phase in the previous section.

8.3.2.7


Figure 8-12 Thermograms of the second crystallisation at 2 0C/min in a DSC polyamide-6/phenothiazine ampoule material.

Figure 8-12 depicts the thermograms from the second crystallisation in the DSC at 2 0C/min of samples originally formed in ampoules. 250

The small differences to be seen in Figure 8-12 compared to the first DSC run Figure 8-8 are due to the loss of some phenothiazine by evaporation. 74PA6PTh and 61PA6PTh stand out most noticeably in this respect. This latter has lost enough phenothiazine to result in no separate phenothiazine crystallisation “spike” in the range 135-165 0C.

The only other significant

change is the 29PA6PTh sample has lost enough phenothiazine that there is now

a

separate,

approximately

Gaussian,

polyamide-6/phenothiazine peaking at 143 0C.

crystallisation

of

This means that there is

polyamide-6 solidified with the phenothiazine in a phase depressed 50 0C below the normal polyamide-6 crystallisation temperature.

8.4 Fourier Transform Infra Red Spectroscopy FTIR was carried out on ampoule samples in the Mid Range IR with photoacoustic and in the Near IR with DRIFT to search for hydrogen bond interactions between the two materials. A close inspection of the results in both ranges did not show any such interaction.

Results may be seen in

Appendix D on CD.

8.5 Summary The

results

in

this

chapter

have

demonstrated

similar

effects

in

dissolution/melting and in crystallisation to those seen in some earlier chapters, sometimes with the roles of polyamide and diluent reversed because melting temperatures are in reverse order.

The lower melting

material has melted at depressed temperatures, dissolving some of the higher melting material until the solution became saturated.

The higher

melting material then required higher temperatures before more could be dissolved into the saturated solution. Another observation common with polyamide-4,6/carbazole and polyamide4,6/phenothiazine blends was some polyamide melting and crystallising exactly at the normal polyamide melting temperatures. This shows reduced compatibility of the polyamide-6 in polyamide-6/phenothiazine solutions at high temperatures. The crystallisation of pure polyamide-6 was occurring at fast and slow cooling rates when the level of polyamide was medium to low. It only occurred at fast cooling rates when the level of polyamide-6 was moderate to high, showing that the two materials could be marginally

251

compatible at high temperatures if the polyamide-6 level was sufficiently high. In addition to the overall common factors, a part of the phenothiazine melted at exactly the normal phenothiazine melting temperature in the same way as seen strongly with polyamide-4,6/phenothiazine in Chapter 7.

The

difference here was the presence of double peaks at both phenothiazine depressed and at normal phenothiazine melting temperatures. The melting at normal phenothiazine melting temperatures constituted the majority of the double peak for the highest level of phenothiazine with pan blended samples.

These curves indicate that dissolution/melting temperature

depression is taking place up to a certain level of phenothiazine and then phenothiazine does not continue melting until the temperature has been raised further. polyamide.

At that stage the temperature is too low to dissolve more

That situation differs from the various polyamides combined

with carbazole where the lower melting material completely melted at a depressed temperature and went into dissolution of the higher melting material as the temperature was raised sufficiently. This combination of a polyamide with phenothiazine showed with FTIR that there was no hydrogen bond interaction between the polyamide-6 and phenothiazine

as

for


polyamide/carbazole combinations.

252

and

all

the

Chapter 9

POLYAMIDE-6,9 WITH PHENOTHIAZINE CONTENTS 9.1

Introduction

253

9.2


254

9.3


255

9.3.1

Pan Melt Blending

9.3.1.1 9.3.1.2 9.3.1.3 9.3.1.4

9.3.2

Ampoule Material

9.3.2.1 9.3.2.2 9.3.2.3 9.3.2.4 9.3.2.5 9.3.2.6 9.3.2.7

255

0

Melting Temperatures for first heating ramp of dry powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min 0 Melting Peak Temperatures for second heating ramp at 5 C/min 0 Crystallisation Peak Temperatures for second cooling ramp at 25 C/min 0

255 257 259 261

262

Melting Temperatures (First melt in DSC at 5 C/min) Overall Crystallinity 0 First crystallisation in the DSC at 2 C/min of ampoule samples Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC 0 Third Melting (at 5 C/min) of ampoule materials/Second DSC Melt Third Crystallisation of Materials/Second DSC Crystallisation

262 264 265 266 267 268 269

9.4


270

9.5

Summary

270

9.1 Introduction Chapter 9 deals with the blending of polyamide-6,9 with phenothiazine. In this case the melting temperature of the phenothiazine is lower than that of the polyamide-6,9 rather than the situation in Chapter 5 with the carbazole having a higher melting temperature than the polyamide-6,9. Polyamide-6,9 is different from the other polyamides studied here in that it is not a polyamide-n type and it is an even-odd rather than polyamide-m,n.

even-even

The length of the diamine section of the repeat unit is

moderate at 6 carbons and the diacid is reasonably long at 9 carbons. This means that the amide density is lower than the industry-standard polyamide-6 and polyamide-6,6 which should give more flexibility of the chains. The even-odd status will play its part in the ability of the chains to form hydrogen bonds as they crystallise and causes this polyamide to have 253

the lowest melting temperature of the group of polyamides being studied here. The way in which this polyamide and phenothiazine interact is not easily predictable because of the less favourable matching of amide groups to form the bonds within the crystalline structure. The previous two chapters on polyamide-4,6 or polyamide-6 blended with phenothiazine have started to show slightly less compatibility compared to their blends with carbazole and this chapter points again to some increase in phase separation.

9.2 Thermogravimetric Analysis TGA results of the plateau level at 300 0C for samples taken from ampoules are compared in Figure 9-1 with the expected levels for polyamide from the weights of materials used in the ampoules. In most cases the actual values are close to the expected ones. polyamide-6,9/phenothiazine

There is only one sample in this set of

ampoule

samples

where

the

actual


concentration deviated a considerable amount from the expected. 100 PA69PTh 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 9-1 Actual vs expected weight percentage polyamide in samples from polyamide-6,9/phenothiazine ampoule samples.

That deviant data point had 27% polyamide with the “R4S4” TGA method and a separate sample from nearby in the bulk had shown 21% polyamide with the “straight” method, a reasonably consistent result. The ampoule had been one of those from just before the furnace profiles had been completely 254

refined.

This is, most likely, the reason for the deviation so far from the

expected.

9.3 Differential Scanning Calorimetry It was mentioned in Chapter 5 that polyamide-6,9 has a melting temperature close to 210 0C with crystallisation in the range 175-193 0C, depending on cooling rate. 9.3.1 9.3.1.1

Pan Melt Blending Melting Temperatures for first heating ramp of dry powders at 5 0C/min

DSC thermograms are shown in Figure 9-2 for the first melting of powder samples of polyamide-6,9, phenothiazine and their mixtures at a heating rate of 5 0C/min.

The thermograms for the first melting in DSC pans of

polyamide-6,9, phenothiazine and three mixtures of the powders are shown in Figure 9-2.

Figure 9-2 DSC thermograms during the first melting at 5 0C/min of polyamide-6,9/phenothiazine and powder mixtures blended in DSC pans.

a) The polyamide-6,9 sample has a weak endothermic peak near 160 0C for the melting of the metastable form resulting from previous thermal 255

history. A marked exotherm for the re-crystallisation into the stable form lies prior to the main peak for the melting of the stable form. There is a minor peak above the main melting peak that is unknown in origin. Several repeat DSC runs have been made of virgin vacuum dried powder from several vacuum dryings and the peaks are to be found in all of them. They would have been found with the other polyamides if it were contamination due to the grinding and vacuum drying. b) The 68PA69PTh thermogram has a complex, flat triple peak spreading over the range 160-190 0C with the two strongest peaks at 165 and 181 0C.

One of the two materials, most likely the polyamide-6,9, is

beginning to melt and dissolve the other when another process begins, leading to the second peak. A better understanding of what is happening will come from the following two samples and the remelting described in a later section.

The third peak is a shoulder on the trailing edge of the

combined peak and appears to be at the normal phenothiazine melting temperature.

The thermogram is flat above 190 0C indicating that all

polyamide-6,9 has been consumed by this stage.

In addition, the

complex peaks are preceded by an almost imperceptible dip centred in the range 145-150 0C. This appears to be related to the slight melting of the metastable crystalline form seen in the polyamide-6,9 thermogram. This dip is also to be seen in the following two samples. c) The thermogram for sample 50PA69PTh is similar to that for 68PA69PTh except that the two main peaks of the composite peak are further resolved into two peaks.

There are some other small differences.

The small

shoulder at the same temperature as the phenothiazine melting temperature is missing.

There is a faint deviation from the horizontal

over the range 185-195 0C that is most likely the TLS peak for dissolution of the small remaining amount of polyamide-6,9 in the sample.

Noise

levels in the DSC signal at temperatures above the melting peaks appear to be higher than normal. There seems to be no reason for that and the noise disappears at 160 0C as the sample temperature is lowered in cooling as can be seen in the figure immediately below.

Intermittent

problems had been encountered with the DSC equipment regarding noisy

256

signals and the manufacturer’s service personnel had difficulties in rectifying the problems. d) The thermogram for 39PA69PTh is nearly the same form as the 50PA69PTh. A difference is that the relative peak heights of the main two peaks are now unequal, with the first peak smaller than the second, the second taking place a little higher at very close to the phenothiazine melting temperature.

The TLS peak shoulder also extends to slightly

higher temperatures.

The difference in heights relate to the higher

phenothiazine concentration in this sample and allows the assignment of the second peak in all three to phenothiazine rather than polyamide-6,9. We see here the multi-stage melting/dissolution of polyamide-6,9 powder with phenothiazine in which the polyamide appears to melt/dissolve first at substantially depressed temperatures below that of phenothiazine. The phenothiazine has a limited melting temperature depression due to polyamide-6,9

interactions.

Viewed

over

the

three

samples,

the

phenothiazine has a TLS peak as it is taken up in the saturated solution. The sample with the lowest concentration of phenothiazine has a small subpeak on the trailing edge of the composite peak. The small shoulders on the higher temperature side of the multiple peaks are above the phenothiazine melting temperature and must be the dissolution of residual polyamide. That implies that the dissolution of polyamide grains in the eutectic solution is retarded due to kinetic effects or that the system is in a region of phase space where the formation of a solution is unfavourable. 9.3.1.2 The

Crystallisation for first cooling ramp of the molten blend at 25 0C/min DSC

thermograms

from

the

first

crystallisation

ramp

for

polyamide-6,9/phenothiazine powder mixtures taken to the melt and cooled at 25 0C/min are to be seen in Figure 9-3. a) The crystallisation thermogram of 63PA69PTh shows a small peak for nearly pure polyamide-6,9 followed by a treble peak comprising (sequentially)

the

crystallisation

of

phenothiazine,

a

polyamide-6,9/phenothiazine peak on the falling side and a final peak over 10 0C lower at just under 140 0C. The first two peaks of the treble peak are almost coincident.

That can be seen more clearly in an

expanded view in Appendix A. The observation of the diluent beginning 257

to crystallise and this triggering the simultaneous crystallisation of diluent with polyamide is also to be seen with some polyamide/diluent combinations in other chapters.

Figure 9-3 DSC thermogram for the first crystallisation of pan blended polyamide-6,9/phenothiazine blends made from powders taken to the melt and being cooled here at 25 0C/min.

b) 50PA69PTh has a small peak near the polyamide-6,9 crystallisation temperature but 4 – 5 0C lower than the equivalent one in 63PA69PTh. Unfortunately, the signal from the DSC for this cooling run was noisy which partly obscures the small peak caused by phase separation of some polyamide-6,9.

The slightly lower temperature indicates a very

small amount of phenothiazine is being incorporated with the polyamide at that point. There is a (double) peak starting at 135 0C which begins with

phenothiazine

crystallising

and

evolves

into

the

concurrent

crystallisation of polyamide-6,9 and phenothiazine. The double peak here is more evident when looking at the expanded curves in Appendix A. c) The crystallisation curve for 39PA69PTh is a curious triple peak, starting with the crystallisation of nearly pure phenothiazine and followed by two subsequent crystallisations of differing proportions of polyamide-6,9 with 258

phenothiazine.

It is quite unlike other cases seen in the rest of this

research work.

It is interesting that, unlike many other cases where

there is some phase separation, there is no crystallisation of nearly pure polyamide for the sample with the highest level of phenothiazine. There is no explanation available for the particular triple peak. Crystallisation of a small amount of nearly pure polyamide has taken place with the two samples having higher polyamide content, apparently after some phase separation.

There has been substantial crystallisation

depression of those two, one of them by nearly 50 0C below the polyamide-6,9 crystallisation temperature.

The sample with the most

phenothiazine goes through a complex triple crystallisation process but does not show evidence of phase separation. 9.3.1.3


Figure 9-4 DSC thermogram at 5 0C/min of second melt of materials previously blended and crystallised in DSC pans for polyamide-6,9/phenothiazine mixtures.

Figure 9-4 is the set of DSC thermograms for the reheating at 5 0C/min of polyamide-6,9/phenothiazine powder mixtures melt blended in DCS pans after having been cooled previously at 25 0C/min. 259

a) The polyamide-6,9 thermogram shows more pronounced melting and recrystallisation of metastable lamellae before the main melting peak as a result of the fast cooling in the previous crystallisation of this sample. b) The 63PA69PTh thermogram for remelting the sample shows further resolution into separate peaks compared to that seen with the initial melting of the two powders. This sample, and the others of this group, had been crystallised at the fast cooling rate of 25 0C/min. That cooling rate leads in the case of pure polyamide-6,9 to a metastable crystalline form that changes to a more stable form as it is heated towards the polyamide-6,9 melting temperature.

There is a weak exothermic peak

near 150 0C that is a washed out re-crystallisation of the polyamide-6,9 lamellae into the stable form. A melting peak at higher temperatures for saturated polyamide-6,9/phenothiazine is followed by the TLS peak for the dissolution of polyamide-6,9/phenothiazine crystallised together. There is melting of some almost pure polyamide-6,9 that had been seen to crystallise near the normal polyamide-6,9 crystallisation temperature in Figure 9-3. c) A double peak and a single peak are seen in the heating ramp for 50PA69PTh. The first endothermic peak, near 157 0C, is most likely to be melting

and

is

followed

by

the

exothermic

re-crystallisation

of

polyamide-6,9 from the metastable crystalline form produced in the prior fast crystallisation. Similar effects can be seen within the thermogram for re-melting of the pan blended 63PA69Car sample in Chapter 5. The main peak is at the same temperature as the 63PA69PTh main peak in Figure 9-4 above. Alternatively, the peaks could be the melting of two different polyamide-6,9/phenothiazine

compositions.

There

is no

significant TLS peak with this sample, indicating that it is close to the eutectic composition.

There is also a small peak at the polyamide-6,9

melting temperature as with 63PA69PTh, and showing the remelting of the small portion that crystallised as polyamide-6,9 in Figure 9-3. d) The 39PA69PTh thermogram shows a main peak 2-3 0C higher than the main peaks of the other two samples. That peak is followed at higher temperatures by a TLS peak.

There is no minor peak near the

260

polyamide-6,9 melting temperature in agreement with none being present during the prior crystallisation. We find here much more defined melting curves than during the first melt. This is similar to that seen with other material combinations. In all cases there is a main peak depressed from near 210 0C to close to 163 0C in eutectic melts with some evidence of phase separation in the prior crystallisation for the higher two concentrations of polyamide-6,9. The two cases with higher polyamide levels also have small polyamide-6,9 melting peaks reflecting the phase separation that took place during crystallisation. 9.3.1.4


Figure 9-5 DSC thermograms of the second crystallisation at 25 0C/min for pan blended polyamide-6,9/phenothiazine samples.

The thermograms In Figure 9-5 for all three mixtures of pan blended polyamide-6,9/phenothiazine show further refinement from the double peaks of the first crystallisation.

This time, all had an initial small

crystallisation near that of pure polyamide-6,9 and a major peak 30-45 0C below the normal polyamide-6,9 crystallisation peak. The 39PA69PTh has by this time lost 33% of the phenothiazine and the sample could more accurately be termed 48PA69PTh. 261

The small peaks near the polyamide-6,9 melting temperature are showing, the

strong

tendency

for

polyamide-6,9

to

crystallise

out

of

the

polyamide-6,9/phenothiazine solution at all concentrations. It can be seen that that peak for 39PA69PTh is slightly higher than the polyamide-6,9 melting peak.

A slightly higher peak had also been seen with the

63PA69PTh sample on the first crystallisation.

It is considered that this

early crystallisation is due to viscosity reduction with phenothiazine enhancing the ability of the polyamide to crystallise. The peak temperature is very close to 190 0C at this fast cooling rate and it will be seen below that the

crystallisation

temperature

for

polyamide-6,9

under

more

ideal

conditions at slower cooling rates is near 193 0C. The crystallisation here is between the temperatures for slow crystallisation and the lower temperature for the faster rate. 9.3.2 9.3.2.1

Ampoule Material Melting Temperatures (First melt in DSC at 5 0C/min)

The DSC thermograms in Figure 9-6 are the first DSC melting at 5 0C/min of samples blended originally in ampoules.

Figure 9-6 DSC thermograms of polyamide-6,9/phenothiazine blends from ampoules in their first melting in the DSC at 5 0C/min.

262

a) The melting thermogram of ampoule sample 63PA69PTh is similar in form to that of polyamide-6,9 but depressed by 15 0C. There appears to be a weak endotherm near 150 0C of the small amount of phenothiazine dissolving into the polyamide and this is followed by the TLS peak of polyamide-6,9 dissolving into the eutectic. b) The 57PA69PTh sample, with a little more phenothiazine, displays the characteristics seen in earlier chapters of a depressed melting/dissolution peak for the materials giving a saturated eutectic solution. That peak is followed at higher temperatures by a TLS peak covering dissolution of the remaining polyamide.

The very large change between this thermogram

and the 63PA69PTh one could mean that the actual differences in percentage polyamide may well be greater than measured by the TGA samples obtained from next to the DSC samples. The 63PA69PTh may be a slightly higher percentage than 63% and/or the 57PA69PTh may be lower than 57% polyamide. c) 44PA69PTh and 47PA69PTh appear to be minor variations of the 57PA69PTh thermogram.

There are some subtleties making them

different. Firstly, the main peak for both is slightly higher in temperature and has the same temperature as the first peak for 23PA69PTh described below.

The size of the TLS peak is larger for the sample with higher

phenothiazine and, as will be seen below, the size of the second peak for 23PA69PTh is larger again. d) The thermogram for 23PA69PTh has a first peak at exactly the same temperature as the 47PA69PTh and 44PA69PTh samples indicating a similar situation. A second peak just below the melting temperature of phenothiazine is much larger than the TLS peaks of the other two. This shows that the composition of the 23PA69PTh sample is further away in polyamide concentration from saturation at 170 0C than the other two. The ampoule samples here show the first melt/dissolution peak is a depressed eutectic melt, followed by residual polyamide only dissolving with a TLS peak at temperatures higher than 170 0C where polyamide concentration is more than 50%.

There is a TLS peak for remaining

phenothiazine dissolving in the saturated solution when the level of phenothiazine in the sample is higher than for a saturated solution. This 263

thermal behaviour is understandable if there is a phase diagram with a eutectic composition near 50% polyamide.

The further depression of the

eutectic melt temperature for 57PA69PTh is typical of low eutectic temperatures near the eutectic composition has been noted with some other combinations of polyamide/diluent. The lead-in to the first major peak for these ampoule blend samples does not show the metastable form melting and recrystallising seen with some of the pan-blended samples that had been cooled at 25 0C/min. 9.3.2.2


The percentages of polyamide were used with the total enthalpies of the first melting heating ramp to calculate the overall crystallinity of ampoule samples in the same manner as in earlier chapters. The results are plotted below in Figure 9-7.


100 90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 9-7 Overall crystallinity determined from TGA and total first DSC melting enthalpy of polyamide-6,9/phenothiazine.

The overall crystallinity from ampoule material samples, as seen in Figure 9-7, slowly decreases as the percentage of polyamide is increased from the high value for phenothiazine to the much lower one for polyamide-6,9. There is only one sample where the crystallinity is a little lower than for polyamide-6,9 but the series for blends are below the linear relation between the phenothiazine and polyamide-6,9. This means that the blending process is suppressing overall crystallinity to some extent. 264

9.3.2.3

First crystallisation in the DSC at 2 0C/min of ampoule samples

Figure 9-8 DSC thermograms of the first crystallisation in DSC at 2 0C/min of polyamide-6,9/phenothiazine ampoule material.

a) 63PA69PTh has a polyamide-6,9 crystallisation peak depressed 11 0C by the presence of phenothiazine. b) The 57PA69PTh thermogram has a polyamide-6,9/phenothiazine peak depressed 37 0C below the polyamide-6,9 crystallisation temperature and followed by a small (depressed) phenothiazine peak depressed by 24 0C from the phenothiazine crystallisation temperature.

Both of these

depressions show that there are Flory-Huggins style interactions between the materials at the time of crystallisation. c) The thermogram for sample 47PA69PTh has a single peak depressed some 50 0C below the polyamide-6,9 crystallisation temperature and nearly 20 0C below the phenothiazine crystallisation temperature. That peak actually comprises initial crystallisation of phenothiazine that goes over smoothly into the crystallisation of polyamide-6,9/phenothiazine in the latter stages. This can be seen compared to a normal phenothiazine crystallisation in a more expanded version in Appendix A. 265

d) The thermogram for 44PA69PTh has a small peak near the polyamide crystallisation temperature caused by some phase separation.

This is

followed at lower temperatures by a double peak slightly higher than for 47PA69PTh in which excess phenothiazine to the saturated solution crystallises

at

a

temperature

crystallisation temperature.

depressed

from

the

phenothiazine

That peak runs into a second peak of

polyamide-6,9/phenothiazine within the same peak envelope. e) The 23PA69PTh thermogram takes matters a step further with the near complete separation of the two peaks. Those peaks are both at slightly higher temperatures than the previous samples, the first because the level of polyamide-6,9 “contaminating” the phenothiazine is less and the second because the polyamide-6,9 is having more problems in remaining in solution at those temperatures. This set of thermograms shows a maximum depression by 50 0C of the polyamide-6,9 crystallisation peak with 47PA69PTh. This is near equal concentrations of each constituent material.

The maximum depression of

the phenothiazine crystallisation peak is with the 57PA69PTh sample and is by 24 0C. The phenothiazine, on the other hand, has a more well-behaved and consistent depression of the crystallisation temperature with increasing polyamide concentration in a style typical of the Flory-Huggins theory. There is only one sample here with crystallisation at the polyamide crystallisation temperature indicating phase separation. 9.3.2.4


The crystallinity of phenothiazine and polyamide as displayed in Figure 9-9 from the first cooling ramp in the DSC does have some limitations because of the evaporative loss of phenothiazine to varying degrees from the samples. The phenothiazine enthalpy of crystallisation decreases to zero by 63% polyamide.

There is a wider scatter of results in this decrease than was

found with the polyamide/diluent combinations having higher polyamide than diluent melting temperature. The limited results for this combination do make it difficult to have a definitive pronouncement on this point. The crystallinity of the polyamide-6,9 is unusual in that it increases with decreasing polyamide, at least down to 23% polyamide. 266


100 phenothiazine polyamide-6,9

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 9-9 Crystallinity of Phenothiazine and Non-Phenothiazine parts determined from the first crystallisation from the melt of polyamide-6,9/phenothiazine samples that had been taken to the melt in the DSC and cooled at 2 0C/min.

9.3.2.5


230

Liquid

220 210 200 0

Temperature ( C)





190 180


170


160 150

Solid & liquid Solid Liquid & crystallites Solid Solid & liquid


140 130

Solid

120 110 0

10

20


80

90

100

Figure 9-10 Non-equilibrium phase diagrams for polyamide-6,9, phenothiazine and their blends.

Figure 9-10 shows the phase diagram characteristics seen in chapters 3, 7 and 8 of Flory-Huggins crystallisation temperature depressions, dips near 267

the crossover concentration and crystallisation of small amounts of near-pure polyamide at or above that normal for the polyamide. 9.3.2.6

Third Melting (at 5 0C/min) of ampoule materials/Second DSC Melt

The ampoule samples in pans from the first DSC runs were passed through a repeat melt/crystallisation cycle in the DSC, as was done in earlier chapters.

Figure 9-11 below shows the DSC thermograms of the melt

portions of the repeat DSC runs. Some evaporation of phenothiazine has generally taken place between the start of the first and second DSC runs of the ampoule samples. The effect will be greater than in the pan blended samples because of the protracted times spent at high temperatures due to the slower cooling ramp used with these samples.

Broadly, the second melting in the DSC of the ampoule

samples are similar to the first ones: a) The small dip in 63PA69PTh near 150 0C has disappeared and there is an increase in temperature of 3 0C for the main peak, indicative of a small loss in phenothiazine through evaporation.

Figure 9-11 DSC thermograms of the second melt at 5 0C/min in DSC of polyamide-6,9/phenothiazine blends of ampoule material.

268

b) There is a reduction in the peak height, and hence crystallinity, for the main peak of 57PA69PTh. The TLS peak has become more pronounced. Both these are due to some phenothiazine evaporative loss. c) The peak for 47PA69PTh has become slightly larger and is 2 0C higher in temperature. d) The thermogram for 44PA69PTh, as with the 57PA69PTh, has become slightly more pronounced.

There is also a small peak at the

polyamide-6,9 melting temperature.

These changes together perhaps

indicate some phase separation having taken place, particularly when there was a small peak seen at the polyamide crystallisation temperature in the first crystallisation in the DSC. e) The 23PA69PTh thermogram for the second DSC melt has an almost identical first peak to the first time melting in the DSC but the second peak is slightly larger and there is a small peak at the polyamide-6,9 melting temperature. These also point to slight phase separation of some regions into the polyamide-rich and diluent-rich phases. The picture here is similar to the first melting in the DSC but modified slightly by evaporation of phenothiazine seen at lower phenothiazine levels where it is more evident and by some phase separation at higher phenothiazine levels. 9.3.2.7


The DSC thermograms for the second cooling ramp in the DSC at 2 0C/min of polyamide-6,9/phenothiazine samples from ampoules are shown below in Figure 9-12.

The small differences between the first and second DSC

crystallisations are due to phenothiazine evaporation and, at higher phenothiazine concentration, of phase separation.

There is a quite faint

peak in the 23PA69PTh thermogram at the polyamide-6,9 crystallisation temperature consistent with the peak seen at high temperatures during remelting.

269

Figure 9-12 Thermograms of the second crystallisation in the DSC at 2 0C/min of polyamide-6,9/phenothiazine ampoule material.

9.4 Fourier Transform Infra Red Spectroscopy Polyamide-6,9/phenothiazine samples from the ampoules were examined with FTIR in the mid range using photoacoustic techniques and in the Near Infra Red using DRIFT.

These showed no evidence of changes in the

hydrogen bonding status of either polyamide-6,9 or phenothiazine. Spectra may be seen in Appendix D on CD.

9.5 Summary The material combination of polyamide-6,9 and phenothiazine melt blended in pans or in ampoules displays similarities to the polyamide-4,6 and polyamide-6 combinations with phenothiazine in that there is more evidence of phase separation than in the combinations with carbazole. It is not clear at this stage whether this is related to the melting temperature of phenothiazine being lower than the polyamide or to the nature of the phenothiazine molecules. One of the more interesting things found in these trials is that the melting of the powders together in a DSC pan leads to a melt peak depression of 45 0C for the polyamide-6,9 as it initially melts and dissolves phenothiazine up to 270

the point of saturation. There is evidence that the phenothiazine has some difficulty in dissolving easily into the solution.

The temperature must be

close to the phenothiazine melting temperature before the process is complete. This is confirmed to some extent by the tendency of polyamide-6,9 to crystallise as solutions are cooled at fast or slow rates.

There is an

optimal concentration near equal parts of both materials where the bulk of the

crystallisation

takes

place

at

45-50 0C

below

crystallisation temperature, regardless of cooling rate. rate

leads

to

single

peaks

for

the

the

polyamide

The faster cooling crystallisation

of

polyamide-6,9/phenothiazine but the materials tend to crystallise separately at depressed temperatures with a slower cooling rate. The substantial melting and crystallisation temperature depressions show that there are considerable Flory-Huggins style interaction between the two materials. The difficulties in achieving dissolution of the powders in the first instance are thus showing that there are kinetic effects at play in slowing the dissolution. FTIR experiments in the Mid Range IR and in the Near Infra Red both show that hydrogen bond interactions do not play a part in the melting and crystallisation behaviour of these two materials.

271

Chapter 10

POLYAMIDE-6,12 WITH PHENOTHIAZINE CONTENTS 10.1 Introduction

272

10.2 Thermogravimetric Analysis

273

10.3 Differential Scanning Calorimetry

274

10.3.1

Pan Melt Blending

10.3.1.1 10.3.1.2 10.3.1.3 10.3.1.4

10.3.2

Ampoule Material

10.3.2.1 10.3.2.2 10.3.2.3 10.3.2.4 10.3.2.5 10.3.2.6 10.3.2.7

274

0

Melting Temperatures for first heating ramp of the powders at 5 C/min 0 Crystallisation for first cooling ramp of the molten blend at 25 C/min 0 Melting Peak Temperatures for second heating ramp at 5 C/min Crystallisation Peak Temperatures for second cooling at 25 0C/min

274 276 278 280

281

0

Melting Temperatures (First melt in DSC at 5 C/min) Overall Crystallinity 0 DSC Crystallisation Temperatures at 2 C/min for remelted ampoule material. Crystallinity from first crystallisation in the DSC Phase Diagrams for first time heating and cooling ampoule material in DSC Third Melting of materials/Second DSC Melt Third Crystallisation of Materials/Second DSC Crystallisation

281 282 283 286 286 287 288

10.4 Fourier Transform Infra Red Spectroscopy

289

10.5 Summary

289

10.1 Introduction Chapter 10 covers the melt blending of polyamide-6,12 and phenothiazine. It is the last of the chapters dealing with experimental trials on small scale melt blending of powders and their DSC investigation and in the making of bulk materials in ampoules plus their investigation with various techniques. Polyamide-6,12 is an even-even polyamide and has the lowest density of amide groups in the group of four polyamides studied. should be the closest to a polyolefin.

In principle, it

It does display a low melting

temperature consistent with that aspect but the polyamide-6,9 has a slightly lower melting temperature as a result of it being an even-odd polyamide. The

melting

temperature

of

polyamide-6,12

phenothiazine that it is being melt blended with.

272

is

above

that

of

the

This chapter will show thermal behaviour of polyamide-6,12/phenothiazine blends in pans similar to that seen in the previous chapter with polyamide-6,9/phenothiazine blends. There was one exception compared to all other tests done. One sample had no depression of the phenothiazine crystallisation when cooled quickly. The maximum depression of the polyamide crystallisation is much less than seen with polyamide-6,12/carbazole and it occurs at 36% polyamide rather than near 75% polyamide-6,12 with the carbazole. The comparisons will be brought together with much more detail in the General Conclusions chapter and that chapter will cover the comparisons between the different polyamides more closely.

10.2 Thermogravimetric Analysis TGA results for samples taken from ampoules are compared in Figure 10-1 with the expected levels for polyamide from the weights of materials used in the ampoules. The TGA results are those for the plateau level at 300 0C after


heating the samples at 10 0C/min in a nitrogen gas stream. 100 PA612PTh 1:1 line

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Expected polyamide concentration (%wt) Figure 10-1 Actual vs expected weight percentage polyamide in samples from polyamide-6,12/phenothiazine blends from ampoules The plateau levels at 300 0C are compared with the percentages from weights of the constituents added to the ampoules for blending.

Some of the data points in Figure 10-1 for ampoule samples show a divergence between the expected and actual concentrations of polyamide in 273

the samples. These ones are all from early in the trials where the ampoule cooling process had not been fully refined, leading to less consistent polyamide concentration in the ampoule samples. They have been included in the data used for this chapter because they extend the concentration range investigated.

There could potentially be some deviations from the

general behaviour during the first melting in the DSC because their immediate thermal history has differed slightly.

10.3 Differential Scanning Calorimetry Polyamide-6,12 has a melting temperature in the range 216-218 0C with crystallisation in the range 185-197 0C, depending on the cooling rate and as stated in Chapter 6. 10.3.1 Pan Melt Blending 10.3.1.1

Melting Temperatures for first heating ramp of the powders at 5 0C/min

Figure 10-2 DSC thermograms during the first melting at 5 0C/min of polyamide-6,12, phenothiazine and powder mixtures of the two.

The DSC thermograms for the first melting in DSC pans of polyamide-6,12, phenothiazine and three mixtures of the powders at 5 0C/min are shown in Figure 10-2. Similarities will be seen with the polyamide-6,9/phenothiazine combination of Chapter 9. 274

a) The polyamide-6,12 is showing a minor endotherm near 160 0C, and an ill-defined exotherm immediately before the main endotherm which peaks at 216 0C. The minor peaks can be better seen in the Appendix A, along with a horizontal line to differentiate exotherms from endotherms. They represent the effects of the thermal history of the sample. The sample is displaying some of the characteristics of granule manufacture, grinding stress and the later thermal treatment during vacuum drying prior to these trials. b) The thermogram for 71PA612PTh comprises a broad endotherm over the range 160-180 0C followed by a broad, skewed peak near 210 0C. That first endotherm is the melting/dissolution of some of the polyamide-6,12 with phenothiazine. This is followed by the TLS peak of polyamide-6,12 dissolving into the saturated solution as the temperature nearly reaches the normal polyamide melting temperature. c) 60PA612PTh has a thermogram with three peaks, a double peak at 172/183 0C and a very broad weak peak that extends from the double peak to 204 0C. The second peak of the double peak is much weaker and could be a TLS peak of phenothiazine dissolving when considered with the 36PA612PTh thermogram. Alternatively it could possibly be a TLS peak for polyamide-6,12 if raising the temperature above 190 0C places the sample in a part of phase space where polyamide-6,12 has very limited solubility. The very weak broad peak above this is most likely the slow dissolution of remaining polyamide-6,12 into the saturated solution [159]. d) The thermogram for 36PA612PTh has a double peak at 174/184 0C, 1-2 0C higher than the double peak of 60PA612PTh. The relative heights of the peaks in the double peak are in reverse order with the second now far stronger. The higher level of phenothiazine in this sample combined with the lesser amount of polyamide-6,12 points to the first peak for both samples being the dissolution of available polyamide with phenothiazine to form a saturated solution and the second peak being the TLS peak for the consumption of excess phenothiazine. There is a separate small peak near the polyamide-6,12 melting temperature showing that there was some residual polyamide available to melt. 275

The existence of the small

peak may be due the dissolution being slow from reduced solubility of polyamide-6,12 at the higher temperature, as described for 60PA612PTh. The set of thermograms for the first melting of powders in DSC pans is showing

a

similar

behaviour

to

that

seen

with

the

polyamide-6,9/phenothiazine blends with a saturated solution of the two materials forming nearly 40 0C below the polyamide melting temperature and approximately 15 0C below the phenothiazine melting temperature when there is at least 35% phenothiazine in the mixture. Excess phenothiazine will only dissolve in the high temperature solution with a TLS peak at elevated temperatures (up to the phenothiazine melting temperature). Some of the polyamide appears to have difficulty in dissolving when the phenothiazine level is high and will only melt at close to the normal polyamide melting temperature. 10.3.1.2


Figure 10-3 DSC thermogram for first crystallisation of pan blended polyamide-6,12/phenothiazine from the melt at 25 0C/min.

The thermograms in Figure 10-3, for the first cooling at 25 0C/min of polyamide-6,12, phenothiazine powders and their mixtures that had been taken to the melt, are described below: 276

a) Sample 71PA612PTh has a small crystallisation of nearly pure polyamide ahead of the major peak for the slightly depressed crystallisation of polyamide-6,12. It is noteworthy that the trailing edge of the main peak extends down to approximately 150 0C before petering out but no explanation is available for that. b) The

thermogram for the

71PA612PTh, polyamide-6,12.

with

the

60PA612PTh crystallisation

sample of

a

begins, similar

to

small

of

amount

It is only at 40 0C lower that we see the depressed

crystallisation of polyamide-6,12 with phenothiazine in a double peak that changes into the crystallisation of phenothiazine with some polyamide-6,12.

The much less “spiky” form of the latter peak is an

indication that there is a moderate amount of the polyamide crystallising with the phenothiazine. c) The thermogram for the sample 36PA612PTh is unlike the previous two discussed in that there is no crystallisation peak near the normal polyamide-6,12 crystallisation temperature. The thermogram also differs from other crystallisation thermograms seen in previous material combinations in that there is a major phenothiazine crystallisation peak at virtually the same temperature as the pure phenothiazine. In previous thermograms, we have usually seen some depression of major diluent crystallisation by polyamide.

This shows that there is virtually no

interaction between the two materials in this particular case.

The

crystallisation is part of a double peak where the phenothiazine crystallisation changes into a depressed polyamide-6,12 crystallisation as the phenothiazine crystallisation progresses. The only exception to this was with the ampoule samples 25PA6Car and 28PA69Car at its first crystallisation where they phase separated and crystallised 1 – 2 0C lower under the differing cooling rate. This set of thermograms shows some interesting features.

With past

material combinations we have only seen odd cases where the crystallisation of almost pure phenothiazine changes into the depressed crystallisation of polyamide. We have not previously seen the crystallisation of phenothiazine from a polyamide/phenothiazine solution where the crystallisation takes place at the same temperature as for pure phenothiazine. 277

10.3.1.3


The DSC thermograms found in Figure 10-4

for the repeat melting at

5 0C/min of polyamide-6,12/phenothiazine samples melt blended previously in DSC pans are discussed below:

Figure 10-4 DSC thermograms at 5 0C/min for the second melt of polyamide-6,12/phenothiazine materials previously melt blended in pans and crystallised at 25 0C/miin.

a) The polyamide-6,12 thermogram, where the sample has been remelted after being cooled at 25 0C/min, shows the typical shape of melting of the metastable lamellae followed by recrystallisation into the stable form before undergoing full melting of the stable form. The exothermic peak for the crystallisation rises quite noticeably above the baseline of the thermogram, indicating a substantial drop in free energy by the crystallographic reorganisation. b) The

71PA612PTh

sample

undergoes

the

same

processes

polyamide-6,12 but is depressed in temperature because

as

the

of the

phenothiazine present in the sample. The melting/recrystallisation of the metastable form is depressed by 15 0C for this sample but the main melting peak is depressed by less than 10 0C. The tail of the main peak, 278

overlapping the normal melting region for polyamide-6,12, most likely incorporates the melting of polyamide-6,12 seen to crystallise as near pure polyamide in Figure 10-3. c) The 60PA612PTh thermogram shows a treble peak for the melting and dissolution of metastable polyamide-6,12 into a high temperature solution

followed

by

a

TLS

peak

for

the

dissolution

of

excess

phenothiazine or polyamide-6,12 only at further temperature elevation. There is minor melting of residual, nearly pure, polyamide-6,12 at the normal polyamide-6,12 melting temperature, confirming the previous crystallisation seen in Figure 10-3. The TLS peak could possibly be for either material but two factors point towards it being for the polyamide. Firstly, there is generally a nexus between the concentration for crystallisation of both materials at the same time and the eutectic concentration. We will see at Figure 10-5 in the next section that the crystallisation at the same temperature would be found between 60 and 36% polyamide for the next crystallisation of these samples. It will be found close to 36% from the first crystallisation of the ampoule samples in Figure 10-8.

Secondly, the concentration region where the eutectic

concentration for ampoule samples lies is between 30 and 50% polyamide-6,12.

The eutectic concentrations of pan and ampoule

samples were found to be virtually identical for other diluent/polyamide combinations. The connection between eutectic concentration for melting and the concentration where diluent and polyamide crystallise at the same time is also the same for other diluent/polyamide combinations. It therefore seems likely that this TLS peak is for the polyamide because the polyamide concentration is higher than the above ranges. d) The initial melting of metastable polyamide-6,12 in the 36PA612PTh sample and recrystallisation before the main melting is not so clearly seen but can be picked out more clearly in Appendix A.

The main melting

peak is close to that for 60PA612PTh. The process of forming a saturated solution should be nearly the same in both cases, so temperatures of the peaks were close, as expected. Saturated solution formation is followed by a phenothiazine TLS peak.

279

This thermograms display behaviour expected from previous chapters for remelting polyamide/phenothiazine samples that had been cooled rapidly. That sort of melting of metastable crystalline forms of the polyamide and recrystallisation was also seen in some of the polyamide/carbazole blends. We have also seen here the remelting at normal polyamide-6,12 melting temperatures of near pure polyamide that was seen to crystallise at near normal crystallising temperatures for the polyamide.

That melting of the

near-pure polyamide-6,12 lamellae has not happened during the general dissolution of polyamide-6,12/phenothiazine in the eutectic melt so either the polyamide is constrained from melting because it is within a binodal region of phase space or it is resistant to earlier dissolution because it is in a particularly stable lamellar/crystallographic configuration. 10.3.1.4

Crystallisation Peak Temperatures for second cooling at 25 0C/min

Figure 10-5 DSC thermograms of the second crystallisation at 25 0C/min of pan blended polyamide-6,12/phenothiazine samples.

Figure 10-5 depicts the thermograms for the repeat crystallisation cycle at 25 0C/min of polyamide-6,12/phenothiazine samples originally blended in DSC pans from powder mixtures.

They are nearly identical to the first

crystallisations, as expected from earlier chapters. The slight differences are 280

due to the loss phenothiazine in the intervening high temperature heating regimes, resulting in minor polyamide concentration changes in the samples. 10.3.2 Ampoule Material 10.3.2.1

Melting Temperatures (First melt in DSC at 5 0C/min)

The DSC thermograms of eight blends of polyamide-6,12/phenothiazine and the raw materials from ampoule samples during the first heating ramp at 5 0C/min are shown in Figure 10-6.

Figure 10-6 DSC thermograms of polyamide-6,12/phenothiazine ampoule samples first melting in the DSC at 5 0C/min.

It should firstly be noted here that the samples 30, 36 and 58PA612PTh were from early ampoules and deviated in the initial cooling from the way the standard process developed later.

In particular, the sample 29PA612PTh

would have deviated much more than the other two from the standard cooling set-up. The cooling for 29PA612PTh had been done manually which meant that the set-point for the furnace had been set lower by 20 0C every 10 minutes. The result was an undulating cooling ramp averaging 2 0C/min but with steeper sections.

That occurred until the thermal mass of the

furnace, combined with the smaller temperature difference from ambient, slowed the cooling rate right down to a minimal rate. 281

The results are

included here to give a broad picture of the likely thermal responses in the DSC, but obviously with some caution. The large set of thermograms here will be treated together to minimise repetition. The two samples near 80% polyamide have skewed main peaks after a weak endotherm near 165 0C. This weak endotherm can be seen as the process of melting/dissolution of phenothiazine with polyamide-6,12 to form a saturated solution at those temperatures. This is followed by the TLS peak for the remaining polyamide-6,12. The peak temperatures of the remaining samples lie closely together in a 2 0C range near 170 0C. They represent the eutectic melting temperature of the


blends.

Slight

differences

in

the

temperature had also been seen with the pan blended samples. The other curves

are

followed

by

smaller

TLS

peaks

as

samples

with

less

polyamide-6,12 are taken until the TLS peak for polyamide-6,12 disappears near either 50 or 35% polyamide. There are some minor exceptions in the whole set of curves but there is a strong trend to support this notion. We see the lead-in to the main peaks do not show the melting/recrystallising of metastable polyamide. The absence was expected from the slower cooling rates employed with the ampoules. The 29PA612PTh sample has a small peak at the polyamide-6,12 melting temperature but just from this thermogram it is not known if this is due to the slightly different cooling that had occurred with this sample during formation in the ampoule. We will see in the section on crystallisation in the DSC that a small portion of the polyamide crystallises out during cooling. This minor feature of the thermogram is thus not due to the original semicontrolled cooling regime for that sample in the ampoule but due to the temperature/concentration conditions the sample is experiencing. 10.3.2.2


The overall crystallinity is plotted in Figure 10-7 for the series of ampoule samples taken to the melt in the DSC. We see a generally decreasing level of crystallinity for higher and higher concentrations of polyamide-6,12 in the figure.

282


100 90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 10-7 Overall crystallinity determined from the first DSC melting enthalpy of polyamide-6,12/phenothiazine ampoule samples and TGA vs weight percentage polyamide from TGA plateau level at 300 0C.

The level of crystallinity at approximately 80% polyamide actually dips slightly below that of the polyamide.

The lack of a linear relationship

proportional to concentration shows that the crystallinity of blends is being suppressed overall by the melt blending process. 10.3.2.3


Figure 10-8 shows the thermograms of the crystallisation of material meltblended in ampoules, taken to the melt in the DSC and then crystallised for the first time in the DSC at 2 0C/min. This set of DSC thermograms for the first cooling ramp of the ampoule samples in the DSC will be treated as a whole, as with the first melting of ampoule samples in the DSC.

The alignment of the thermograms from

highest downwards to lowest concentration of polyamide in the samples makes it clear that there are several processes taking place in the series of thermograms. The 29PA612PTh thermogram is the only one to have crystallisation of nearly pure polyamide-6,12 with the exception of the 55PA612PTh sample having a very faint broad exotherm in that region. The exotherms are due to

283

a small amount of the high temperature solution beginning to phase separate. There is an increased depression of the polyamide crystallisation with lower polyamide levels in the blend. That occurs down to the 36PA612PTh sample, where the crystallisation starts as a phenothiazine crystallisation and changes into (highly depressed) crystallisation of the polyamide-6,12 within the same peak.

Figure 10-8 DSC thermograms of the first crystallisation in the DSC at 2 0C/min of polyamide-6,12/phenothiazine ampoule material.

The 29PA612PTh sample has a polyamide peak at higher temperatures than the 36PA612PTh sample so there appears to be a maximum in the polyamide-6,12 crystallisation temperature depression, as was found with ampoule samples for polyamide-6,9/phenothiazine in Chapter 9.

A

minimum of the polyamide crystallisation was not found with the other polyamide/diluent

combinations,

possibly,

in

some

cases

because

sufficiently low polyamide concentrations were not explored and in others because the diluent crystallised first. The peak near 155 0C for 29PA612PTh incorporates a small sub-peak on the leading edge but the form is Gaussian 284

rather than the “spiky” peak of phenothiazine crystallisation. The reason for that shoulder peak is unknown at this stage. “Spiky” peaks for the crystallisation of quite pure phenothiazine can be seen depressed progressively less as the level of polyamide is reduced. At 58% polyamide-6,12 there is strong crystallisation of phenothiazine depressed by 18 0C from the normal phenothiazine crystallisation temperature.

The

temperature that the crystallisation occurs at and the size of the peak increase as the level of polyamide is reduced further until there is a 10 0C depression for the 36PA612PTh sample. polyamide

does

actually

have

a

The last thermogram, at 29%

slightly

higher

depression

of

the

phenothiazine crystallisation by 2 0C. There are also small secondary phenothiazine crystallisations at lower temperatures than the main phenothiazine crystallisation peaks with this polyamide/diluent combination.

They are the crystallisation of residual

domains of phenothiazine as far down as 120 0C, some 45 0C below the normal phenothiazine crystallisation temperature under these cooling rates. The amounts of material involved in these cases are of the order of 5 to 10 µg in a 10 mg sample. The secondary crystallisations are from small pockets where the mobility of the phenothiazine through the thickening amorphous polyamide is still sufficiently high that a few molecules can group. They can then crystallise into a phenothiazine micro-domain before the viscosity drops too far and they become individually trapped in the amorphous region. The large number of ampoule samples spanning a wide range in polyamide concentration in this data give a good opportunity to see overall trends taking place. The generally increased depression of polyamide crystallisation with increased concentration of phenothiazine and the increased depression of the main phenothiazine crystallisation temperature with increased polyamide concentration can quite clearly be seen and both show the FloryHuggins style interaction between the materials. Interest in the different behaviour of the 29PA612PTh sample with respect to the others should be tempered by the knowledge that the ampoule was originally made in a different manner to the others. This different behaviour would have to be confirmed by other ampoule samples made with similar concentrations of polyamide-6,12 under standard furnace conditions. 285

10.3.2.4


The crystallinity of phenothiazine and polyamide is displayed in Figure 10-9 from the first cooling ramp in the DSC.


100 phenothiazine polyamide-6,12

90 80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

70

80

90

100

Polyamide concentration (%wt) Figure 10-9 Crystallinity of Phenothiazine and Non-Phenothiazine parts from the first crystallisation of polyamide-6,12/phenothiazine ampoule samples from the melt at 2 0C/min in the DSC versus weight percentage of polyamide-6,12.

The crystallinity of the phenothiazine, as with the crystallisation depression, is very variable but reduces to zero by 80% polyamide-6,12. It is another example of the scatter in the phenothiazine enthalpy of crystallisation seen with polyamide/diluent combinations having a higher polyamide than diluent melting temperature. The crystallinity of the non-phenothiazine portion gradually increases to 40% as the level of polyamide is reduced. 10.3.2.5


The phase diagrams in Figure 10-10 are similar to those of previous chapters where the polyamide normally crystallises at higher temperatures than the diluent. The only difference here is that there is an upturn at low polyamide concentration so that the polyamide crystallises before the diluent below

35% wt

of

polyamide.

The

first

temperature implying a eutectic. 286

melting

is

at

near-constant

230 220

Liquid


210

Liquid & crystallites

Temperature (0C)

200 190


Liquid & crystallites

180 170

Solid & liquid

160

Solid & liquid Liquid & crystallites

150 140


130

Solid Liquid



Solid

120 110 0

10

20


80

Figure 10-10 Non-equilibrium phase diagrams for polyamide-6,12, phenothiazine and their blends with Flory-Huggins style melt/crystallisation depressions.

10.3.2.6


Figure 10-11 DSC thermograms of the second melt at 5 0C/min in the DSC of polyamide-6,12/phenothiazine ampoule material.

287

90

100

The ampoule samples in pans from the first DSC runs were passed through a repeat melt/crystallisation cycle in the DSC as was done in earlier chapters. Figure 10-11 below shows the DSC thermograms of the melting portions of the repeat DSC runs made at 5 0C/min heating rate for ampoule samples. Generally, the trends in this set of thermograms are similar to those seen in the first melt. The differences that can be seen can be attributed to the loss of some phenothiazine due to evaporation. This can be seen, for example, with some smaller main peaks and larger TLS peaks, indicating a move in the direction of increased polyamide concentration in the samples. There is some reduction in crystallinity with a few of the samples, most notably the 29PA612PTh sample. That sample had originated from early ampoule trials. Differences from the first melting may not be fully representative because we are now looking at the crystallinity embodied from a proper 2 0C/min cooling rather than the crystalline state after a partly uncontrolled cooling. 10.3.2.7


Figure 10-12 Thermograms of the second crystallisation at 2 0C/min in a DSC of polyamide-6,12/phenothiazine ampoule material.

288

The set of thermograms seen in Figure 10-12 for the second DSC crystallisation of ampoule samples of polyamide-6,12/phenothiazine blends are generally similar to the set in Figure 10-8 for the first crystallisation. The small differences that can be seen are due to the evaporation of phenothiazine. The exception to this is the 29PA612PTh sample where the phenothiazine crystallisation peak has disappeared and the main peak has moved

higher

by

15 0C.

The

crystallisation

peak

of

nearly

pure

polyamide-6,12 remains. Again, with this material combination we see some very small spiky peaks below the main peaks where extremely small amounts of phenothiazine are crystallising in ampoule samples.

10.4 Fourier Transform Infra Red Spectroscopy Samples from all the ampoules made with polyamide-6,12/phenothiazine blends underwent FTIR with photoacoustic detection in the mid range IR and again with DRIFT in the Near Infra Red region. In no case was there any evidence of hydrogen bond interactions between the polyamide-6,12 and the phenothiazine. Results may be seen in Appendix D on CD.

10.5 Summary There are similarities with the polyamide-6,9/phenothiazine blends made in DSC pans during the first melting. Melting/dissolution of polyamide-6,12 occurs here in a saturated solution with phenothiazine approximately 40 0C below the polyamide melting temperature. This is followed by the dissolution of excess polyamide or phenothiazine with a TLS peak at temperatures below the respective normal melting temperatures.

Some of the polyamide-6,12

appears to have difficulty in dissolving into the saturated high temperature solution until the normal polyamide melting temperature is reached as with the polyamide-6,9. This is confirmed during crystallisation by the tendency of polyamide-6,12 to crystallise out when cooled at 25 0C/min. One aspect found with the fast crystallisation in pans with this material combination, and not found elsewhere in the thesis, is the crystallisation of phenothiazine

in

a

blend

(36PA612PTh)

without

temperature depression for the phenothiazine.

any

crystallisation

That is indicating a less

favourable interaction between polyamide-6,12 and phenothiazine under those conditions. 289

The remelting of those samples crystallised quickly shows an initial melting/recrystallisation of the metastable form of the polyamide into the stable form prior to the main melt. That behaviour has been seen previously with other polyamides in blends with both carbazole and with phenothiazine. Blends made in DSC pans and remelted showed temperature limited solubility of phenothiazine in mixtures with 60% polyamide and less. The range of ampoule samples available for polyamide-6,12/phenothiazine blends allowed clear trends to be seen in the melting temperature limited solubility peaks and crystallisation temperature depression of the samples. In

particular,

the

depression

of

polyamide

crystallisation

by

the

phenothiazine and the depression of phenothiazine crystallisation by the polyamide-6,12 was very clear. The behaviour of polyamide-6,12/phenothiazine blends differs from that with polyamide-6,12/carbazole blends in that the maximum depression of polyamide crystallisation is now 40 0C rather than the 60 0C seen with carbazole. In this case, the maximum depression is seen at 36% polyamide rather than near 75% polyamide. The

melt

blending

of

polyamide-6,12

with

phenothiazine

has

been

investigated by FTIR in two spectral regions with differing detection systems and found not to involve hydrogen bond interactions between the two materials. This is the same result as has been found for all combinations of the four polyamides studied with either carbazole or phenothiazine. The similarities and differences between the blending of each of the four polyamides with carbazole and with phenothiazine will be covered fully in the chapter on General Conclusions that follows.

290

Chapter 11

GENERAL CONCLUSIONS CONTENTS 11.1 Thesis arguments

291

11.1.1

Effects of relative crystallisation temperatures

291

11.1.2

Effects of polyamide repeat unit

298

11.2 Overall common relationships

300

11.3 Hydrogen bonding

300

11.4 Other areas of potential study for future workers

302

11.5 Practical implications of the work

303

11.6 Summary of conclusions

303

The thesis has explored the melt blending of four different aliphatic polyamides with each of two small molecule organic diluents. In conjunction with this, their solidification at different cooling rates were investigated. The polyamides studied were polyamide-4,6, polyamide-6, polyamide-6,9 and polyamide-6,12, representative of a range of amide densities and polyamide types.

The resulting combinations of melting temperatures of the

polyamides and the diluents, carbazole and phenothiazine, allow conclusions to be made about the effects of relatively mobile diluent molecules on the crystallisation processes.

The range of polyamides studied also allows

further conclusions to be made about the effects of polyamide repeat structure on eutectic compositions and maximum crystallisation and melting depressions.

11.1 Thesis arguments 11.1.1 Effects of relative crystallisation temperatures The main differences in morphology resulting from crystallisation processes at the relatively slow cooling rate of 2 0C/min are related to whether the crystallisation temperature of the polyamide is higher or lower than the diluent.

The combinations PA46PTh, PA46Car, PA6PTh, PA612PTh and 291

PA69PTh all have a polyamide crystallisation temperature greater than the diluent to varying degrees. In contrast, the group PA6Car, PA612Car and PA69Car have a polyamide crystallisation temperature lower than the diluent. The

group

with

polyamide

crystallising

above

the

diluent

are

all

characterised by crystallisation depression phase diagrams demonstrating: a)

Some possibility of a small amount of crystallisation within 10-15 0C of the normal polyamide crystallisation temperature.

b)

Major

near-Gaussian

crystallisation

peaks

at

increasingly

lower

temperatures as the level of polyamide in the blend is reduced. c)

Major “spiky” style crystallisation peaks for the diluent at increasingly lower temperatures as the level of the diluent in the blend is reduced.

These characteristics can be seen for the first DSC crystallisations of ampoule material in Figures 3-13, 3-19, 7-8, 8-8, 9-8 and 10-8 plus their second crystallisations in the DSC and stand out more clearly in the relevant phase diagrams in Figures 3-22, 7-10, 8-10, 9-10 and 10-10 where transition temperatures are plotted directly against concentration. In some cases there is a minor variant in the polyamide crystallisation, where the depression of the peak is greatest at the point where polyamide and diluent crystallise at virtually the same time. This polyamide crystallisation temperature

depression

concentrations.

is

often

not

so

great

at

lower

polyamide

Examples of a minimum in polyamide crystallisation

temperature for the two materials crystallising at the same time are seen in Figures 9-8 and 10-8 or the phase diagrams Figures 9-10 and 10-10. These characteristics can also be seen in samples melt blended in pans and cooled at 25 0C/min, although more crystallisation occurred near the polyamide crystallisation temperature. This pattern can clearly be seen in Figures 3-6, 7-3, 8-3, and 10-3. The pan blended samples for PA69PTh are not quite so obviously of the same mould due to large evaporative loss with 39PA69PTh and some other minor differences. This increased crystallisation at the polyamide crystallisation temperature is due to the lowered viscosity combined with rapid cooling giving a deeper supercooling to drive the crystallisation.

This temporarily allowed more crystallisation before the

increase in viscosity at low temperatures brings crystallisation to a halt. 292

The melting behaviour for this group have phase diagrams that display nearly constant first transition temperatures as polyamide concentration is varied and appear to be eutectic melting. The above is in stark contrast to the crystallisation characteristics for the group with a polyamide crystallisation temperature less than that of the diluent.

The crystallisation for PA6Car, PA69Car and PA612Car was

characterised by eutectic types of phase diagrams, Figures 4-13, 5-10 and 6-10, showing: a) Little depression in the crystallisation temperature of the diluent until polyamide concentration was raised to 20-30% and then dropping rapidly. There was no crystallisation of diluent at temperatures below that of the polyamide. b) There was almost no crystallisation at the polyamide crystallisation temperature. c)

There was a lower Gaussian peak at a virtually constant temperature, although this was sometimes slightly lower at the concentration where the diluent was crystallising concurrently.

There was no evidence during crystallisation of a peak at eutectic temperatures for polyamide levels higher than the eutectic concentration. The further depression at the eutectic concentration appeared to be of the same nature that Berghmans noted [135 p. 219]

(apparent eutectic

temperature depressed by non-equilibrium conditions). Results from the fast cooling used for pan blended samples were consistent with the above results except that a very small amount of crystallisation was often encountered at the polyamide crystallisation temperature when the diluent concentration was very high. Melting transitions for these ampoule samples can be seen in the phase diagrams to be virtually the same as the crystallisation transitions but occur at 10-25 0C higher temperatures Further evidence of marked differences between the two groups is in the enthalpy of crystallisation for blends, in regards to the calculated “crystallinity” of the diluent. It can be seen in Figures 3-20, 7-9, 8-9, 9-9 and 10-9 that there were high values of diluent crystallinity at certain 293

concentration values and low values only a few percent different in concentration, producing a scattered result for diluent crystallinity. The corresponding figures for PA6Car, PA69Car and PA612Car (Figures 4-12, 5-9 and 6-9) follow a much more regular pattern with a linear drop-off in diluent crystallinity from the very high values for pure diluent to zero at some concentration. This linear drop-off appears similar to those found in van der Heijden’s thesis [7]. It is difficult to compare these results due to the differences in focus. He dealt with liquid-liquid phase separation and with an amorphous polymer whereas I used a true monotectic with liquid-solid phase separation, although semicrystalline polyamides are obviously also semi-amorphous. The above evidence shows that different morphologies will result from quite different crystallisation processes taking place depending on the relative crystallisation temperatures of polyamide and diluent.

The reasons for the

two sets of morphological behaviour are linked to the very different molecular condition of the two types of material. Firstly, we should look at the crystallisation of the pure materials.

The

diluent by itself will normally lose entropy sharply as the small molecules in the molten material rapidly stack themselves into crystals with little hindrance from other molecules.

Being small molecules, they have a low

viscosity, particularly in the energetic state at the elevated temperatures involved in this study. In contrast, polymers are severely restricted in their ability to crystallise because the long chains must act co-operatively under strong driving forces in order to achieve partial crystallinity. The Olmsted group [30] concluded that the formation of crystallites only takes place after precursor liquid-liquid decomposition of denser, more organised, domains. Thus, slower crystallisation only takes place at much larger supercooling than experienced with small molecules. Crystallisation of polymer chains is also hindered by the limited mobility of the surrounding and entwined non-crystallising chains, which rapidly lose their mobility as the temperature is reduced. It is important to remember that even highly crystallisable polymers are semi-amorphous and this research dealt with a two-phase system in the solid state when considering just the straight polymers. This issue is discussed later in further detail. 294

The study here concerns polyamides that have considerable hydrogen bond linkages.

The effects for the solid material are well known: high melting

temperatures caused by the lamellae being linked more solidly together at a chain-to-chain level and the surrounding amorphous material not dropping in viscosity so quickly as the temperature is raised to the melt.

The

situation above the melt and for cooling is also strongly affected by the hydrogen bonding.

There is a high viscosity just above the melting

temperature because of the linkages forming and reforming, even if the residence time of the bonds is short. The chains are being pulled strongly by the hydrogen bonds into crystalline lamellae as the temperature drops to the crystallisation temperature. They are restricted more in that action if there is a partial mismatch requiring contortions of the chain backbone to line up the hydrogen bonds (resulting in lower crystallisation temperatures as with polyamide-6,9).

The viscosity of the amorphous melt is high and rapidly

increasing as the temperature is reduced. Consider now the situation with a molten blend having two dissimilar crystallisable polymers. There are different possible outcomes. We may have liquid-liquid phase separation before one and then the other type of domain crystallises. Otherwise we will have the higher melting polymer attempting to crystallise first as the temperature of the high temperature solution is reduced if the polymers are sufficiently compatible in the melt.

The first

material crystallising will attempt to expel the other from the lamellae into the interlamellar region and will crystallise at a lower temperature than normal (higher driving force required). The second crystallising polymer will be in a non-crystallising state at that time and will act as a highly viscous diluent having problems in ‘moving out of the way’ because of the co-operation needed in moving the long chains.

Eventually the second

material will crystallise in the interlamellar space or outside the spherulites if it has been able to move far enough away [160]. This sets the framework for a description of the two cases being studied here where there is either a polymer with high crystallisation temperature and diluent with lower crystallisation temperature or the reverse. Three characteristics of the diluent are that it is highly crystallisable, it will be highly mobile and that it will be either partially compatible or 295

non-compatible with the polyamide at a molecular level. It will initially act as a plasticiser of the amorphous portion of the semicrystalline solid at low diluent concentrations and become a solvent at high concentrations. We will now look at the expected crystallisation process in different circumstances: a)

Polyamide crystallisation temperature higher than diluent and a high polyamide level. Some samples exhibit only a little crystallisation at the normal polyamide crystallisation temperature.

The dynamic cooling

used means that there is interplay between the increased supercooling and the reduction in mobility of the diluted polyamide at lower temperatures. The limited or no crystallisation in these situations is related to the kinetics involved. The presence of some crystallisation at the polyamide crystallisation temperature could be due to either of two possibilities. The viscosity of the normally viscous polyamide melt may be reduced by the diluent, allowing easier crystallisation of the polymer chains despite the presence of “foreign” material. Alternatively, there could be some minor phase separation beginning between the two materials, driving the polyamide into a condition more amenable for crystallisation. This would be in a manner similar to the single-polymer phase separation in the melt prior to actual crystallisation of Olmsted et al [30].

There are some cases such as in Figure 3-19 where

crystallisation temperature has been raised and is possible even above the normal polyamide crystallisation temperature. In this group of polyamide/diluent combinations we see depression of both polyamide and diluent by each other in the manner of a FloryHuggins interaction during crystallisation. We have dilution in solution as the major factor suppressing the (possibly second) polyamide crystallisation at high polyamide levels and temperatures well below the normal polyamide crystallisation. The formation of polyamide lamellae in the amorphous polyamide/diluent at temperatures possibly well above the diluent melting temperature may lead to complete expulsion of diluent from the interlamellar region due to its high mobility. Alternatively, some of the diluent may still reside in the interlamellar regions and later lead to a very retarded crystallisation of small

296

amounts of diluent, or no crystallisation. The remaining diluent would reside in the amorphous polyamide/diluent at lower temperatures. We see little or no crystallisation of diluent at polyamide levels greater than 75% polyamide.

Here, the polyamide crystallisation is being

slightly delayed and, in the cases of polyamide-6, polyamide-6,9 and polyamide-6,12 with phenothiazine, the crystallinity of the polyamide portion of the blend is enhanced.

This is possibly due to a lowered

viscosity of the melt by the diluent at these lower temperatures, leading to easier alignment of polymer chains into a crystallisable condition under the high driving force of greater supercooling. The reason for the apparent randomness of diluent crystallinity has no clear explanation. b)

Polyamide crystallisation temperature higher than diluent and a high diluent level. The comments regarding crystallisation at the polyamide crystallisation temperature are also applicable here. Slightly suppressed crystallisation of the diluent takes place first (if there has been no minor crystallisation at the polyamide crystallisation temperature).

This is due to the presence of a few polymer chains

restricting the ability of the diluent to crystallise. crystallises later amongst the diluent crystallites.

The polyamide

The reason for the

slightly lesser crystallisation depression at polyamide concentrations below the point where both crystallise together is not fully understood. It may possibly be related to the above earlier diluent crystallisation shifting the concentration of polyamide in the remaining material to higher levels. c)

Polyamide crystallisation temperature lower than diluent and a high polyamide level.

The only guidance on slow crystallisation for these

conditions is the one sample, 89PA6Car. That sample crystallised towards the high end of a normal polyamide-6 crystallisation. At that point the small amount of diluent in the sample was already more than 30 0C below its normal crystallisation temperature. d)

Polyamide crystallisation temperature lower than diluent and a

high

diluent level. Blends in this category have all the hallmarks of concentration dependence of the diluent crystallisation for being a liquid-solid phase separation, ie. monotectic. This process is followed at 297

a relatively constant temperature by a eutectic-like crystallisation. In this case, excess diluent to the eutectic solution is rapidly crystallised, the eutectic reaction takes place at the eutectic temperature and remaining diluent-infused amorphous polyamide solidifies.

The low

temperatures of this liquid at the eutectic temperature and the large supercooling of the diluent probably mean that some diluent crystallises at the same time as the polyamide, and most likely in the interlamellar regions. We have noted in the various chapters that crystallinity of the polyamide often appears to be enhanced for this group in the region where the diluent is in excess. Part of that excess may well be the concurrent crystallisation of some diluent. The almost linear drop in diluent enthalpy of crystallisation with increasing polyamide concentration has some parallels to the results of van der Heijden, although he worked with a non-crystallisable polymer. Zero crystallinity is reached at the Berghmans Point where liquid-liquid phase separation meets the Kelley-Bueche line for glass transition with dilution.

The situation here is liquid-solid phase separation with a

crystallisable polymer. The major mediator in the morphology of polyamide/diluent melt blends with carbazole or phenothiazine as the diluent is the relative crystallisation temperatures of the two components. Cases with the polyamide having a greater crystallisation temperature lead to a combination of Flory-Huggins style of crystallisation depressions for each material by the other with possible crystallisation of some polyamide before this.

Systems with a

higher diluent crystallisation temperature and excess diluent lead to liquid-solid phase separation of the diluent followed by a eutectic-like crystallisation of polyamide (perhaps with some diluent), leaving residual diluent uncrystallised in the amorphous remainder. 11.1.2 Effects of polyamide repeat unit Three of the four polyamides studied, polyamide-6, polyamide-6,9 and polyamide-6,12 have melting temperatures within a 12 0C range in between the melting temperatures of carbazole and phenothiazine but have quite different repeat unit structures. One is a polyamide-n type and the others are even-odd and even-even polyamide-m,n types, respectively. 298

There are

some differences between these three polyamides regarding melting and crystallisation depressions and concentrations. The melting temperature depression for ampoule samples is 50 0C for PA6PTh but 29 0C for PA69PTh with respect to the polyamide and 22 0C for PA612PTh but 14 0C for PA6PTh and PA69PTh with respect to the phenothiazine. The

best

estimate

of

molar

concentration

for the

eutectic

melting

temperature varies for these ampoule samples from 46% polyamide for PA6PTh to 30% polyamide for PA612PTh. During the first DSC crystallisation of ampoule samples we found that there was a slight amount of crystallisation at the polyamide crystallisation temperature for PA6Car but none for PA69Car or PA612Car. For those types, the eutectic-like crystallisation peaks are depressed by 11 0C for PA69Car but by 21 0C for PA612Car. The weight percentage polyamide for eutectic crystallisation of ampoule samples varies between 75% for PA6Car to 70% for PA612Car and this becomes 81 and 56% respectively when calculated on a molar basis. The closest approximation to the work here regarding the effects of polyamide type is perhaps the research of Kim, Char and Kim [10 Fig.4]. They

studied,

amongst

other

things,

PEG

with

a

400 Dalton

poly(ethyleneglycol) melt blended with polyamide-6, polyamide-11 and polyamide-12. Their work uses a polymer as a diluent, and one that does have

hydrogen bonding with the

polyamides. They had the

(even)

polyamide-12 with longer aliphatic chains between the amide groups where, here, we had the even-even polyamide-6,12 with longer chains between some of the amide groups. They also used the (odd) polyamide-11 (melting at a lower temperature than polyamide-12 because of the less favourable chain configuration polyamide-6,9. order

for

hydrogen

bonding),

similar

to

the

situation

for

In their case the liquid-liquid phase separation is in the

polyamide-11,

polyamide-12,

and

polyamide-6

with

increasing

polyamide level whilst the best estimates of eutectic compositions for the three here are greater than 64% for polyamide-6,9, 70% for polyamide-6,12 and 75% for polyamide-6.

The comparison could well be there but the

approximations are large. 299

There is a wide range of these types of differences between the behaviour of the different polyamides blended with either carbazole or phenothiazine and cooled at a slow or fast rate. The differences go further than what could be expected from the slight melting temperature differences.

It is concluded

that they are related to the polyamide repeat unit structure, although the mechanisms are not currently understood.

11.2 Overall common relationships Generally, there were interesting features in the first melting of powders for pan blending where features were washed out and often displayed quite a different thermal behaviour on a second melting of the material once it had been crystallised from the melt. Sometimes unusual peaks occurred in this first melting which points to areas of incompatibility between the materials. The dissolution of powder grains in contact with other powder grains or a thin film of liquid was often influenced by the kinetics of dissolution. The polymer grains being bulk particles with dual amorphous and crystalline characteristics further complicated this dissolution. Melting of all material combinations (after previously crystallising from the melt) appears to generally take the form of a classic, simple eutectic melt. Levels of polyamide below 20% have not been explored here and there is little data for blends at over 80% polyamide.

Often a peak at the eutectic

temperature is missing or faint at polyamide concentrations over 70%. In all cases there is a dip in the relationship between overall crystallinity (from the first DSC melting of ampoule samples) and polyamide concentration, meaning that there is negative enthalpy of mixing for blends. It is also noteworthy that there seems to be a very approximate correlation across all material combinations that the greater the diluent crystallisation temperature

is

than

the

polyamide

temperature,

the

higher

the

concentration where both materials crystallise together. This runs from near 0% for PA46PTh at –111 0C to 70-75% for PA6Car and PA612Car near +45 0C excess diluent crystallisation temperature. Results for PA69Car are indeterminate due to a paucity of results but appear to be greater than 64%.

11.3 Hydrogen bonding The particular diluents had originally been chosen as potential model polyamide-polyamide hydrogen bond disruptors. The finding is that these 300

two hydrogen bond donors do not interact with the carbonyl acceptors in the polyamides. Fourier

Transform

Infra-Red

(FTIR)

was

carried

out

with

sensitive

photoacoustic spectroscopy in the mid range IR on all blend samples, leaving them in their original morphological state. Expectations from the literature on hydrogen bond complexation are that N-H peak values would shift by 15 to 30 cm-1. “Free” N-H peaks, such as normally with the diluents, should shift to lower wavenumbers as the hydrogen bond interactions take place. Peaks for “bound” N-H should move to higher wavenumbers as the diluents form hydrogen bonds with carbonyl oxygen atoms, displacing the normal amide N-H to O bonds. The outcome should have been a move of the blend N-H peaks towards each other. In no blend sample was there a shift in peak frequency greater than 3 cm-1.

Often slight shifts seen with the peaks

moving away from each other compared with simple summation of the spectra.

This can be inspected for each material combination in the PDF

files on the attached CD displaying actual spectra and the addition result from the spectra of the constituent materials. Near Infra-Red experiments were carried out on all blends using the DRIFT technique in sensitive hydrogen bonding overtone regions.

The peaks

chosen were those close to the ones identified by Wu and Siesler [141] in their work on hydrogen bonding with polyamide-11.

Spectra had been

manually displaced vertically in Omnic software to check the match between the blend and its polyamide equivalent peak for each peak position.

The

original spectral overlays of NIR measurements on blends and constituent materials can be seen in the PDF files on CD. All peaks in the hydrogen bonding region of interest on all blend samples coincided exactly with the overtone peaks of the relevant polyamide. The lack of N-H shifts for polyamide and diluent peaks in the Mid-range IR using the photoacoustic FTIR technique leads us to the conclusion that any interactions between the two materials on a molecular level do not involve hydrogen bonding between them. This was vindicated by using a different FTIR technique in the Near Infra-Red and finding that sensitive hydrogen bonding overtone bands of blends were unchanged from those of the raw materials treated in the same manner. 301

There is FTIR peak height information in the Mid-range IR band 1700 to 400 cm-1 that is ambiguous because of the non-linear response of photoacoustic measurements, making it difficult to assess whether there were actual differences in infra-red absorption bands or not.

Added to this was the

observation of very bright pink colours in some samples and pink “freckles” visible under a binocularscope in pink hued samples. Together these could mean that pi-conjugation or charge transfer processes are taking place due to non-hydrogen bond interactions between the diluent and polyamide, perhaps with some phase separation. It is suggested that future researchers in this area may wish to pursue UV-visible spectroscopy and other techniques such as dielectric spectral analysis in addition to far more extensive NMR to investigate these aspects with a view to commercial applications.

11.4 Other areas of potential study for future workers This work has covered a broad range of melting and crystallisation temperatures concentrations.

of

the

various

material

combinations

and

polyamide

The use of dynamic DSC crystallisation rather than

isothermal crystallisation has been appropriate for the scale of the investigation.

Other researchers may desire to look at specific material

combinations using isothermal crystallisation because it is easier to treat the results from a theoretical perspective.

The slow cooling experiments will

have much in common, morphologically, with the “two step” isothermal crystallisation used by some authors [6, 160] because the crystallisations of the blend are staged. The fast cooling work here should approximate their “one step” crystallisations. The fast cooling here may, in fact, elucidate the order of crystallisation in their work in cases where the materials crystallise almost simultaneously. Polyamide-6,6 had initially not been studied because the even-even polyamide-4,6 and polyamide-6,12 were already being studied and suitable samples uncontaminated with additives (fire retardant, processing, etc.) had not been located. Polyamide-6,6 has a melting temperature close to 260 0C which means that, based on the behaviour of the polyamides studied, crystallisation would be in the range 16-19 0C lower for cooling at 2 0C/min and 31-36 0C lower at 25 0C/min. This puts its crystallisation temperature 302

approximately in the range 251-233 0C depending on cooling rate.

These

two values straddle the known crystallisation of carbazole at 237-233 0C at the

same

cooling

rates.

Experiments

by

other

researchers

with

polyamide-6,6 and carbazole could use cooling rate to actively drive crystallisation into one morphology or the other. A lower molecular weight polyamide-6,6 could be used if the polyamide crystallisation temperature is too high or, alternatively, one of polyamides –4,8, -8,4 or –8,6 may be suitable, based on melting temperatures from Jones, Aitken and Hill [96]. The use of DSC as the only technique for investigating the morphology does have limitations because it is only “seeing” the crystalline portion of the sample.

Temperature modulated DSC during cooling would be more

appropriate for thorough investigation of glass transition temperatures of highly crystalline materials and for liquid-liquid demixing.

Simultaneous

Small Angle X-ray Scattering with Wide Angle X-ray Diffraction monitoring from the melt down to room temperature would allow investigations into the formation of the rich morphology that has been found.

A further

understanding of Brill transitions taking place and other crystallographic changes could also be studied at the same time.

11.5 Practical implications of the work The most likely practical implications of the work are considered to be specialised membrane formation, drug delivery (and stability) and in melt processing for some specialised applications.

11.6 Summary of conclusions The research in this project has led to increased understanding in areas of semicrystalline/crystalline

binary

polymer-diluent

systems

and

the

implications of relative melting temperatures of the components on the morphology of melt blends. It has given insight into the effects of polyamide repeat units (as distinct from melting temperature) on the behaviour the blends with carbazole or phenothiazine.

It has also shown that any

interactions there may be between the specific diluents and any of the four polyamides chosen do not involve the replacement of polyamide-polyamide hydrogen bonds with diluent-polyamide ones.

303

Appendix A

FURTHER DETAIL FROM DSC THERMOGRAMS There are a number of individual thermograms that show specific details not sufficiently evident in the combined thermograms. These specific thermograms have been reproduced in an expanded form to highlight the subtleties.

A.1 Chapter 5 polyamide-6,9 with carbazole Section 5.3.2.6 Third Crystallisation of Materials/Second DSC Crystallisation at 2 0C/min.

Figure A-1 Crystallisation of a small amount of pure polyamide-6,9 at 190 0C can be seen in the second DSC crystallisation of the ampoule sample 28PA69Car.

Evidence for a minimal amount of pure polyamide-6,9 can be seen to crystallise for the ampoule sample 28PA69Car at 190 0C in Figure A-1.

304

A.2 Chapter 6 polyamide-6,12 with carbazole Section 6.3.2.3 DSC Crystallisation Temperatures at 2 0C/min for remelted ampoule material. The nearly concurrent crystallisation of polyamide-6,12 and carbazole for the ampoule sample 70PA612Car in its first DSC crystallisation can be seen in Figure A-2.

Figure A-2 DSC thermogram of the first crystallisation at 2 0C/min in the DSC of 70PA612Car ampoule sample showing a slow start to the crystallisation followed by a near vertical section of carbazole crystallising and followed by the crystallisation of polyamide-6,12 with a small peak visible at 169.9 0C.

305

A.3 Chapter 8 polyamide-6 with phenothiazine. Section 8.3.1.1 Melting Temperatures for first heating ramp of the dry powders at 5 0C/min The DSC heating ramp for the dry powders of 83PA6PTh in Figure A-3 shows detail of the complex dissolution of phenothiazine into amorphous and/or crystalline polyamide-6 showing the faint peaks at 160 and 1700C before the main melting peak in the range 190 to 220 0C.

Figure A-3 DSC thermogram of the lead processes before the main melting peak for 83PA6PTh in taking the powders to the melt.

306

Section 8.3.1.2 Crystallisation for first cooling ramp of the molten blend at 25 0C/min. The first crystallisation of the pan blended 83PA6PTh is a skewed main peak that on closer examination can be shown to be a double peak in Figure A-4.

Figure A-4 DSC thermogram of the pan blended 83PA6PTh sample crystallising for the first time at 25 0C/min and showing a peak for the crystallisation of pure polyamide-6 just before the main crystallisation peak and demonstrating some phase separation.

307

A.4 Chapter 9 polyamide-6,9 with phenothiazine Section 9.3.1.2 Crystallisation for first cooling ramp of the molten blend at 25 0C/min The treble peak for the first crystallisation of the pan blended 63PA69PTh with the first two close to each other can be seen better here in Figure A-5 than in Chapter 9.

Figure A-5 First crystallisation at 25 0C/min of the pan blended 63PA69PTh with the first two peaks of the treble peak being almost co-incident in time. The treble peak starts with a very small crystallisation, presumably of polyamide-6,9 fthat immediately changes into the distinctive phenothiazine crystallisation before changing concentrations induce polyamide-6,9 crystallisation.

The first crystallisation of the pan blended 50PA69PTh in Figure A-6 incorporates three parts, a slow start to the crystallisations taking place, a straight section with the DSC signal rising linearly as with other phenothiazine and carbazole crystallisations, that peaking near 132 0C and changing into the main peak for the polymer crystallising that has a maximum near 130 0C.

308

Figure A-6 Double peak in the first crystallisation of pan blended 50PA69PTh showing the transition between phenothiazine crystallising with a maximum rate at 132 0C and polyamide-6,9 crystallising with a maximum at 130 0C.

309

Section 9.3.2.3 First time crystallisation in the DSC at 2 0C/min of ampoule samples

Figure A-7 Comparison between the first crystallisation in the DSC of ampoule samples 47PA69PTh and phenothiazine showing the polymer crystallisation peak near 142.5 0C and the much broader base of the initially “spiky” peak

310

A.5 Chapter 10 polyamide-6,12 with phenothiazine Section 10.3.1.1 Melting Temperatures for first heating ramp of the dry powders at 5 0C/min Figure A-8

shows the residual thermal activity before the main

melting peak for polyamide-6,12 powder reflecting stress relaxations and the melting/recrystallisation of metastable lamellae.

Figure A-8 Melting polyamide-6,12 powder at 5 0C/min displaying the thermal activity before the main melting peak due to stress relief and melting/recrystallisation of metastable lamellae. A baseline has been inserted to make the endothermic and exothermic activity clearer.

311

Section 10.3.1.3 Melting Peak Temperatures for second heating ramp at 5 0C/min The pan blended sample 36PA612PTh shows a slight amount of melting and recrystallisation in the lead up to the main melting peak during its second DSC heating ramp in Figure A-9. This is helped by studying the first derivatives of the heat flow signal.

Figure A-9 Heat Flow and the first derivative of heat flow for the second DSC melting ramp of 36PA612PTh with the lead up to the main melt showing slight melting and recrystallisation of metastable lamellae by 161 0C.

312

Appendix B

DSC, TMDSC AND LISSAJOUS FIGURES Lissajous Figures for Understanding Temperature Modulated Differential Scanning Calorimetry of Nylons This appendix covers some work that is associated with the thesis and is worth recording but does not fit so neatly in the main thrust of the work

B.1 Introduction Differential Scanning Calorimetry (DSC) has been a valuable tool for investigating the melting, crystallisation and recrystallisation characteristics of polymers. Standard DSC instruments can be of two types “Power Compensated”-and “Heat Flux”. “Power Compensated” instruments maintain the reference and the sample pans at the same temperature during the applied heating cycle. Separate heaters in the DSC cell for each pan maintain a zero temperature difference between sample and reference pans.

The power supplied to each pan is

monitored so the difference in the applied power can be used to calculate the heat flow. “Heat Flux” instruments use an alternative method where both pans are heated from a single heating source in the DSC cell walls. The temperature difference between the two pans is monitored using thermocouples positioned under the sample and reference pans. The actual heat flow into or out of the sample pan is calculated from the measured temperature difference. This approach is that used by TA Instruments equipment in this work. A Heat Flux instrument, model 2920 supplied by TA Instruments and equipped with a standard cell, was used for this work.

This cell is

constructed using heating elements wound around a vertical silver cylinder. A thin constantan disk is fitted horizontally across the middle of the DSC cell. The constantan disk supports the sample and reference pans on raised 313

dimples equidistant from the heating block.

Two fine wires welded

underneath the sample and reference pan dimples to chromel disks immediately below, form area thermocouples with the constantan. These are connected back-to-back so that the difference in temperature between the sample and reference is directly measured. The construction geometry of the cell is such that the thermal resistance from the heating block to the both pans is the same. Newton’s Law of Heat Flow is used to calculate the heat flow in and out of the sample Newton’s Law, between two locations 1 and 2 of differing temperature, is essentially given by Equation B-1.

dQ12 = λ × ∆T12 dt

Eqn. (B-1)

where dQ12/dt is the overall rate of heat flow, t is time, λ is the thermal conductivity of the path between the two locations and ∆T12 is the difference in temperature between the two locations. Essentially this specifies that the heat flows from the heating block (nominally through the constantan disk) to the dimples are proportional to the temperature differences between the heating block temperature (Tb) and either of the dimples. The cell relies on the thermal resistance between walls and the reference or sample dimples in the constantan disk being the identical.

Newton’s constant (λ) is thus the same in each case and is

effectively incorporated into the experimentally determined Cell Constant K. Thus:

dQbr = λ × (Tb − Tr ) dt

Eqn. (B-2)

dQbs = λ × (Tb − Ts ) dt

Eqn. (B-3)

and

Subtracting Equation B-2 from Equation B-3 effectively eliminates the block temperature (Tb) and provides the heat flow difference, dQsr/dt, between the sample and reference dimples.

This simplification demonstrates that the

difference in heat flow between the sample and reference (dQrs/dt) from or to the block can be determined as shown in Equation B-4.

dQrs = λ × (Tr − Ts ) dt 314

Eqn. (B-4)

Back-to-back thermocouples are used to directly measure the heat flow into the sample relative to that into the reference utilising the constant of proportionality from block-to-dimple thermal resistance. The cell is used with a purge gas flow into the lower and upper halves of the cell ensuring a quick response to changes in heat flow. Two gasses that are commonly used are nitrogen and helium.

Helium has greater heat

conductivity than nitrogen, thereby transporting heat to and from the cell walls, the constantan disk and the samples more efficiently than nitrogen. Nitrogen is commonly used in preference to helium however due to its much lower cost. The idealised DSC cell assumes all heat to and from the reference and sample occurs via the thermally conductive constantan disk.

Wunderlich

[161] has also argued that the lower thermal conductivity of nitrogen gas (compared to helium gas) ensures the DSC operates more ideally, whereby heat transfer occurs more readily through the constantan disk and less readily via the gas. However, in reality, the constantan disk and gas operate in parallel to form the thermal paths. In general, the heat flow into or out of a sample of material is made up of thermodynamic effects and kinetic effects. The thermodynamic component is dependent only upon the heat capacity of the material and the rate of change of temperature. The kinetic effects will be dependent on a function of the temperature and time history of the sample material. This is written [129] for a sample in Equation B-5 as:

dQs dT = - Cp × s + f [T, t ] dt dt

Eqn. (B-5)

where Cp is the Heat Capacity function at constant pressure and f(T,t) is a function of temperature and time. It is important to appreciate that in DSC and TMDSC that Cp is not constant but is variable with respect to temperature. It can be seen that the heat capacity can be determined by deliberately changing the heating rate and examining the way that heat flow changes. Thermodynamic changes can be classified as reversible within the context of 315

an experiment, provided that thermal processes occur at a rate that is significantly faster than the heating rate changes.

Examples of thermal

properties that are generally considered to be reversible are glass transitions and heat capacity. On the other hand, the kinetic component, dependant on temperature and time, is independent of the heating rate.

Examples of these are non-

reversible processes such as decomposition, evaporation, curing reactions, and changes from metastable crystalline structures to more stable ones. Sometimes there are events that appear to have both reversible and nonreversible characteristics.

It has been postulated that thermodynamically

reversible events may occur on a time frame that is similar to or slower than the speed of variation in heating rate.

A transition will seem in such

situations to have a mixed reversible/non-reversible character [134] or in some instances may appear completely non-reversible. An example of this is the melting of polymer crystals. There have been careful experiments by Okazaki and Wunderlich [40] showing that melting of poly(ethylene terephthalate) PET can be completely reversible over small temperature ranges where macromolecular chains partly melt from polymer crystals. A fully molten macromolecule would require (non-reversible) re-nucleation due to a free-enthalpy barrier to the crystallisation. Glass transitions can also display some non-reversible character [133, 162] when evaluated by TMDSC. Standard DSC merely measures the total heat flow rather than separating out the components. Heat Capacity (Cp) can be determined by measuring different samples at different heating rates in conventional DSC.

The original thermal history

would be destroyed in the first run if the same sample were to be run twice. Use of two samples also means a loss in accuracy because of the potential for weight measurement errors.

The time taken to carry out the

experimental runs is also longer, reducing throughput. Another problem with DSC is a requirement for small samples to achieve good resolution, which conflicts with the requirement for larger samples to achieve higher sensitivity.

The temperature differential within (normally)

poorly conducting polymer samples is greater if the sample is larger. This leads to a spreading of sharp transitions due to different parts of the sample 316

being at slightly different stages of the heating ramp, and thus a loss in resolution. 1993 saw the advent of Temperature Modulated Differential Scanning Calorimetry (TMDSC) [131, 163-166] . TMDSC superimposes a modulated heating rate on top of a constant heating ramp normally used in most DSC experiments.

Most equipment manufacturers for TMDSC use sinusoidal

modulation, but sawtooth modulation has also been employed. The period of the modulations is generally limited by physical factors and occurs in a range 10 to 100 s for commercially available equipment.

The discussion

here considers only the mathematical and graphical ramifications of sinusoidal modulations. Two further parameters are necessary to define the experimental system; the amplitude As (0C) of the sinusoidal temperature modulation at the sample: and P (sec), the period of the modulation [129]. Temperature (Ts) of the sample at a particular time (t) once the system has achieved steady state is given [40] by Equation B-6:

Ts = T0 + q × t - q ×

Cs

λ

+ A s × sin (ω × t - ε )

Eqn. (B-6)

The corresponding linear temperature expression for conventional DSC in steady state is given by Equation B-7:

Ts = T0 + q × t - q × Cs

Eqn. (B-7)

where: Ts is the sample temperature at time t. T0 is the starting temperature of the heating block, the reference pan and the sample pan with its sample, is the average or underlying heating rate, t is the time (s). ω is the modulation frequency (s-1) and equals 2Π/P. ε is the phase lag in radians between the block and the sample pan. λ is the thermal conductivity of block-pan path as mentioned previously. 317

and Cs is the heat capacity of the sample pan (including any sample). The term with Cs is caused by the temperature lag of the sample behind the temperature applied to the block.

There is an exponential lead-in to the

steady state condition at the start of the experiment and tends to zero as steady state is approached. There are virtually identical equations for Tr, the reference temperature with subscripts r instead of s and a phase lag of φ for the empty reference pan that is different from ε for the matched sample pan with sample. The sinusoidally modulated heat flow into and out the sample pan, due to the heat capacity of the sample itself, is still determined by Equation 4 and is proportional to the instantaneous difference in sample and reference temperatures. It will have a sinusoidally changing amplitude with a phase difference δ between block and ∆T. There will then be a measurable phase lag of Φ between the measured sample temperature Ts and the measured heat flow given by the difference between ε and δ. This difference Φ in phase between the sample temperature and the heat flow will become more apparent in the discussion of Lissajous figures and is shown in the figures presented later in this paper. A common procedure in TMDSC experiments is to apply a modulation amplitude such that, whilst heating, the temperature rate never goes negative, or during cooling, never goes positive. For example, an amplitude of 0.212 0C results in heating rates varying from zero to close to 4 0C/min at an average heating rate of 2 0C/min and period of 40 s. The advantages of such an experimental method are numerous. 1.

Both sensitivity and resolution are increased. The portions of a

modulated heating ramp with a high heating rate contribute a large signal, increasing sensitivity. The reason for this can easily be seen from Equation B-4 in that the heat flow signal will increase markedly as the heating rate increases. At the same time, the overall heating rate is slower than if the sample had been ramped continuously at the maximum rate for the whole time. 318

The

slower average heating rate with TMDSC overcomes the problem that would have been encountered with temperature gradients and thermal lags [134]. This means that samples as low as 2 or 3 mg may still be used in combination with the high and low heating rates of TMDSC to achieve a good signal. This compares with the 10 to 20 mg usually used in standard DSC. Commonly sample sizes of 5 to 10 mg are used with TMDSC and these are preferable, in the case of polymers to samples of 20 mg because of the delays inherent in heat transfer from the outside to inside of a sample. The small sample size is necessary because of the generally poor heat conductivity of polymers (vide supra). One of the conditions of accurate TMDSC work is that there is minimal temperature variation within the sample. 2.

The sample never goes into cooling during heating with a modulation

procedure of this nature. Whilst this is less important for the analysis of rapid thermodynamic transitions such as glass transitions that are experimentally reversible. This is not the case for slower reversible transitions, such as polymer melting.

It is possible during melting that

unwanted recrystallisation may occur during a short cooling phase of the modulation and that is not considered to be beneficial.

No such cooling

phase occurs using the heating profile described above. 3.

The experimental setup for the described modulation is such that the

heating rate is actually 0 0C/min at the slowest part of the heating cycle. We can see from Equation B-5 that at this point dT/dt=0. Therefore all heat flow is non-reversible, that is dQ/dt= f(T,t). It is possible to use this feature of TMDSC to check that the cell constant and heat capacity constant are correct by the alignment of the modulated heat flow at dT/dt = 0 with the derived continuous non-reversible heat flow. Another criterion is that the response is linear [132, 133, 167, 168]. Linearity occurs where doubling the temperature modulation amplitude leads to a measured doubling of the heat flow response. That is generally not the case during major transitions where materials are rapidly changing from one phase or morphology into another [169]. Equation B-5 clearly demonstrates that reversible heat flow is a function of heating rate and sample heat capacity. There is an enormously large and rapid increase in heat capacity during melting. 319

This, combined with the

modulated heating rate that also varies, results in an extraordinary variation in thermal conductivity of a polymer sample during melting. As a result, it would appear that accurate measurement of heat capacity, and hence reversing heat flow, is unlikely during melting of a sample. A restriction on the applicability of TMDSC is that the heat capacity of the material should not change significantly during a modulation cycle [132, 133]. This also calls into question the applicability of TMDSC to melting (or crystallisation). Accurate TMDSSC measurements require a DSC cell that responds quickly and precisely with the modulated temperature signal to rapid heating and cooling rates.

It is also necessary to have a quick response between the

thermal driving force (electric heating elements) or cooling system and the sample & reference pans. Multiple thermal paths between the walls of the cell and the pans and between the two pans can influence TMDSC measurements.

The primary

heat transfer path is generally considered to be via the constantan disk and another path is via the gas flowing in the cell. Most authors [129, 170, 171] have chosen to ignore or play down the role of the gas in thermal transport. These two paths have different speeds of transferring heat to or from the walls, and possibly between thermocouple and sample. This may influence TMDSC measurements that are a dynamic situation with varying heating rates as distinct from the more equilibrium situation of conventional DSC with constant heating rates. Newton’s constant λ is not easily measured for each of the heat transfer paths on their own (eg. constantan disk and purge gas). Therefore a cell constant is determined which effectively incorporates Newton's for all heat transfer media together. Only one set of calibration and cell constants is possible with the currently available software. Different calibration and cell constants are necessary for different experimental conditions. It means that if the heating ramp is run under one set of conditions eg a ramp rate of 2 0C/min and cooling is under other conditions eg. 0.5 0C/min then only the one part is able to properly run under calibrated conditions.

As a result, instrument makers are

continuously striving to improve cell design providing greater flexibility. 320

The measured heat capacity is found in TMDSC with its varying heating rates by comparing the amplitudes of the driving modulated heating rate and the resulting heat flow. The Reversing component can be isolated from the total by considering only the amplitudes of the swings in heat flow and heating rate [133, 170].

This is achieved by taking a discrete Fourier

transform of the data at the modulation frequency. C p = K Cp ×

Modulated Heat Flow amplitude Modulated Heating Rate amplitude

Eqn. (B-8)

where KCp is an instrument heat capacity calibration constant for the particular experimental conditions. The derivative of Equation B-6 with respect to time

(dTs/dt) is given by

Equation (B-9) and describes the modulated temperature (heating) rate of the sample.

dTs = q − ωAs cos(ωt − ε ) dt

Eqn. (B-9)

From Equation B-6, the amplitude of the modulated sample temperature is given by As and we can see from Equation B-9 that the amplitude of the sample temperature (heating) rate is given by ωAs. This results in a phase change difference between the modulated sample temperature (Ts) and the modulated heating rate of the sample (dTs/dt) of Π/2. The difference has ramifications discussed in the Lissajous section below. The use of the derivative of heating rate has advantages for non-isothermal work in that it removes the effects of the sample temperature creeping upwards at the average rate of the underlying heating rate . There is another form of the relationship as seen in Equation B-10 that occurs regularly [129 p. C-17, 170, 172] with quasi-isothermal work where the average heating rate is zero. There, the denominator is the amplitude of the temperature swings as a result of some simplifications for this special case. C p = K Cp ×

Maximum Heat Flow amplitude Maximum Temperatur e amplitude × ω

321

Eqn. (B-10)

TMDSC measures the Total Heat Flow (from the average over a cycle). The heat capacity Cp can continuously be determined using Equation (B-8). The heat capacity must be accurately calibrated in order to calculate the correct “Reversing” Heat Flow. The “Non-reversing” heat flow will also be wrong if this is incorrectly determined because it is found in Equation (B-11) by the difference between the Total Heat Flow and the Reversing Heat Flow from Equation (B-9). Non - Reversing Heat Flow = Total Heat Flow - Reversing Heat Flow

Eqn. (B-11)

The Total Heat Flow is the experimentally determined heat flow from the average over a cycle and the Reversing Heat Flow is obtained from the modulated heat flow signal after the equipment has been calibrated under identical conditions using the heat capacity of a known material. It is necessary to firstly measure the Cp of a material with well-characterised Cp (eg sapphire/Al2O3).

The experimentally determined heat capacity

constant KCp is essentially a second multiplier constant determined under modulating conditions of testing which becomes a multiplier of the experimentally determined cell constant (which itself is a multiplier to the experimentally determined heat flow).

This must be done under the same

conditions of purge gas type and flow, average ramp rate, period and amplitude that are to be used for the experimental work. The Heat Capacity (Cp) calibration is an extremely important part of obtaining correct Reversing Heat Flow and Non-Reversing Heat Flow results from the TMDSC approach. The determination of the correct Reversing Heat Flow signal is important because the Non-Reversing result is obtained by subtracting the Reversing signal from the Total Heat Flow signal. That Total Heat flow is obtained by taking the average heat flow signal over one cycle. Sapphire (Al2O3) is a material with heat capacity well known in the literature from various thermal measurements. It has the advantage that it has no thermal transitions from sub-zero temperatures to well over the 600 0C maximum working temperature of the instrument. The values of Cp for the material increase monotonically from 0.718 at 0 0C to 1.000 at 180 0C and 1.0895 at 300 0C. The KCp to achieve the correct Cp value varies with temperature. The results over a range of interest are averaged to give the best overall result. This will 322

obviously mean that inaccuracies in Cp measurement at temperatures other than the middle of the range can creep in. For example, the KCp required in one calibration run for 167 0C was 0.6897 and at 247 0C it was 0.7214, a 4.5% change over 80 0C.

An 80 0C temperature range of melting

temperatures in a single experiment can be quite normal and it cannot always be predicted beforehand what the exact range of interest will be. This means that the Reversing signal is under or over-estimated, and thereby the Non-Reversing signal is over-or under-estimated to an equivalent amount. The nature of the TA Instrument 2920 DSC software is that it is virtually impossible to adjust the results afterwards once the single KCp has been entered before the experimental run. That is, unless the data is exported to an ASCII datafile, imported into a spreadsheet such as Excel and the data manipulated further. An error in the calibration of heat capacity in TMDSC experiments will be clearly reflected in the measurements of both the reversing and nonreversing heat flows. It is an indicator for possible incorrect Cp calibration when exact “mirroring” of Reversing and Non-reversing components [171] is seen.

Incorrect Cp calibration leads to the Reversing and Non-reversing

components being “measured” inaccurately.

That is why heat capacity

calibration is so important. The correct allocation of reversing and non-reversing heat flows allows the separation of the thermodynamic and kinetic components of the material. Scherrenberg [172] raises the point that if the time scale of reversible processes in the polymer differ too much from the period of the measurement, they will not be “seen” by the experiment. This is also alluded to in Simon & McKenna’s paper [170] and Xu, Li and Feng’s work [171]. B.1.1 Lissajous Figures A Lissajous Figure is generated when one sinusoidal curve is plotted on the y-axis against another sinusoidal curve on the x-axis. This will result in a closed ellipse if the frequencies of the two curves are the same. There will be a straight line produced if the two curves are in phase. An ellipse will be produced when there is a phase lag between the two sinusoidal curves. Plotting the heat flow in a TMDSC experiment against the time derivative of sample temperature (or against the temperature for a quasi-isothermal 323

experiment) will create a Lissajous Figure. These can be seen in the work of several authors [40, 171, 173-179]. The slope of the straight line in the very ideal case of zero phase difference will be proportional to the heat capacity of the sample when matched pans are used for reference and sample. This is depicted in Figure B-1 for two differing heat capacities and in-phase sinusoidal curves in the use of TMDSC.

Figure B-1 Representation of Lissajous figures at two heat capacities in the ideal case with zero phase lag in the cell between sample and reference.

There are a number of causes for phase lags to occur.

Phase lag occurs

because there is a time difference between the modulated temperature signal measured at the reference and the sample thermocouples.

This time lag

may be instrumental. Factors such as: a)

Reference and sample thermocouple not symmetrically positioned.

b)

Variation in the constantan disk creating uneven heat flow to the sample and reference thermocouples.

c)

Chromel disks and thermocouples for the reference and sample may not be identical resulting in heat flow variations.

d)

Non-uniform heating block. 324

e)

Uneven heat flow via the purge gas.

f)

Unequal thermal resistance between the constantan dimples and the two calorimeter pans.

In addition to this, thermal conductivity of samples can vary during thermal transitions (such as melting) due to associated large variations in sample heat capacity. Rapid, large thermal conductivity variations by a sample can induce short-term phase lags between the sample and reference pans. Zero phase lag would be ideal for TMDSC analysis, or failing this, possibly a constant phase lag, However, in reality this is not the case. An ellipse will be generated for TMDSC when the modulated heat flow is out of phase with the sample temperature derivative and a phase lag Φ has developed between the signals. Drift of the Lissajous figure can occur if: a)

the kinetic component of heat flow f(T,t) is changing,

b)

the heat capacity is changing, in particular during transitions, or

c)

dTs/dt is varying, for example during transitions.

Distortions arise if the heat flow process is nonlinear.

B.2 Experimental B.2.1 The Choice of TMDSC Experimental Conditions For Trials The choice of experimental conditions for DSC work is important, but this is much more the case for TMDSC work. Constraints on experimental conditions can sometimes be conflicting: a)

The period should be as long as possible but at the same time should give 3 to 4 cycles during transitions.

b)

The instrument should retain accurate control of heating and cooling the pans in order to provide repeatable results.

A maximum cooling

rate of 25 0C/min used during cooling to maintain instrumental control with the TA Instruments 2920 operating in the range 300 0C down to 25 0C and the LNCA setting on switch position 3. c)

Non-isothermal modulation should not have an amplitude great enough to give negative heating rates during ramping through melt transitions, 325

otherwise unwanted re-crystallisation can occur during the cooling phase, unduly influencing the results. d)

The empty reference pan should match the sample pan (and lid) to within 100 µg [129].

The method chosen to comply with the last point above was to select a series of hermetic pans and lid combinations that were spaced apart in mass by not more than 200 µg for use as reference pans after crimping. Pans and lids used for samples could then easily be paired with an appropriate reference pan that was within the 100 µg variation limits. Lids were usually chosen to provide pairings that were much better than the above limits. An ATI Cahn C-34 balance capable of weighing to 3 to 5 µg was used. Many of the transitions experienced with sample materials used in general with research work had sharp transitions and there is a requirement from the Fourier transformation process that at least four cycles should occur over a transition to avoid artefacts [148]. Therefore the cycle period used in this experimental work was 40 s. Helium purge gas was used for polyamide4,6 with the cycle periods of 40 s, better than the instrument manufacturer’s recommendations [129] for use with 30 s or less, and in the light of the findings here on Lissajous Figures. The appropriate amplitude for a heating ramp rate of 2 0C/min with a 40 s period was 0.212 0C in order to maintain a zero or positive heating rate during the whole ramp [148]. B.2.2 Experimental Runs Baseline calibrations over the range 25-320 0C were run with an empty cell. Cell Constant & Temperature calibrations were carried out with indium over the range 140-170 0C at a ramp rate of 2 0C/min for work with either helium or nitrogen purge gas flow in both parts of the cell. The indium calibrant was pre-melted and solidified in the pan to give good thermal contact. The Heat Capacity constant KCp was calibrated with Al2O3 at 2 0C/min in the range 170-230 0C and/or 260-310 0C, using a modulation amplitude of 0.212 0C and 40 s period. All calibrations were carried out using identical helium or nitrogen gas flow and average ramp rate to be used in the final experiments. 326

TMDSC Measurements were carried out for polyamide-4,6 in the range 25-315 0C at 2 0C/min using helium purge gas at 50 ml/min. Polyamide-4,6 melts at approximately 290 0C.

B.3 Results and Discussion The Lissajous plot in Figure B-2 shows that, under modulated conditions, a number of 40 s loops taking several minutes to stabilise, are required to achieve stabilisation. There is a feedback loop so that the block temperature modulations are modified dynamically to give the required modulation amplitude of the sample temperature. A major transition, such as melting, must be expected to take some time to restabilise the modulations.

Figure B-2 Lissajous Figure of an Al2O3 sample achieving thermal control at the start of a TMDSC experiment. Average heating rate is 2 0C/min with a period of 40 s, an amplitude of 0.212 0C and 50 ml/min helium purge gas.

The comparison between Figure B-3 and Figure B-4 is instructive because it displays the difference in Lissajous figures of Al2O3 in changing from helium to nitrogen purge gas at the same flow rate and with all other instrumental conditions being identical. Only one loop is displayed in each figure in order to demonstrate the delays clearly but the loops are very typical of those found experimentally. It is clear from Figure B-3 and Figure B-4 that the ellipse from the nitrogen gas is pointing at approximately 90o to that 327

observed for helium. It is also much rounder than that seen with helium. The difference in phase Φ is determined from maximum vertical tangent clockwise to the next horizontal tangent. These show that the lower heat conductivity of nitrogen, compared to helium, results in substantial thermal delays in transferring heat to and from the sample material.

Figure B-3 Lissajous Figure for Al2O3 sample displaying the phase lag Φ with TMDSC for a 50 ml/min helium purge gas flow, heating rate of 2 0C/min, period of 40 s and an amplitude of 0.212 0C.

It is quite clear that the trial with helium purge gas has a delay of approx 2 s whilst the trial with nitrogen shows a phase delay of approximately 15 s with respect to the sample temperature. The measured heat flow is actually not that for the sample material or even the sample and sample pan but is based on the difference in temperature between the sample thermocouple and reference thermocouple.

There will be a measurable thermal resistance

between thermocouple and pan and between pan and sample, even where the sample has already been pre-melted onto the base of the pan. A paper to be published soon will address these issues.

328

Figure B-4 Lissajous Figure for Al2O3 sample displaying the phase lag Φ with TMDSC for a 50 ml/min nitrogen purge gas flow, heating rate of 2 0C/min, period of 40 s and an amplitude of 0.212 0C.

Figure B-5 Lissajous Figure for Al2O3 sample with TMDSC for a 25 ml/min nitrogen purge gas flow, heating rate of 2 0C/min, period of 40 s and an amplitude of 0.212 0C.

329

The Lissajous ellipse for helium becomes rounder when the gas flow is reduced substantially. That is also demonstrated by Cser’s work [148] where he shows that even at 50 ml/min there is starting to be a change from the results at 75 and 100 ml/min. This can be seen in Figure B-5 where the helium flow was set at 25 ml/min in both purge gas streams rather than 50 ml/min.

In this case, the thermal delay in heat flow has increased

markedly with the lowered flow rate. The above shows that thermal transport with gas is significantly more important than the transport via the constantan disk or there would be little change in direction of the major axis of the ellipse or the shape of the ellipse. The TA Instruments recommendation to use helium for experiments with short periods seems well placed and should probably be extended to all periods. The nitrogen results are much further from the in-phase line than the helium results.

It is probable that other physical cell designs and

material choices could lead to a more responsive system.

Figure B-6 Lissajous Figure of polyamide-4,6 before, during and after melting in a TMDSC experiment with a 50 ml/min helium purge gas flow, heating rate of 2 0C/min, period of 40 s and an amplitude of 0.212 0C. The three sections are displayed but with portions of the trace leading into and out of the melting removed for clarity. They have not been shifted vertically. CHI_fig6X

330

A Lissajous plot for polyamide-4,6 before, during and after melting is presented in Figure 6 and demonstrates the distortions to a linear response during melting [162, 177]. The Lissajous loops in Figure B-6 just at the time of going into and out of the melt have been excluded from display for clarity. The results are typical of other polyamides such as polyamide-6,12, polyamide-6 and polyamide-6,9 measured with TMDSC. The changing shape of the melting section of the Lissajous Figure B-6 is due to: a)

The loops descending on the page and then rising again to the post-melt situation.

This is due to the average envelope of the heat flow

oscillations dipping during the melting. The downward drift of the loops as the polyamide melts are exactly the same as the dip in modulations seen in the inset in Figure B-7.

(The perspective in Figure B-7 is a

normal one of heat flow versus temperature.) b)

The height of the oscillations increases markedly. This is due to the increase in heat capacity at the melting temperature.

In theory Cp

should become infinite just at the melt because of the latent heat of fusion. c)

The top of the oscillations dips because there is a kinetic component to the melting. There is absorption of energy by kinetic processes at zero heating rates.

Much of this is due to the disengagement of

macromolecules from lamellae. d)

The loops become slightly wider at the peak of the melting (the bottom). The heat flow into the sample whilst the reference continues to heat causes an increase in ∆T and this is reflected in the wider range of heating rate because the extra temperature difference must be achieved in the same modulation cycle.

That requires a greater heating rate

during part of the cycle. The broadening in the temperature derivative direction seen during the polyamide-4,6 melt is similar to that observable in the Lissajous plot in Fig. 8 of Xu et al. [171] for crystallisation of amorphous metal alloy. e)

The loops become rounder indicating phase lags [134].

There are

delays built into the melting process. The advantage of studying the 331

Lissajous figures is that the distortion towards the bottom loop is clearly evident due to large phase changes between the signals It may be that the “anomalous behaviour” referred to by Cser et al. [148] is due to loss of linearity during the melting of HDPE, leading to non-elliptical Lissajous plots rather than the Fourier transformation process. It can be seen that the top of the Lissajous ellipses do not quite coincide with a zero heat flow even though the rate of heating drops to zero for each modulation. The ellipses are often slightly positive. Dr Mike Reading, the inventor of TMDSC, was queried about this in early 2001. He suspected that the positive heat flow may be due to slight mismatches in pan weights. This does not appear to be the case as in the one experimental run on polyamide4,6 displayed in Figure B-6 it can clearly be seen that the tops of the ellipses move from the range -0.002 to -0.004 before the melt to the range -0.025 to -0.028 after melting.

This is the same run with the same sample and

reference pan so that cannot be the case. These differences may possibly be due to baseline curvature. The amplitude of the modulations has been chosen so that there is zero heating during part of the cycle.

That means the Reversing component

should be zero at that stage. It would also infer that the “Non-reversing” signal should be touched by the Modulated Heat Flow curves at those points. Figure B-7 shows that this has not occurred and there is a minor deviance of the Non-reversing heat flow signal. It must be noted here that this will only be the case where the there is no rapid change in the Non-reversing signal because of the timing effects of the 1½ cycle delay in “temperature” between the instantaneously derived Modulated Heat Flow signal and the computed Non-reversing signal.

Further work will be required to clarify this issue

which may also be related to the issue above where the top of the Lissajous ellipse, under equilibrium conditions, has a slight negative heat flow rather than zero as expected. The system has to recover quickly from the perturbation of a melt and that recovery is driven by the block temperature. The sample temperature is utilised as the control element by the instrument, hence there is a feedback response loop involved. A quick response to the perturbations at the sample requires a quick thermal path from the block to the sample pan. This infers 332

a low thermal resistance that will obviously have some implications with instrument sensitivity.

The results will not be valid whilst the system is

trying to recover [167].

Figure B-7 Non-reversing and Modulated Heat flow versus Temperature for the region around the melting of polyamide-4,6 in TMDSC for a 50 ml/min helium purge gas flow, heating rate of 2 0C/min, period of 40 s and an amplitude of 0.212 0C.

B.4 Conclusions TMDSC does have the potential to increase sensitivity whilst improving resolution and to separate “Reversible” processes from “Non-Reversible” ones on the time-scale of the experiment. Heat Capacity (Cp) can be determined on a continuous basis in a relatively quick experiment with TMDSC from a single run on a single small sample and achieve high sensitivity and resolution. The restriction is that accurate heat capacity measurements are most unlikely during major thermal transitions, such as melting and crystallisation. The method does allow exploration of the effects of in-built stress or ageing on samples, of glass transitions (with some caveats) [180] and of liquid-liquid demixing in polymer-diluent systems [7]. There are limitations as to which transitions can be accurately investigated due to the limitations on TMDSC 333

of sample size, linearity and the stability of Cp during some thermal transitions. The results obtained in TMDSC experiments are quite dependent on the experimental conditions but results can easily be monitored by Lissajous plots. The plots of Modulated Heat Flow versus the Derivative of Modulated Temperature can be used to alert to unfavourable experimental conditions where large thermal delays or loss of system linearity can be seen. TMDSC does require an extra calibration step for accurate determination of heat capacity. Accuracy is extremely important as the results obtained are much more dependent on experimental conditions. Sample preparation is also very important [171]. The choice of purge gas and gas flow rate should aim to approximate the ideal theoretical situation of in-phase signals. In this way transitions can properly be followed in many systems. Helium purge gas is recommended for all TMDSC measurements in TA Instruments Standard Cells because the heat path delays between cell block and the sample and reference pans have been shown to be adversely affected by the use of nitrogen. The flow rate should be at least 50 ml/min. Other types of cells may also require helium purge gas, depending upon their type of construction and the relevant heat flow paths.

334

Appendix C

MID RANGE IR ASSIGNMENTS FOR POLYAMIDE TYPES STUDIED IN THE THESIS Appendix C gives a table of FTIR spectroscopy peak frequencies in the mid Infra Red region for the four polyamides studied here.

Results from the literature in differing crystallographic environments are provided along with the

measured values from photoacoustic experiments on the polyamides with their morphology from formation in ampoules.

C.1 Mid range IR and hydrogen bond interactions The table presented below is a compilation of assignments from various literature sources along with measured values of 335

peak wavelengths found from ampoule samples of the polyamides and from some quenched samples also measured. Table C-1 Assignments for polyamide bands Bold=[99, pp.85-88], Italic [48, p. 504]. Normal from ampoule samples and PA46 Gaymans are solution cast FTIR peaks from [122] in cm-1 FTIR Assignment

PA46

PA6 α or

PA6 γ

PA46

PA46

PA6

Gaymans

Meas.

Amorph.

PA6

3305

3300

3304

3290

3290

3302

N-H band

3066

3070

3067

3070

3090

CH2 asym stretch

2944

2945

2944

2930

CH2 sym stretch

2872

2870

2872

PA69

PA69

PA612

PA612

Meas

Meas.

3302

3302

3300

3069

3090

3090

3063

2930

2935

2930

2930

2921

2865

2860

2868

2854

2854

2853

Amide I unassoc

1667

1650

1652

1652

1652

Amide I ordered

1647

1643

1647

1645

1645

1642

1642

1636

1636

1634

Hydrogen bonded N-H

Meas.

β

stretch

Amide I

1652

1638

1652

1636

FTIR Assignment

PA46

PA46

PA46

PA6

Gaymans

Meas.

Amorph.

PA6 α or

PA6 γ

PA6

PA69

Meas.

β

PA69

PA612

Meas

PA612 Meas.

C=O stretch Amide II

1540

1540

1545

1562

1559

1540

1560

1541

1558

1558

C-N stretch + C(O)-N-H bend Amide II unassoc.

1560

1541

336

N vic,CH2 bend (α)

1476

CH2 bend

1464

CH2 bend

1438

CO vic, CH2 bend (α)

1418

1419

1417

1363

1363

1374 1265

Amide III (γ)

1279

1476 1459 1440

1280

1281

1281

1476

1471

1471

1474

1463

1463

1466

1466

1467

1436

1442

1437

1436

1436

1436

1438

1418

1420

1419

1419

1419

1369

1373

1371

1368

1369

1370

1269

1266

1277

1276

1277

1250

1237

1237

1218

1217

1236 1201

1170

(γ),amorph Amorph

1140

1142

1027 984

944 906

1170

940

943 906

1170

1242 1201

1194

1196

1188

1188

1170

1180

1189

1180

1188

1131

1126

1116

1116

1124 1121

1123

1111

1112

1060

1079

1071

1072

1072

1064

1064

1029

1048

1029

1026

1017

1036

1026

988

988

988

987

959

977

968

930

922

929

940

942

937

938

900

899

897

899

1002

989

C-CO stretch α or γ C-CO stretch α or γ

1199

1200

1474

1464

1249

(α)

1540

FTIR Assignment

CH2 wag

PA46

733

PA46

PA46

PA6

Gaymans

Meas.

Amorph.

730

PA6 α or

PA6

PA69

PA69

PA612

PA612

β

Meas.

834

833

853

842

777

797

799

791

792

730

730

727

728

731

730

720

721

690

692

731

731

PA6 γ

Meas

Meas. 853

712

Amide V (γ) N-H out of plane Amide V (α & β)

693

690

692

691

691

689

623

Amide VI (γ)

695 627

612

C=O out of plane Amide VI (α)

581

575

579

581

Amide VI amorph

579

588

578

337

524

520

525

574

582

582

539

539

574 522

521

522

532

535

Appendix D (on CD only)

FOURIER TRANSFORM INFRARED SPECTRA (PART 1) D.1 N-H stretches with Photoacoustic FTIR in Mid Range Infrared D.1.1

Introduction

Appendix D comprises four sections: a) Part 1 with mid range infrared spectra of the N-H stretches and found in this file Appx_D1.pdf. b) Part 2 with mid range infrared spectra in the region 1700 to 400 cm-1 and found in the file Appx_D2.pdf. c) Part 3 with near infrared spectra in the region 7300 to 5700 cm-1 and found in the file Appx_D3.pdf. d) Part 3 with near infrared spectra in the region 5000 to 4300 cm-1 and found in the file Appx_D4.pdf. This part of Appendix D provides Fourier Transform Infrared (FTIR) spectra supporting the argument that the diluents do not significantly interact by hydrogen bonding with the polyamides.

It displays the spectra for Mid Infrared range

measurements carried out with Photoacoustic Spectroscopy (PAS). The PAS results focus on the Amide-A N-H stretches where the N-H of the diluent should shift from “free’ to “bound” and the polyamide where the N-H stretch should change from “bound” to “free”. Figures are presented per material combination with either 3 or 4 samples representative of the available concentration range measured. This is displayed in spectra covering either the range 3450 to 3250 cm-1 where carbazole is the diluent or 3400 to 3200 cm-1 for phenothiazine as the diluent. The spectra for the blends from ampoule materials are shown with composite spectra by appropriately summing spectra from the two constituents so that minor peaks of each constituent are D-1

Mid IR range FTIR spectra of N-H stretches in the correct ratios, as described in Chapter 1. This does lead to major peaks often differing strongly in height from those of the blends because the intensity of strong photoacoustic peaks for the pure materials are non-linear. This does not affect the frequencies for peaks but does make the peak heights for the N-H stretches of the two constituents appear to be in the wrong ratio. The model helps accurately evaluate possible peak shifts because the peak for carbazole is on a sharply falling portion of the polyamide and vice versa. In the case of phenothiazine as a diluent, the N-H stretch is very close to that of the polyamides and the rounding caused by the non-linear response distorts the model substantially. The cases of polyamide/phenothiazine blends are perhaps better evaluated by comparing the blend peaks with those of the constituents. The models have been left in the figures for reference. Each figure has all the spectra displayed full scale to facilitate viewing the comparisons of peak positions. D.1.2

Notation

The notation used for Differential Scanning Calorimeter (DSC) samples in the main section of the thesis has been based upon the known percentage of polyamide in the sample, the polyamide type and the diluent type eg. 54PA46Car for 54%polyamide in a polyamide-4,6/carbazole blend.

The weight percentages for ampoule samples were obtained from

Thermogravimetric Analysis samples taken right next to the DSC sample. The percentage polyamide in different parts of an ampoule sample had often been found to differ markedly within the ampoule. Fourier Transform Infraied samples are much larger than the DSC samples and the spectra are an average over a larger area but relevant for a depth into the sample depending on measuring technique and conditions. They may well not be close in constitution to DSC samples that had been taken from the same ampoule. Also, fresh samples taken for near infrared measurements using Diffuse Reflectance Infrared Fourier Transform at a later stage of the experimental work may differ again in composition for the sampling volume. For this reason, the samples in FTIR are referred to only by the blend type and the ampoule they came from, eg. PA6PTh_A48 for a polyamide-6/phenothiazine sample from Ampoule 48. In a few cases, noteworthy features such as the colour being bright red are noted as well. D-2

Mid IR range FTIR spectra of N-H stretches Models by spectral summation were created for the mid range IR to match the salient features at selected frequencies by combining appropriate proportions of the spectra of the constituents, as described in Chapter 1. Absolute values for peak heights of each constituent can vary, depending on the exact positioning of the sample surface. Therefore the percentages of each spectrum added for the model do not necessarily reflect the weight proportions of the materials in the spectra they match. The same colour coding, as used for thermograms, but based on the percentage of photoacoustic spectra added, was used for the blend samples, mainly to make it easier to pick out one figure from another in a series of them. The figures within each materials combination are placed in order of descending level of polyamide.

D-3

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-4,6 and carbazole Polyamide-4,6 and carbazole *Spectral Math Result PA46Car_A57Fawn 55

50

Photoacoustic

D.1.3

45

40

35

30

25

20 3440

3420

3400

3380


3320

3300

3280

3260

Figure D.1-1 Mid range photoacoustic FTIR N-H stretches for Ampoule 57(Fawn) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-4

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-4,6 and carbazole 75

*Spectral Math Result PA46Car_A29

70 65

Photoacoustic

60 55 50 45 40 35 30 25 3420

3400

3380

3360 3340 3320 Wavenumbers (cm-1)

3300

3280

3260

Figure D.1-2 Mid range photoacoustic FTIR N-H stretches from Ampoule 29 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-5

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-4,6 and carbazole

60 55

Photoacoustic

50 45 40 35 30 25

*Spectral Math Result PA46Car_A57BrightRed 3440

3420

3400

3380


3320

3300

3280

3260

Figure D.1-3 Mid range photoacoustic FTIR N-H stretches for Ampoule 57 (Bright Red) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-6


60

Photoacoustic

55

50

45

40

35

30 25

*Spectral Math Result PA46Car_A30 3440

3420

3400

3380


3320

3300

3280

3260

Figure D.1-4 Mid range photoacoustic FTIR N-H stretches for Ampoule 30 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-7

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6 and carbazole Polyamide-6 and carbazole

130


120 110 Photoacoustic

D.1.4

100 90 80 70 60 50

3440

3420

3400

3380


3320

3300

3280

3260

Figure D.1-5 Mid range photoacoustic FTIR N-H stretches for Ampoule 61 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-6.

D-8

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6 and carbazole 70


65 60

Photoacoustic

55 50 45 40 35 30 25 3440

3420

3400

3380


3320

3300

3280

3260


D-9

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6 and carbazole *Spectral Math Result PA6Car_A59 65 60

Photoacoustic

55 50 45 40 35 30 25 3440

3420

3400

3380


3320

3300

3280

3260


D-10

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6 and carbazole 80


75 70

Photoacoustic

65 60 55 50 45 40 35 30 3440

3420

3400

3380


3320

3300

3280

3260


D-11

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,9 and carbazole Polyamide-6,9 and carbazole 55


50

45

Photoacoustic

D.1.5

40

35 30 25

20

15 3440

3420

3400

3380


3320

3300

3280

3260


D-12



55

Photoacoustic

50 45 40 35 30 25 20

3440

3420

3400

3380


3320

3300

3280

3260


D-13

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,9 and carbazole 70 65 60

Photoacoustic

55 50 45 40 35 30 25


3420

3400

3380


3320

3300

3280

3260


D-14

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,12 and carbazole Polyamide-6,12 and carbazole 60

*Spectral Math Result 60PA6Car_A34

55 50 Photoacoustic

D.1.6

45 40 35 30 25 20 3440

3420

3400

3380


3320

3300

3280

3260


D-15

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,12 and carbazole 50 45

Photoacoustic

40

35

30

25

20

15


3420

3400

3380


3320

3300

3280

3260

Figure D.1-13 Mid range photoacoustic FTIR N-H stretches for ampoule 41 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-6,12.

D-16

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,12 and carbazole 70 65 60

Photoacoustic

55 50 45 40 35 30 25


3420

3400

3380


3320

3300

3280

3260


D-17

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-4,6 and phenothiazine Polyamide-4,6 and phenothiazine 150 140

PA62PTh PA46_A44 *Spectral Math Result PA46PTh_A47

130 120 Photoacoustic

D.1.7

110 100 90 80 70 60 50 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220

Figure D.1-15 Mid range photoacoustic FTIR N-H stretches for Ampoule 47 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-4,6 and the spectra of the constituents.

D-18

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-4,6 and phenothiazine 70 65

phenothiazine polyamide-4,6 *Spectral Math Result PA46PTh_A42

60

Photoacoustic

55 50 45 40 35 30 25 20 3450

3400


3250


D-19

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-4,6 and phenothiazine 24


22

Photoacoustic

20

18

16

14

12

10 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-20

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6 and phenothiazine Polyamide-6 and phenothiazine 52 50 48

phenothiazine polyamide-6 *Spectral Math Result PA6PTh_A43

46 44 42 Photoacoustic

D.1.8

40 38 36 34 32 30 28 26 24 22 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220

Figure D.1-18 Mid range photoacoustic FTIR N-H stretches for Ampoule 43 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6 and the spectra of the constituents.

D-21

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6 and phenothiazine 30 28

phenothiazine polyamide-6 *Spectral Math Result PA6PTh_A18Yellow

26

Photoacoustic

24 22 20 18 16 14 12 10 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220

Figure D.1-19 Mid range photoacoustic FTIR N-H stretches for Ampoule 18 (Yellow) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6 and the spectra of the constituents.

D-22


phenothiazine polyamide-6 *Spectral Math Result PA6PTh_A49

65

Photoacoustic

60 55 50 45 40 35 30 25 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-23


phenothiazine Polyamide-6 *Spectral Math Result Pa6PTh_A48

70 65

Photoacoustic

60 55 50 45 40 35 30 25 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-24

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,12 and phenothiazine Polyamide-6,9 and phenothiazine 130 120

phenothiazine polyamide-6,9 *Spectral Math Result PA69PTh_A52GreyTough

110

Photoacoustic

D.1.9

100 90 80 70 60 50 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220

Figure D.1-22 Mid range photoacoustic FTIR N-H stretches for Ampoule 52 (Grey and tough) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6,9 and the spectra of the constituents.

D-25


A62PTh_B707 polyamide-6,9 *Spectral Math Result PA69PTh_A51

130

Photoacoustic

120 110 100 90 80 70 60 50 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-26



80 75

Photoacoustic

70 65 60 55 50 45 40 35 30 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220

Figure D.1-24 Mid range photoacoustic FTIR N-H stretches for Ampoule 50 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6,9 and the spectra of the constituents,

D-27

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,12 and phenothiazine D.1.10 Polyamide-6,12 and phenothiazine

70

phenothiazine polyamide-6,12 *Spectral Math Result PA612PTh_A55LightGrey

65 60

Photoacoustic

55 50 45 40 35 30 25 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220

Figure D.1-25 Mid range photoacoustic FTIR N-H stretches for Ampoule 55 (Light grey) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6,12 and the spectra of the constituents,

D-28



120

Photoacoustic

110 100 90 80 70 60 50 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-29



60

Photoacoustic

55 50 45 40 35 30 25

3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-30

Mid IR range FTIR spectra of N-H stretches for blends of polyamide-6,12 and phenothiazine

13


12

Photoacoustic

11 10 9 8 7 6 5 3400

3380

3360

3340

3320 3300 3280 Wavenumbers (cm-1)

3260

3240

3220


D-31

Appendix D (on CD only)

FOURIER TRANSFORM INFRARED SPECTRA (PART 2) D.2 1700 to 400 cm-1 portion of the Mid Range Infrared spectrum with Photoacoustic FTIR D.2.1 Introduction Appendix D comprises four sections: a) Part 1 with mid range infrared spectra of the N-H stretches and found in the file Appx_D1.pdf. b) Part 2 with mid range infrared spectra in the region 1700 to 400 cm-1 and found in this file Appx_D2.pdf. c) Part 3 with Near Infrared spectra in the region 7300 to 5700 cm-1 and found in the file Appx_D3.pdf. d) Part 3 with Near Infrared spectra in the region 5000 to 4300 cm-1 and found in the file Appx_D4.pdf. Part 2 does not add to the evidence for there not being hydrogen bond interactions between the constituents in the polyamide/diluent blends but has been added for reference. The non-linearity of the strong photoacoustic peaks in the 1700 to 400 cm-1 region, combined with the effects of the quite different morphology of the blends to the constituent materials make forming conclusions very difficult. Some of the effects of morphology changes on peak positions in this region can be seen from the table in Appendix C and there are references in Chapter 1 to the work of others in assigning peak shifts in this spectral region to changes in crystallographic form or morphology of individual polymers. It is believed that there were interactions between the materials in some cases because blend samples were often purple, bright pink or mottled pink, pointing to charge transfer, or pi-pi conjugation taking place.

D-32

Mid IR range FTIR Spectra The figures use models from appropriately summing the spectra of the pure constituents. The model formation is covered in Chapter 1.

Figures are presented per material combination with either 3 or 4 samples representative of the available

concentration range measured.

It should be noted that the models matched the heights of certain peaks to obtain close to

the correct ratios of the constituent spectra. The figures are displayed in Full Scale mode in this part of the appendix and therefore do not necessarily show the relevant peaks in the same scaling. D.2.2 Notation Notation and colours for spectra in this section of Appendix D are described in Part 1 of the appendix.

D-33

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and carbazole D.2.3 Polyamide-4,6 and carbazole 65 60 55

Photoacoustic

50 45 40 35 30 25 20

*Spectral Math Result PA46Car_A57Fawn 1600

1400


800

600

Figure D.2-1Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm---111 region for Ampoule 57 (Fawn) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-34

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and carbazole 55 50 45

Photoacoustic

40 35 30 25 20 15 10


1400


D-35

800

600

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and carbazole Figure D.2-2 photoacoustic FTIR spectrum in the 1700 to 400 cm---111 region for ampoule 29 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

80

*Spectral Math Result PA46Car_A57BrightRed

75 70

Photoacoustic

65 60 55 50 45 40 35 30 1600

1400


800

600

Figure D.2-3 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm---111 region for Ampoule 57 (Bright Red compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-36

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and carbazole 70


65 60

Photoacoustic

55 50 45 40 35 30 25

1600

1400


800

600

Figure D.2-4 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm---111 region for Ampoule 30 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-4,6.

D-37

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and carbazole D.2.4 Polyamide-6 and carbazole 150 140 130 120

Photoacoustic

110 100 90 80 70 60 50 40


1400


800

600

Figure D.2-5 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm---111 region for Ampoule 61 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-6.

D-38

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and carbazole 65 60 55

Photoacoustic

50 45 40 35 30 25 20 15


1400


800

600


D-39

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and carbazole 70


65 60

Photoacoustic

55 50 45 40 35 30 25 1600

1400


800

600


D-40

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and carbazole 90


85 80 75

Photoacoustic

70 65 60 55 50 45 40 35 30 1600

1400


800

600


D-41


Photoacoustic

50 45 40 35 30 25 20 15


1400


800

600


D-42



70 65 60

Photoacoustic

55 50 45 40 35 30 25 20 1600

1400


800

600


D-43

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,9 and carbazole

70


65 60

Photoacoustic

55 50 45 40 35 30 25 20 1600

1400


800

600

Figure D.2-11 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for Ampoule 71 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents carbazole and polyamide-6,9.

D-44


Photoacoustic

55 50 45 40 35 30 25

*Spectral Math Result 60PA6Car_A34 1600

1400


800

600


D-45

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,12 and carbazole

50

45

Photoacoustic

40

35

30

25

20

15


1400


800

600


D-46



65 60

Photoacoustic

55 50 45 40 35 30 25 1600

1400


800

600


D-47

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and phenothiazine D.2.7 Polyamide-4,6 and phenothiazine 140 130 120

Photoacoustic

110 100 90 80 70 60 50 40 30

*Spectral Math Result PA46PTh_A47 1600

1400


800

600

Figure D.2-15 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for Ampoule 47 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-4,6.

D-48

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and phenothiazine 75 70

*Spectral Math Result PA46PTh_A42

65 60

Photoacoustic

55 50 45 40 35 30 25 20 15 1600

1400


800

600


D-49

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-4,6 and phenothiazine 26 24 22

Photoacoustic

20 18 16 14 12 10 8 6


1400


800

600


D-50

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and phenothiazine D.2.8 Polyamide-6 and phenothiazine 60 55 50

Photoacoustic

45 40 35 30 25 20 *Spectral Math Result 15

PA6PTh_A43 1600

1400


800

600

Figure D.2-18 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for Ampoule 43 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6.

D-51

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and phenothiazine 36 34 32 30 28 Photoacoustic

26 24 22 20 18 16 14 12 10

*Spectral Math Result PA6PTh_A18Yellow 1600

1400


800

600

Figure D.2-19 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for Ampoule 18 (Yellow) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6.

D-52

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and phenothiazine 80 75


70 65

Photoacoustic

60 55 50 45 40 35 30 25 20 15 1600

1400


800

600


D-53

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6 and phenothiazine 90 85

*Spectral Math Result Pa6PTh_A48

80 75 70

Photoacoustic

65 60 55 50 45 40 35 30 25 20 15 1600

1400


800

600


D-54

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,9 and phenothiazine D.2.9 Polyamide-6,9 and phenothiazine 150 140 130 120

Photoacoustic

110 100 90 80 70 60 50 40 30 20

*Spectral Math Result PA69PTh_A52ToughGrey 1600

1400


800

600

Figure D.2-22 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for Ampoule 52 (Grey and tough) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6,9.

D-55

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,9 and phenothiazine 180 170 160 150

Photoacoustic

140 130 120 110 100 90 80 70 60 50 40


1400


800

600


D-56

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,9 and phenothiazine 110 100 90

Photoacoustic

80 70 60 50 40 30 *Spectral Math Result PA69PTh_A50 1600

1400


800

600


D-57

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,12 and phenothiazine D.2.10 Polyamide-6,12 and phenothiazine 85 80 75 70 65

Photoacoustic

60 55 50 45 40 35 30 25 20 15 10

*Spectral Math Result PA612PTh_A55LightGrey 1600

1400


800

600

Figure D.2-25 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for Ampoule 55 (Light grey) compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6,12.

D-58

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,12 and phenothiazine 170


160 150 140 130 Photoacoustic

120 110 100 90 80 70 60 50 40 30 1600

1400


800

600

Figure D.2-26 Mid range photoacoustic FTIR spectrum in the 1700 to 400 cm-1 region for ampoule 54 compared with the model based just upon the spectral addition of the (non-linear) spectra of the constituents phenothiazine and polyamide-6,12.

D-59

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,12 and phenothiazine 70


65 60

Photoacoustic

55 50 45 40 35 30 25 20 1600

1400


800

600


D-60

Mid IR range FTIR spectra (1700 to 400 cm-1) for blends of polyamide-6,12 and phenothiazine

9


8

Photoacoustic

7

6

5

4

3 2 1600

1400


800

600


D-61

Near Infra-Red FTIR spectra for 7300 to 5700 cm-1 Appendix D (on CD only)

FOURIER TRANSFORM INFRARED SPECTRA (PART 3) Support In The Near Infrared For There Being No Polyamide/Diluent Hydrogen Bond Interactions D.3 Diffuse Reflection Infrared Fourier Transform in the Near Infrared range 7300 to 5700 cm-1. D.3.1 Introduction Appendix D comprises four sections: a) Part 1 with mid range infrared spectra of the N-H stretches and found in the file Appx_D1.pdf. b) Part 2 with mid range infrared spectra in the region 1700 to 400 cm-1 and found in the file Appx_D2.pdf. c) Part 3 with near infrared spectra in the region 7300 to 5700 cm-1 and found in this file Appx_D3.pdf. d) Part 4 with near infrared spectra in the region 5000 to 4300 cm-1 and found in the file Appx_D4.pdf. Part 3 of Appendix D displays Near Infrared (NIR) spectra in the region 7300 to 5700 cm-1 and covers work done after the Mid range IR spectra for N-H stretches of polyamide/diluent blends showed no significant peak shifts due to diluent N-H becoming bound and polyamide N-H becoming free. The work was carried out to validate the absence of hydrogen bond interactions by using another part of the IR spectrum and happens to also use a different FTIR technique. The portions of the NIR region sensitive hydrogen bonding in polyamides can be split into 7300 to 5700 cm-1 and 5000 to 4300 cm-1, as found by Wu and Siesler [135]. This part of the appendix covers the first of those NIR regions and shows that the blend hydrogen bonding overtone and combination bands are only a combination of the spectra of the polyamide and diluent.

D-62

Near Infra-Red FTIR spectra for 7300 to 5700 cm-1 NIR spectra did not require modelling because the peaks are not so sharp. Each of the spectral peaks had been moved up or down on the screen to confirm exact matches in peak frequency. There are two regions of interest in the NIR and so two graphs per sample. Graphs are presented per material combination with a minimum of 3 and a maximum of 5 representative samples representative of the available concentration range measured. D.3.2 Notation Notation for spectra in this section of Appendix D are described in Part 1 of the appendix and colours correspond to .those of the equivalent ampoule samples in Part 1 of Appendix D.

D-63

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and carbazole, 7300 to 5700 cm-1 D.3.3 Polyamide-4,6 and carbazole

0.64

carbazole polyamide-4,6 PA46Car_A57Fawn

0.62

Absorbance

0.60 0.58 0.56 0.54 0.52 0.50

7200

7000

6800


6200

6000

5800

Figure D.3-1Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 57 (Fawn) compared with the spectra for polyamide-4,6 and carbazole.

D-64

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and carbazole, 7300 to 5700 cm-1

0.280

carbazole polyamide-4,6 PA46Car_A29 NotRed

0.270 0.260

Absorbance

0.250 0.240 0.230 0.220 0.210 0.200 0.190 0.180 0.170 7200

7000

6800


6200

6000

5800

Figure D.3-2 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 29 (Not Red) compared with the spectra for polyamide-4,6 and carbazole.

D-65

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and carbazole, 7300 to 5700 cm-1 0.180

carbazole polyamide-4,6 PA46Car_A57BrightRed

0.170 0.160

Absorbance

0.150 0.140 0.130 0.120 0.110 0.100 0.090 0.080 7200

7000

6800


6200

6000

5800

Figure D.3-3 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 57 (Bright Red) compared with the spectra for polyamide-4,6 and carbazole.

D-66

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and carbazole, 7300 to 5700 cm-1 0.270 0.260

carbazole polyamide-4,6 PA46Car_A30

0.250 0.240

Absorbance

0.230 0.220 0.210 0.200 0.190 0.180 0.170 0.160 0.150 0.140 7200

7000

6800


6200

6000

5800

Figure D.3-4 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 30 compared with the spectra for polyamide-4,6 and carbazole.

D-67

Near Infra-Red FTIR spectra for blends of polyamide-6 and carbazole, 7300 to 5700 cm-1 D.3.4 Polyamide-6 and carbazole 0.490

carbazole polyamide-6 PA6Car_A61

0.480 0.470

Absorbance

0.460 0.450 0.440 0.430 0.420 0.410 0.400 0.390 0.380 7200

7000

6800


6200

6000

5800

Figure D.3-5 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 61 compared with the spectra for polyamide-6 and carbazole.

D-68

Near Infra-Red FTIR spectra for blends of polyamide-6 and carbazole, 7300 to 5700 cm-1

0.50


0.48

Absorbance

0.46

0.44

0.42

0.40

0.38

0.36 7200

7000

6800


6200

6000

5800


D-69

Near Infra-Red FTIR spectra for blends of polyamide-6 and carbazole, 7300 to 5700 cm-1 0.470 0.460


0.450 0.440

Absorbance

0.430 0.420 0.410 0.400 0.390 0.380 0.370 0.360 0.350 7200

7000

6800


6200

6000

5800


D-70


0.300


0.290 0.280

Absorbance

0.270 0.260 0.250 0.240 0.230 0.220 0.210 0.200 7200

7000

6800


6200

6000

5800


D-71

Near Infra-Red FTIR spectra for blends of polyamide-6,9 and carbazole, 7300 to 5700 cm-1 D.3.5 Polyamide-6,9 and carbazole 0.500 0.490


0.480 0.470

Absorbance

0.460 0.450 0.440 0.430 0.420 0.410 0.400 0.390 0.380 7200

7000

6800


6200

6000

5800


D-72



0.56

Absorbance

0.54 0.52 0.50 0.48 0.46 0.44 0.42 7200

7000

6800


6200

6000

5800


D-73



0.360 0.350 0.340 Absorbance

0.330 0.320 0.310 0.300 0.290 0.280 0.270 0.260 0.250 7200

7000

6800


6200

6000

5800


D-74



0.390 0.380

Absorbance

0.370 0.360 0.350 0.340 0.330 0.320 0.310 0.300 0.290 7200

7000

6800


6200

6000

5800


D-75



0.540 0.530 0.520

Absorbance

0.510 0.500 0.490 0.480 0.470 0.460 0.450 0.440 0.430 0.420 0.410 7200

7000

6800


6200

6000

5800


D-76



0.42

Absorbance

0.40

0.38 0.36

0.34

0.32

0.30

7200

7000

6800


6200

6000

5800


D-77

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and phenothiazine, 7300 to 5700 cm-1 D.3.7 Polyamide-4,6 and phenothiazine 0.410 0.400

phenothiazine polyamide-4,6 PA46PTh_A46

0.390 0.380 0.370

Absorbance

0.360 0.350 0.340 0.330 0.320 0.310 0.300 0.290 0.280 7200

7000

6800


6200

6000

5800

Figure D.3-15 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 46 compared with the spectra for polyamide-4,6 and phenothiazine. nb. This figure does not have an equivalent in Parts 1 & 2 of this appendix as the Mid range IR spectrum file was corrupted.

D-78

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and phenothiazine, 7300 to 5700 cm-1 0.410 0.400


0.390 0.380 0.370

Absorbance

0.360 0.350 0.340 0.330 0.320 0.310 0.300 0.290 0.280 0.270 7200

7000

6800


6200

6000

5800

Figure D.3-16 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 47 compared with the spectra for polyamide-4,6 and phenothiazine.

D-79

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and phenothiazine, 7300 to 5700 cm-1

0.52


0.50 0.48

Absorbance

0.46 0.44 0.42 0.40 0.38 0.36 0.34 7200

7000

6800


6200

6000

5800


nb. Parts 1 and 2 have spectra for Ampoule 8 but that ampoule sample had not been measured in the NIR.

D-80

Near Infra-Red FTIR spectra for blends of polyamide-6 and phenothiazine, 7300 to 5700 cm-1 D.3.8 Polyamide-6 and phenothiazine 0.76 0.74

phenothiazine polyamide-6 PA6PTh_A43

0.72 0.70

Absorbance

0.68 0.66 0.64 0.62 0.60 0.58 0.56 0.54 0.52 0.50 7200

7000

6800


6200

6000

5800

Figure D.3-18 Near Infrared Diffuse Reflectance FTIR spectrum in the 7300 to 5700 cm-1 region for a sample from Ampoule 43 compared with the spectra for polyamide-6 and phenothiazine.

nb. Parts 1 and 2 have spectra for Ampoule 18 (Yellow) but that ampoule sample had not been measured in the NIR.

D-81

Near Infra-Red FTIR spectra for blends of polyamide-6 and phenothiazine, 7300 to 5700 cm-1 0.430 0.420


0.410 0.400 0.390

Absorbance

0.380 0.370 0.360 0.350 0.340 0.330 0.320 0.310 0.300 0.290 7200

7000

6800


6200

6000

5800


D-82

Near Infra-Red FTIR spectra for blends of polyamide-6 and phenothiazine, 7300 to 5700 cm-1 0.370 0.360


0.350 0.340 0.330

Absorbance

0.320 0.310 0.300 0.290 0.280 0.270 0.260 0.250 0.240 7200

7000

6800


6200

6000

5800


D-83

Near Infra-Red FTIR spectra for blends of polyamide-6,9 and phenothiazine, 7300 to 5700 cm-1 D.3.9 Polyamide-6,9 and phenothiazine 0.450


0.440 0.430

Absorbance

0.420 0.410 0.400 0.390 0.380 0.370 0.360 7200

7000

6800


6200

6000

5800


D-84

Near Infra-Red FTIR spectra for blends of polyamide-6,9 and phenothiazine, 7300 to 5700 cm-1 0.260


0.255 0.250 0.245

Absorbance

0.240 0.235 0.230 0.225 0.220 0.215 0.210 0.205 0.200 0.195 7200

7000

6800


6200

6000

5800


D-85



0.400 0.390

Absorbance

0.380 0.370 0.360 0.350 0.340 0.330

7200

7000

6800


6200

6000

5800


D-86

Mid IR range FTIR graphs for blends of polyamide-6,12 and phenothiazine, 7300 to 5700 cm-1 D.3.10 Polyamide-6,12 and phenothiazine 0.450


0.440 0.430

Absorbance

0.420 0.410 0.400 0.390 0.380 0.370 0.360 0.350 0.340 7200

7000

6800


6200

6000

5800


D-87

Mid IR range FTIR graphs for blends of polyamide-6,12 and phenothiazine, 7300 to 5700 cm-1 0.58 0.56


0.54 0.52

Absorbance

0.50 0.48 0.46 0.44 0.42 0.40 0.38 0.36 0.34 0.32 7200

7000

6800


6200

6000

5800


D-88

Mid IR range FTIR graphs for blends of polyamide-6,12 and phenothiazine, 7300 to 5700 cm-1 0.42


0.40

Absorbance

0.38

0.36

0.34

0.32

0.30

0.28

7200

7000

6800


6200

6000

5800


D-89



0.340 0.330

Absorbance

0.320 0.310 0.300 0.290 0.280 0.270 0.260 0.250 0.240 7200

7000

6800


6200

6000

5800


D-90

Near Infra-Red FTIR spectra for 5000 to 4300 cm-1 Appendix D (on CD only)

FOURIER TRANSFORM INFRARED SPECTRA (PART 4) Support In The Near Infrared For There Being No Polyamide/Diluent Hydrogen Bond Interactions D.4 Diffuse Reflection Infrared Fourier Transform in the Near Infrared range 5000 to 4300 cm-1. D.4.1 Introduction Appendix D comprises four sections: a) Part 1 with mid range infrared spectra of the N-H stretches and found in the file Appx_D1.pdf. b) Part 2 with mid range infrared spectra in the region 1700 to 400 cm-1 and found in the file Appx_D2.pdf. c) Part 3 with near infrared spectra in the region 7300 to 5700 cm-1 and found in this file Appx_D3.pdf. d) Part 4 with near infrared spectra in the region 5000 to 4300 cm-1 and found in the file Appx_D4.pdf. Part 3 of Appendix D displays Near Infrared (NIR) spectra in the region 5000 to 4300 cm-1 and covers work done after the Mid range IR spectra for N-H stretches of polyamide/diluent blends showed no significant peak shifts due to diluent N-H becoming bound and polyamide N-H becoming free. The work was carried out to validate the absence of hydrogen bond interactions by using another part of the IR spectrum and happens to also use a different FTIR technique. The portions of the NIR region sensitive hydrogen bonding in polyamides can be split into 7300 to 5700 cm-1 and 5000 to 4300

cm-1,

as

found by Wu and Siesler [135]. This part of the appendix covers the first of those NIR regions and shows that the blend hydrogen bonding overtone and combination bands are only a combination of the spectra of the polyamide and diluent.

D-91

Near Infra-Red FTIR spectra for 5000 to 4300 cm-1 NIR spectra did not require modelling because the peaks are not so sharp. Each of the spectral peaks had been moved up or down on the screen to confirm exact matches in peak frequency. There are two regions of interest in the NIR and so two graphs per sample Graphs are presented per material combination with either 3 or 4 samples representative of the available concentration range measured. D.4.2 Notation Notation for spectra in this section of Appendix D are described in Part 1 of the appendix and colours correspond to .those of the equivalent ampoule samples in Part 1 of Appendix D.

D-92

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and carbazole, 5000 to 4300 cm-1 D.4.3 Polyamide-4,6 and carbazole

0.80

carbazole polyamide-4,6 PA46Car_A57Fawn

0.78

Absorbance

0.76

0.74

0.72

0.70

0.68

0.66 5000

4900

4800


4500

4400

Figure D.4-1Near Infrared Diffuse Reflectance FTIR spectrum in the 5000 to 4300 cm-1 region for a sample from Ampoule 57 (Fawn) compared with the spectra for polyamide-4,6 and carbazole.

D-93


carbazole polyamide-4,6 PA46Car_A29 NotRed

0.670 0.660

Absorbance

0.650 0.640 0.630 0.620 0.610 0.600 0.590

5000

4900

4800


4500

4400

Figure D.4-2 Near Infrared Diffuse Reflectance FTIR spectrum in the 5000 to 4300 cm-1 region for a sample from Ampoule 29(Not Red) compared with the spectra for polyamide-4,6 and carbazole.

D-94


carbazole polyamide-4,6 PA46Car_A57BrightRed

0.270

Absorbance

0.260

0.250

0.240

0.230 0.220

0.210 0.200 5000

4900

4800


4500

4400

Figure D.4-3 Near Infrared Diffuse Reflectance FTIR spectrum in the 5000 to 4300 cm-1 region for a sample from Ampoule 57 (Bright Red) compared with the spectra for polyamide-4,6 and carbazole.

D-95



0.330 0.320

Absorbance

0.310 0.300 0.290 0.280 0.270 0.260 0.250 0.240 5000

4900

4800


4500

4400


D-96

Near Infra-Red FTIR spectra for blends of polyamide-6 and carbazole, 5000 to 4300 cm-1 D.4.4 Polyamide-6 and carbazole 0.575 0.570


0.565 0.560

Absorbance

0.555 0.550 0.545 0.540 0.535 0.530 0.525 0.520 0.515 5000

4900

4800


4500

4400


D-97


0.610


0.600

Absorbance

0.590

0.580

0.570

0.560 0.550

0.540

5000

4900

4800


4500

4400


D-98

Near Infra-Red FTIR spectra for blends of polyamide-6 and carbazole, 5000 to 4300 cm-1 0.280


0.270 0.260

Absorbance

0.250 0.240 0.230 0.220 0.210 0.200 5000

4900

4800


4500

4400


D-99

Near Infra-Red FTIR spectra for blends of polyamide-6 and carbazole, 5000 to 4300 cm-1 0.420


0.410 0.400 0.390 Absorbance

0.380 0.370 0.360 0.350 0.340 0.330 0.320 0.310 5000

4900

4800


4500

4400


D-100



0.540

Absorbance

0.535 0.530 0.525 0.520 0.515 0.510 0.505 5000

4900

4800


4500

4400


D-101



0.700 0.690 0.680

Absorbance

0.670 0.660 0.650 0.640 0.630 0.620 0.610 0.600 0.590 5000

4900

4800


4500

4400


D-102



0.450 0.440

Absorbance

0.430 0.420 0.410 0.400 0.390 0.380 0.370 0.360 5000

4900

4800


4500

4400


D-103



0.450

Absorbance

0.445 0.440 0.435 0.430 0.425 0.420 0.415 0.410 5000

4900

4800


4500

4400


D-104



0.630

Absorbance

0.620 0.610 0.600 0.590 0.580 0.570

5000

4900

4800


4500

4400


D-105



0.60 0.58

Absorbance

0.56 0.54 0.52 0.50 0.48 0.46 0.44 5000

4900

4800


4500

4400


D-106

Near Infra-Red FTIR spectra for blends of polyamide-4,6 and phenothiazine, 5000 to 4300 cm-1 D.4.7 Polyamide-4,6 and phenothiazine 0.560 0.550


0.540 0.530 0.520

Absorbance

0.510 0.500 0.490 0.480 0.470 0.460 0.450 0.440 0.430 0.420 5000

4900

4800


4500

4400

Figure D.4-15 Near Infrared Diffuse Reflectance FTIR spectrum in the 5000 to 4300 cm-1 region for a sample from Ampoule 46 compared with the spectra for polyamide-4,6 and phenothiazine. nb. This figure does not have an equivalent in Parts 1 & 2 of this appendix as the Mid range IR spectrum file was corrupted.

D-107



0.510 0.500

Absorbance

0.490 0.480 0.470 0.460 0.450 0.440 0.430 0.420 5000

4900

4800


4500

4400


D-108



0.62

Absorbance

0.60 0.58 0.56 0.54 0.52 0.50 5000

4900

4800


4500

4400


nb. Parts 1 and 2 have spectra for Ampoule 8 but that ampoule sample had not been measured in the NIR.

D-109

Near Infra-Red FTIR spectra for blends of polyamide-6 and phenothiazine, 5000 to 4300 cm-1 D.4.8 Polyamide-6 and phenothiazine 1.04 1.02


1.00 0.98 0.96 0.94 Absorbance

0.92 0.90 0.88 0.86 0.84 0.82 0.80 0.78 0.76 0.74 0.72 5000

4900

4800


4500

4400


nb. Parts 1 and 2 have spectra for Ampoule 18 (Yellow) but that ampoule sample had not been measured in the NIR.

D-110

Near Infra-Red FTIR spectra for blends of polyamide-6 and phenothiazine, 5000 to 4300 cm-1 0.520


0.510

0.500

Absorbance

0.490

0.480

0.470

0.460

0.450 0.440 5000

4900

4800


4500

4400


D-111

Near Infra-Red FTIR spectra for blends of polyamide-6 and phenothiazine, 5000 to 4300 cm-1 0.470


0.460 0.450

Absorbance

0.440 0.430 0.420 0.410 0.400 0.390

5000

4900

4800


4500

4400


D-112

Near Infra-Red FTIR spectra for blends of polyamide-6,9 and phenothiazine, 5000 to 4300 cm-1 D.4.9 Polyamide-6,9 and phenothiazine 0.510


0.505 0.500

Absorbance

0.495 0.490 0.485 0.480 0.475 0.470 0.465 0.460 5000

4900

4800


4500

4400

Figure D.4-21 Near Infrared Diffuse Reflectance FTIR spectrum in the 5000 to 4300 cm-1 region for a sample from Ampoule 52 compared with the spectra for polyamide-6,9 and phenothiazine.>>>>>>>>>>>> ZZZZ note that the sample peak being slightly to the right near 4875 cm-1 is due to the steep slope for the pth at that section.

D-113

Near Infra-Red FTIR spectra for blends of polyamide-6,9 and phenothiazine, 5000 to 4300 cm-1

0.360


0.350

Absorbance

0.340 0.330 0.320 0.310 0.300 0.290 0.280 5000

4900

4800


4500

4400


D-114



0.520 0.510 0.500

Absorbance

0.490 0.480 0.470 0.460 0.450 0.440 0.430 0.420 0.410 0.400 5000

4900

4800


4500

4400


D-115

Mid IR range FTIR graphs for blends of polyamide-6,12 and phenothiazine, 5000 to 4300 cm-1 D.4.10 Polyamide-6,12 and phenothiazine 0.520 0.515


0.510 0.505

Absorbance

0.500 0.495 0.490 0.485 0.480 0.475 0.470 0.465 5000

4900

4800


4500

4400


D-116



0.76 0.74 0.72 0.70 Absorbance

0.68 0.66 0.64 0.62 0.60 0.58 0.56 0.54 0.52 0.50 0.48 5000

4900

4800


4500

4400


D-117



0.520 0.510

Absorbance

0.500 0.490 0.480 0.470 0.460 0.450 0.440 0.430 0.420 5000

4900

4800


4500

4400


D-118

Mid IR range FTIR graphs for blends of polyamide-6,12 and phenothiazine, 5000 to 4300 cm-1 0.430


0.420

Absorbance

0.410 0.400 0.390 0.380 0.370 0.360 0.350 5000

4900

4800


4500

4400


D-119

Bibliography 1.

Myers, M.E., Wims, A.M., Ellis, T.S., and Barnes, J., An investigation of the morphology of amorphous/semicrystalline Nylon blends using small-angle x-ray scattering. Macromolecules, 1990. 23(10): 2807-2814.

2.

Liu, L.-Z., Chu, B., Penning, J.P., and Manley, R.S.J., Synchrotron SAXS study of miscible blends of semicrystalline poly(vinylidene fluoride) and semicrystalline poly(1,4-butylene adipate). Macromolecules, 1997. 30(15): 4398-4404.

3.

Chen, H.-L. and Hsiao, M.-S., Morphological structure induced by combined crystallization and liquid-liquid demixing in poly(ethylene terephthalate)/poly(ether imide) blends. Macromolecules, 1998. 31(19): 6579-6584.

4.

You, J.-W., Chiu, H.-J., and Don, T.-M., Spherulitic morphology and crystallization kinetics of melt-miscible blends of poly(3hydroxybutyrate) with low molecular weight poly(ethylene oxide). Polymer, 2003. 44(15): 4355-4362.

5.

Kim, K.J., Cho, H.W., and Yoon, K.J., Effect of P(MMA-co-MAA) compatibilizer on the miscibility of Nylon 6/PVDF blends. European Polymer Journal, 2003. 39(6): 1249-1265.

6.

Qiu, Z., Ikehara, T., and Nishi, T., Miscibility and crystallization in crystalline/crystalline blends of poly(butylene succinate)/poly(ethylene oxide). Polymer, 2003. 44(9): 2799-2806.

7.

van der Heijden, P.C., A DSC-study on the demixing of binary polymer solutions. Thesis, 2001, Universiteit Twente: Enschede, NL.

8.

Cha, B.J., Char, K., Kim, J.J., Kim, S.S., and Kim, C.K., The effects of diluent molecular weight on the structure of thermally-induced phase separation membrane. J. Membrane Sci., 1995. 108(3): 219-229.

9.

Utracki, L.A., Polymer/Polymer miscibility Part 2, in Polymer Alloys and Blends. 1989, Hanser (Karl Hanser Verlag) & Oxford University Press: Vienna/NewYork. 29-69.

10.

Kim, W.K., Char, K., and Kim, C.K., Control of droplet size of polymerdiluent blends through thermally induced phase separation. J. Polym. Sci., Part B: Polym. Phys., 2000. 38(23): 3042-3052.

11.

IUPAC Compendium of chemical terminology. 2nd ed. 1997.

12.

Avrami, M., Kinetics of phase change. I. General theory. J. Chem. Phys., 1939. 7: 1103-1112.

13.

Avrami, M., Kinetics of phase change. II. Transformation-time relations for random distribution of nuclei. J. Chem. Phys., 1940. 8: 212-224. 338

14.

Avrami, M., Granulation, phase change and microstructure. Kinetics of phase change. III. J. Chem. Phys., 1941. 9: 177-184.

15.

Zhong, Z. and Guo, Q., Crystallization kinetics of miscible thermosetting polymer blends of novolac resin and poly(ethylene oxide). Polymer, 1999. 41(5): 1711-1718.

16.

Weng, W., Chen, G., and Wu, D., Crystallization kinetics and melting behaviors of Nylon 6/foliated graphite nanocomposites. Polymer, 2003. 44(26): 8119-8132.

17.

Clarke, N., McLeish, T.C.B., Pavawongsak, S., and Higgins, J.S., Viscoelastic effects on the early stages of phase separation in polymer blends. Macromolecules, 1997. 30(15): 4459-4463.

18.

Lin, Y.H. and Juang, J.H., Onset of entanglement. Macromolecules, 1999. 32(1): 181-185.

19.

Keller, A., Single crystals in polymers: evidence of a folded-chain configuration. Philosophical Magazine, 1957. 2: 1171-1175.

20.

Ishikiriyama, K. and Wunderlich, B., Melting of poly(oxyethylene) analyzed by Temperature-Modulated Calorimetry. Macromolecules, 1997. 30(14): 4126-4131.

21.

Paternostre, L., Damman, P., and Dosiere, M., Time-resolved SAXS study of the crystallization and melting of poly(ethylene oxide) molecular complexes. Macromolecules, 1997. 30(13): 3946-3948.

22.

Hsiao, B.S. New insight into the crystallization behavior of polymers and fibers via time resolved synchrotron X-ray measurements and novel data analysis technique. in CCP13 Workshop. 1997. Daresbury, UK.

23.

Weeks, J.J., Melting temperatures and change of the lamellar thickness of polyethylene with time. J. Research Natl. Bur. Standards, A, 1963. 67: 441-451.

24.

Hoffman, J.D. and Weeks, J.J., X-ray study of isothermal thickening of lamellae in bulk polyethylene at the crystallization temperature. J. Chem. Phys., 1965. 42(12): 4301-4302.

25.

Marand, H., Xu, J., and Srinivas, S., Determination of the equilibrium melting temperature of polymer crystals: linear and nonlinear HoffmanWeeks extrapolations. Macromolecules, 1998. 31(23): 8219-8229.

26.

Welch, P. and Muthukumar, M., Molecular mechanisms of polymer crystallization from solution. Phys. Rev. Lett., 2001. 87(21): 218302-.

27.

Lauritzen, J.I., Jr. and Hoffman, J.D., Theory of formation of polymer crystals with folded chains in dilute solution. J. Research Natl. Bur. Standards, 1960. 64A(1): 73-102. 339

28.

Hoffman, J.D. and Lauritzen, J.I., Jr., Crystallization of bulk polymers with chain folding: theory of growth of lamellar spherulites. J. Research Natl. Bur. Standards, 1961. 65A: 297-336.

29.

Androsch, R. and Wunderlich, B., A study of annealing of poly(ethylene-co-octene) by Temperature-Modulated and Standard Differential Scanning Calorimetry. Macromolecules, 1999. 32(21): 7238-7247.

30.

Olmsted, P.D., Poon, W.C.K., McLeish, T.C.B., Terrill, N.J., and Ryan, A.J., Spinodal-assisted crystallization in polymer melts. Phys. Rev. Lett., 1998. 81(2): 373-376.

31.

Ezquerra, T.A., Lopez-Cabarcos, E., Hsiao, B.S., and Balta-Calleja, F.J., Precursors of crystallization via density fluctuations in stiff-chain polymers. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top., 1996. 54(1): 989-992.

32.

Toma, L., Toma, S., and Subirana, J.A., Simulation of polymer crystallization through a dynamic Monte Carlo lattice model. Macromolecules, 1998. 31(7): 2328-2334.

33.

Jiang, Y., Gu, Q., Li, L., Shen, D.-Y., Jin, X.-G., and Chan, C.-M., Conformational changes in the induction period of crystallization as measured by FT-IR. Polymer, 2003. 44(12): 3509-3513.

34.

Rabani, E., Reichman, D.R., Geissler, P.L., and Brus, L.E., Dryingmediated self-assembly of nanoparticles. Nature, 2003. 426: 271 - 274.

35.

Wurm, A. and Schick, C., Development of thermal stability of polymer crystals during isothermal crystallization. e-Polymers [www.epolymers.org], 2002: Paper No. 24 wurm_180502.pdf 15 pages,.

36.

Doye, J.P.K. and Frenkel, D., The mechanism of thickness selection in the Sadler-Gilmer model of polymer crystallization. J. Chem. Phys., 1999. 110(14): 7073-7086.

37.

Supaphol, P. and Spruiell, J.E., Nonisothermal bulk crystallization of high-density polyethylene via a modified depolarized light microscopy technique: further analysis. J. Appl. Polym. Sci., 2002. 86(4): 10091022.

38.

Bhattarai, N., Kim, H.Y., Cha, D.I., Lee, D.R., and Yoo, D.I., Nonisothermal crystallization and melting behavior of the copolymer derived from p-dioxanone and poly(ethylene glycol). European Polymer Journal, 2003. 39(7): 1365-1375.

39.

Wagner, J. and Phillips, P.J., The mechanism of crystallization of linear polyethylene, and its copolymers with octene, over a wide range of supercoolings. Polymer, 2001. 42(21): 8999-9013.

340

40.

Okazaki, I. and Wunderlich, B., Reversible melting in polymer crystals detected by Temperature-Modulated Differential Scanning Calorimetry. Macromolecules, 1997. 30(6): 1758-1764.

41.

Smith, P. and Manley, R.S.J., Solid solution formation and fractionation in quasi-binary systems of polyethylene fractions. Macromolecules, 1979. 12(3): 483-491.

42.

Huang, X.-D., Goh, S.H., Lee, S.Y., Zhao, Z.D., Wong, M.W., and Huan, C.H.A., Miscibility and interactions in blends and complexes of poly[2-(dimethylamino)ethyl methacrylate] with poly(p-vinylphenol). Macromolecules, 1999. 32(13): 4327-4331.

43.

Flory, P.J., Thermodynamics of high-polymer solutions. J. Chem. Phys., 1942. 10: 51-61.

44.

Huggins, M.L., Theory of solution of high polymers. J. Am. Chem. Soc., 1942. 64: 1712-1719.

45.

Huggins, M.L., Thermodynamic properties of solutions of long-chain compounds. Annals of the New York Academy of Sciences, 1942. 43(Art. 1): 1-32.

46.

Huggins, M.L., Some properties of solutions of long-chain compounds. J. Phys. Chem., 1942. 46: 151-158.

47.

Flory, P.J., Principles of Polymer Chemistry. 1953.

48.

Aharoni, S.M., n-Nylons: Their synthesis, structure and properties. 1997, Chichester, UK: John Wiley and Sons.

49.

Hu, W., Frenkel, D., and Mathot, V.B.F., Lattice-model study of the thermodynamic interplay of polymer crystallization and liquid-liquid demixing. J. Chem. Phys., 2003. 118(22): 10343-10348.

50.

Kelley, F.N. and Bueche, F., Viscosity and glass-temperature relations for polymer-diluent systems. J. Polym. Sci., 1961. 50: 549-556.

51.

Carothers, W.H., Polymerization. Chemical Reviews, 1931. 8: 353-426.

52.

Elias, H.-G., Introduction to plastics. 1993, VCH: New York.

53.

Woebeken, W., Saechtling international plastics handbook for technologists and users. 3rd ed, ed. Hanser. 1995, Cincinnati: Gardner Publishers. 256-257.

54.

Li, D., Wang, H.-C., and Lee, A.F., Polyamides (impact toughness and toughening), in Polymeric Materials Encyclopedia, J.C. Salamone, Editor. 1996, CRC Press: New York. 5409-5418.

55.

Roerdink, E. and Havenith, H., Polyamide 46. A material for high performance application // Polyamid 46: Ein werkstoff fur HighPerformance Anwendungen. Kunststoffe, 1997. 87(9): 1114-1116. 341

56.

Cunliffe, D., Pennadam, S., and Alexander, C., Synthetic and biological polymers--merging the interface. European Polymer Journal, 2004. 40(1): 5-25.

57.

Munoz-Guerra, S., Rodriguez-Galan, A., Fernandez-Santin, J.M., Lopez-Carrasquerro, F., Garcia-Alvarez, M., Bella, J., and Subirana, J.A., Poly(α-alkyl-β−L-aspartate)s: A family of poly(β-amide)s with structure and prpoerties intermediate between Nylons and polypeptides. Trends in Macromol. Res, 1994. 1: 181-200.

58.

Subirana, J.A., Elucidation of chain folding in polymer crystals: comparison with proteins. Trends in Polymer Science, 1997. 5(10): 321-326.

59.

Subirana, J.A. Helical polyamides and rings: A bridge between Nylons and proteins. in XVII Conference of the International Union of Crystallography Aug 8-17 1996. 1996. Seattle, USA.

60.

Skrovanek, D.J., Howe, S.E., Painter, P.C., and Coleman, M.M., Hydrogen bonding in polymers: Infrared temperature studies of an amorphous polyamide. Macromolecules, 1985. 18(9): 1676-1683.

61.

Skrovanek, D.J., Painter, P.C., and Coleman, M.M., Hydrogen bonding in polymers. 2. Infrared temperature studies of Nylon 11. Macromolecules, 1986. 19(3): 699-705.

62.

Coleman, M.M., Lee, K.H., Skrovanek, D.J., and Painter, P.C., Hydrogen bonding in polymers. 4. Infrared temperature studies of a simple polyurethane. Macromolecules, 1986. 19(8): 2149-2157.

63.

Jones, N.A., Atkins, E.D.T., and Hill, M.J., Investigation of solutiongrown, chain-folded lamellar crystals of the even-even Nylons: 6 6, 8 6, 8 8, 10 6, 10 8, 10 10, 12 6, 12 8, 12 10, and 12 12. J. Polym. Sci., Part B: Polym. Phys., 2000. 38(9): 1209-1221.

64.

Franco, L., Cooper, S.J., Atkins, E.D.T., Hill, M.J., and Jones, N.A., Nylon 6 9 can crystallize with hydrogen bonding in two and in three interchain directions. J. Polym. Sci., Part B: Polym. Phys., 1998. 36(7): 1153-1165.

65.

Navarro, E., Franco, L., Subirana, J.A., and Puiggali, J., Nylon 65 has a unique structure with two directions of hydrogen bonds. Macromolecules, 1995. 28(26): 8742-8750.

66.

Aceituno, J.E., Tereshko, V., Lotz, B., and Subirana, J.A., Synthesis and characterization of polyamides n,3. Macromolecules, 1996. 29(6): 1886-1893.

67.

Puiggali, J., Acietuno, J.E., Navarro, E., Campos, J.L., and Subirana, J.A., Structure of n,3 polyamides, a group of Nylons with two spatial hydrogen bond orientations. Macromolecules, 1996. 29: 8170-8179. 342

68.

Navarro, E., Subirana, J.A., and Puiggali, J., The structure of Nylon 12,5 is characterized by two hydrogen bond directions as are other polyamides derived from glutaric acid. Polymer, 1997. 38(13): 3429-3432.

69.

Puiggali, J., Franco, L., Aleman, C., and Subirana, J.A., Crystal structures of Nylon 5,6. A model with two hydrogen bond directions for Nylons derived from odd diamines. Macromolecules, 1998. 31(24): 8540-8548.

70.

Franco, L., Subirana, J.A., and Puiggali, J., Structure and morphology of odd polyoxamides [Nylon 9,2]. A new example of hydrogen-bonding interactions in two different directions. Macromolecules, 1998. 31(12): 3912-3924.

71.

Villasenor, P., Franco, L., Subirana, J.A., and Puiggali, J., On the crystal structure of odd-even Nylons: polymorphism of Nylon 5,10. J. Polym. Sci., Part B: Polym. Phys., 1999. 37(17): 2383-2395.

72.

Mathias, L.J., Powell, D.G., Autran, J.P., and Porter, R.S., Nitrogen-15 NMR characterization of multiple crystal forms and phase transitions in polyundecanamide (Nylon 11). Macromolecules, 1990. 23(4): 963-967.

73.

Atkins, E.D.T., Hill, M., Hong, S.K., Keller, A., and Organ, S., Lamellar structure and morphology of Nylon 46 crystals: a new chain folding mechanism for Nylons. Macromolecules, 1992. 25(2): 917-924.

74.

Roberts, M.F. and Jenekhe, S.A., Site-specific reversible scission of hydrogen bonds in polymers: an investigation of polyamides and their Lewis acid-base complexes by infrared spectroscopy. Macromolecules, 1991. 24(11): 3142-3146.

75.

Franco, L., Navarro, E., Subirana, J.A., and Puiggali, J., Incorporation of glycine residues in even-even Nylons disrupts their characteristic alltrans conformation. Polymer, 1998. 39(22): 5553-5560.

76.

Simal, A.L. and Martin, A.R., Structure of heat-treated Nylon 6 and 6.6 fibers. I. The shrinkage mechanism. J. Appl. Polym. Sci., 1998. 68(3): 441-452.

77.

Simal, A.L. and Martin, A.R., Structure of heat-treated Nylon 6 and 6.6 fibers. II. Recrystallization mechanism. J. Appl. Polym. Sci., 1998. 68(3): 453-474.

78.

Williams, J.C.L., Watson, S.J., and Boydell, P., Properties, in Nylon Plastics Handbook, M.I. Kohan, Editor. 1995, Hanser Publishers: Cincinnati, Ohio, USA. 291-360.

79.

Bermudez, M., Leon, S., Aleman, C., and Munoz-Guerra, S., Structure and morphology of Nylon 46 lamellar crystals: electron microscopy and energy calculations. J. Polym. Sci., Part B: Polym. Phys., 2000. 38(1): 41-52. 343

80.

Stepaniak, R.F., Garton, A., Carlsson, D.J., and Wiles, D.M., An examination of the crystal structures present in Nylon-6 fibers. J. Polym. Sci., Polym. Phys. Ed., 1979. 17(6): 987-999.

81.

Arimoto, H., α-γ transition of Nylon 6. J. Polym. Sci., Part A: Gen Pap, 1964. 3: 2283.

82.

Johnson, C.G. and Mathias, L.J., Structural characterization of Nylon 7 by solid-state nuclear magnetic resonance, differential scanning calorimetry and attenuated total reflectance Fourier transform infrared spectroscopy. Polymer, 1994. 35(1): 66-74.

83.

Ramesh, C., Crystalline transitions in Nylon 12. Macromolecules, 1999. 32(17): 5704-5706.

84.

Goyal, M., Inoue, T., Kakimoto, M.-A., and Imai, Y., High-pressure synthesis and properties of aliphatic-aromatic polyimides via Nylon-salttype monomers derived from aliphatic diamines and benzophenonetetracarboxylic acid. J. Polym. Sci., Part A: Polym. Chem., 1998. 36(1): 39-47.

85.

Lu, Y., Huang, R., and Xie, B., Pressure-induced crystallization of polyamide-6 from the melt. Gaoya Wuli Xuebao (in Chinese), 1998. 12(4): 298-302.

86.

Lu, Y., Huang, R., Cai, B., and Fan, W., Morphology and structure of polyamide-6 crystallized under high pressure. Gongneng Cailiao (in Chinese), 1998. 29(Suppl.): 1138-1140, 1144.

87.

Gogolewski, S. and Pennings, A.J., Crystallization of polyamides under elevated pressure. 3. The morphology and structure of pressurecrystallized Nylon-6 (polycapramide). Polymer, 1977. 18(7): 647-653.

88.

Gogolewski, S. and Pennings, A.J., Crystallization of polyamides under elevated pressure. 4. Annealing of Nylon-6 (polycapramide) under pressure. Polymer, 1977. 18(7): 654-659.

89.

Hirschinger, J., Miura, H., Gardner, K.H., and English, A.D., Segmental dynamics in the crystalline phase of Nylon 66: solid state deuterium NMR. Macromolecules, 1990. 23(8): 2153-2169.

90.

Ramesh, C., Keller, A., and Eltink, S.J.E.A., Studies on the crystallization and melting of Nylon-6,6: 1. The dependence of the Brill transition on the crystallization temperature. Polymer, 1994. 35(12): 2483-2487.

91.

Vasanthan, N., Murthy, N.S., and Bray, R.G., Investigation of Brill Transition in Nylon 6 and Nylon 6,6 by infrared spectroscopy. Macromolecules, 1998. 31(23): 8433-8435.

92.

Cooper, S.J., Atkins, E.D.T., and Hill, M.J., Temperature-induced changes in lamellar crystals of monodisperse Nylon 6 and Nylon 6 6 oligoamides. Macromolecules, 1998. 31(25): 8947-8956. 344

93.

Jones, N.A., Cooper, S.J., Atkins, E.D.T., Hill, M.J., and Franco, L., Temperature-induced changes in chain-folded lamellar crystals of aliphatic polyamides. Investigation of Nylons 2 6, 2 8, 2 10, and 2 12. J. Polym. Sci., Part B: Polym. Phys., 1997. 35(4): 675-688.

94.

Jones, N.A., Atkins, E.D.T., Hill, M.J., Cooper, S.J., and Franco, L., Polyamides with a choice of structure and crystal surface chemistry. Studies of chain-folded lamellae of Nylons 8 10 and 10 12 and comparison with the other 2N 2(N + 1) Nylons 4 6 and 6 8. Macromolecules, 1997. 30(12): 3569-3578.

95.

Atkins, E.D.T., Hill, M.J., Jones, N.A., and Cooper, S.J., Structure and morphology of Nylon 2,4 lamellar crystals: comparison with polypeptides and relationship with other Nylons. J. Polym. Sci., Part B: Polym. Phys., 1998. 36(13): 2401-2412.

96.

Jones, N.A., Atkins, E.D.T., and Hill, M.J., Comparison of structures and behavior on heating of solution-grown, chain-folded lamellar crystals of 31 even-even Nylons. Macromolecules, 2000. 33(7): 2642-2650.

97.

Li, Y., Yan, D., and Zhu, X., Crystalline transition in Nylon 10 10. Macromol. Rapid Commun., 2000. 21(18): 1282-1285.

98.

Huang, Y., Li, W., and Yan, D., Crystalline transition behaviors of Nylons 12 20 and 10 20. European Polymer Journal, 2003. 39(6): 1133-1140.

99.

Kohan, M.I., ed. Nylon plastics handbook. 1995, Hanser Publishers: Cincinnati, Ohio.

100. Qiu, Z., Komura, M., Ikehara, T., and Nishi, T., Poly(butylene succinate)/poly(vinyl phenol) blends. Part 1. Miscibility and crystallization. Polymer, 2003. 44(26): 8111-8117. 101. Bhandari, B.R. and Hartel, R.W., Co-crystallization of sucrose at high concentration in the presence of glucose and fructose. J. Food Sci., 2002. 67(5): 1797-1802. 102. Kristiansen, M., Werner, M., Tervoort, T., Smith, P., Blomenhofer, M., and Schmidt, H.-W., The binary system isotactic polypropylene/Bis(3,4dimethylbenzylidene)sorbitol: Phase behavior, nucleation, and optical properties. Macromolecules, 2003. 36 .(14): 5150-5156. 103. Simek, L., Dostal, J., and Bohdanecky, M., Morphological factor in the melting point depression of polypropylene by alkanes. Polymer, 2001. 42(21): 8897-8900. 104. Kim, J.K. and Kim, B.-K., Effect of the liquid-liquid phase separation on the crystallization behavior and mechanical properties of poly(ethyleneran-vinyl acetate) and paraffin wax blend. J. Polym. Sci., Part B: Polym. Phys., 1999. 37(16): 1991-2005. 345

105. Rocco, A.M., Moreira, D.P., and Pereira, R.P., Specific interactions in blends of poly(ethylene oxide) and poly(bisphenol A-co-epichlorohydrin): FTIR and thermal study. European Polymer Journal, 2003. 39(9): 1925-1934. 106. Dormidontova, E. and ten Brinke, G., Phase behavior of hydrogenbonding polymer-oligomer mixtures. Macromolecules, 1998. 31(8): 2649-2660. 107. Kobori, Y., Yasumitsu, R., Akiba, I., Akiyama, S., and Sano, H., Interfacial irregularity induced by hydrogen bonds at the interface in immiscible polymer blends. e-Polymers, 2004: No pp. given. 108. Moon, H.-S., Ryoo, B.-K., and Jung-Ki, P., Concurrent crystallization in polypropylene/Nylon-6 blends using maleic anhydride grafted polypropylene as a compatibilizing agent. J. Polym. Sci., Part B: Polym. Phys., 1994. 32(9): 1427-1435. 109. Jafari, S.H., Gupta, A.K., and Rana, S.K., Effect of Nylon 6 inclusions on the crystalline morphology of polypropylene-Nylon 6 blends. J. Appl. Polym. Sci., 2000. 75(14): 1769-1775. 110. Murthy, N.S., Wang, Z.h.G., Akkapeddi, M.K., and Hsiao, B.S., Isothermal crystallization kinetics of Nylon 6, blends and copolymers using simultaneous small and wide[hyphen]angle X-ray measurements. Polymer, 2002. 43(18): 4905-4913. 111. Li, Y. and Kaito, A., Crystallization and orientation behaviors of poly(vinylidene fluoride) in the oriented blend with Nylon 11. Polymer, 2003. 44(26): 8167-8176. 112. Damman, P. and Point, J.J., Crystallization and morphology of poly(ethylene oxide-p-nitrophenol molecular complex crystallized from the melt. Macromolecules, 1993. 26(7): 1722-1728. 113. Damman, P. and Point, J.J., Crystalline structure of the poly(ethylene oxide)-p-nitrophenol complex. 2. Fourier Transform Infrared Spectroscopy. Macromolecules, 1994. 27(14): 3919-3925. 114. Paternostre, L., Damman, P., Dosiere, M., and Bourgaux, C., Integral and nonintegral folding in lamellar crystals of molecular complexes of poly(ethylene oxide). Macromolecules, 1996. 29(6): 2046-2052. 115. Kawaguchi, A., Structure of iodine-Nylon 6 complex. 1. The investigation of the lattice constants and hydrostatic compression of the complex crystal. Polymer, 1992. 33(18): 3981-3984. 116. Kawaguchi, A., Structure of iodine-Nylon 6 complex: 2. Arrangement of polyiodides in the complex. Polymer, 1994. 35(12): 2665-2671.

346

117. Kawaguchi, A., Structure of iodine-Nylon 6 complex: 3. Modification of double orientation of the complex with doping. Polymer, 1994. 35(17): 3797-3798. 118. Murthy, N.S., Structure of iodide ions in iodinated Nylon 6 and the evolution of hydrogen bonds between parallel chains in Nylon 6. Macromolecules, 1987. 20(2): 309-316. 119. Penel-Pierron, L., Depecker, C., Seguela, R., and Lefebvre, J.M., Structural and mechanical behavior of Nylon 6 films. Part 1. Identification and stability of the crystalline phases. J. Polym. Sci., Part B: Polym. Phys., 2001. 39(5): 484-495. 120. Joussot-Dubien, C., Engel, R., and Vert, M., Low-temperature extrusion and properties of solid polyamide-6/calcium chloride complexes. Polymer, 1994. 35(17): 3691-3697. 121. Johnson, C.G. and Mathias, L.J., Nylon 13,13 analysis by X-ray and solid state NMR: treatment-dependent crystal forms. Polymer, 1993. 34(23): 4978-4981. 122. Gaymans, R.J., van Utteren, T.E.C., van den Berg, J.W.A., and Schuyer, J., Preparation and some properties of Nylon 46. J. Polym. Sci., Polym. Chem. Ed., 1977. 15(3): 537-545. 123. Haerschnitz, R., Heather, P., Derks, W., and van Leeuwendal, R., Nylon 46 - a design material for demanding engineering applications. Kunststoffe, 1990. 80(11): 1272-1276. 124. Paternostre, L., Damman, P., and Dosiere, M., Morphology and crystal structure of the poly(ethylene oxide)-hydroquinone molecular complex. J. Polym. Sci., Part B: Polym. Phys., 1999. 37(12): 1197-1208. 125. Paternostre, L., Damman, P., and Dosiere, M., Polymorphism and crystal morphology of poly(ethylene oxide)-2-methyl resorcinol supramolecular complexes. Macromolecules, 1999. 32(1): 153-161. 126. Handbook of chemistry and physics. 55th ed, ed. R.C. Weast. 1975, Cleveland, Ohio, USA: CRC Press. 127. Holland, B.J. and Hay, J.N., Thermal degradation of Nylon polymers. Polymer International, 2000. 49(9): 943-948. 128. Wunderlich, B., Thermal analysis. 1990: Academic Press. 129. TA Instruments DSC 2920 Differential Scanning Calorimeter Manual. 1997. 130. Androsch, R., Moon, I., Kreitmeier, S., and Wunderlich, B., Determination of heat capacity with a sawtooth-type, powercompensated temperature-modulated DSC. Thermochim. Acta, 2000. 357-358: 267-278. 347

131. Reading, M., Hahn, B.K., and Crowe, B.S., Method and apparatus for modulated differential analysis. 1993, TA Instruments, Inc. (New Castle, DE): US. 132. Merzlyakov, M. and Schick, C., Complex heat capacity measurements by TMDSC. Part 1. Influence of non-linear thermal response. Thermochim. Acta, 1999. 330(1-2): 55-64. 133. Simon, S.L., Temperature-modulated differential scanning calorimetry: theory and application. Thermochim. Acta, 2001. 374(1): 55-71. 134. Jiang, Z., Imrie, C.T., and Hutchinson, J.M., An introduction to temperature modulated differential scanning calorimetry (TMDSC): a relatively non-mathematical approach. Thermochim. Acta, 2002. 387(1): 75-93. 135. Berghmans, H., Thermal transitions and gelation in polymer solutions, in Calorimetry and Thermal Analysis of Polymers, V. Mathot, Editor. 1993, Hanser Publishers. 206-230. 136. Visjager, J., Tervoort, T.A., and Smith, P., Solid-solution formation and phase segregation in binary systems of homologous extended-chain perfluorinated alkanes. Polymer, 1999. 40(16): 4533-4542. 137. Johnson, C.G. and Mathias, L.J., Solid-state NMR characterization of copolymers of Nylon 11 and Nylon 12. Solid State Nucl. Magn. Reson., 1997. 8(3): 161-171. 138. Paredes, N., Rodriguez-Galan, A., and Puiggalf, J., Synthesis and characterization of a family of biodegradable poly(ester amide)s derived from glycine. J. Polym. Sci., Part A: Polym. Chem., 1998. 36(8): 1271-1282. 139. Moilanen, A.M., Hormi, O.E.O., and Taskinen, K.A., Unexpected strengthening of polyamide 11 with liquid crystalline oligomers of 2alkoxy-4-hydroxybenzoic acids. 1. Macromolecules, 1998. 31(24): 8595-8599. 140. Quintanilla, L. and Pastor, J.M., Structural analysis of polyamide-6,6 reinforced with glass fiber by the use of Fourier transform infra-red spectroscopy with photoacoustic detection and differential scanning calorimetry. Polymer, 1994. 35(24): 5241-5246. 141. Wu, P. and Siesler, H.W., The assignment of overtone and combination bands in the near infrared spectrum of polyamide 11. J. Near Infrared Spectrosc., 1999. 7(2): 65-76. 142. Guerra, G., Choe, S., Williams, D.J., Karasz, F.E., and MacKnight, W.J., Fourier transform infrared spectroscopy of some miscible polybenzimidazole/polyimide blends. Macromolecules, 1988. 21(1): 231-234.

348

143. Xu, Y.-Z., Sun, W.-X., Li, W.-H., Hu, X.-B., Zhou, H.-B., Weng, S.-F., Zhang, F., Zhang, X.-X., Wu, J.-G., Xu, D.-F., and Xu, G.-X., Investigation on the interaction between polyamide and lithium salts. J. Appl. Polym. Sci., 2000. 77(12): 2685-2690. 144. Gao, Q. and Scheinbeim, J.I., Dipolar intermolecular interactions, structural development, and electromechanical properties in ferroelectric polymer blends of Nylon-11 and poly(vinylidene fluoride). Macromolecules, 2000. 33(20): 7564-7572. 145. Wang, F.-Y., Ma, C.-C.M., and Wu, H.-D., Hydrogen bonding in polyamide toughened novolac type phenolic resin. J. Appl. Polym. Sci., 1999. 74(9): 2283-2289. 146. Murthy, N.S., Stamm, M., Sibilia, J.P., and Krimm, S., Structural changes accompanying hydration in Nylon 6. Macromolecules, 1989. 22(3): 1261-1267. 147. Murthy, N.S., Akkapeddi, M.K., and Orts, W.J., Analysis of lamellar structure in semicrystalline polymers by studying the absorption of water and ethylene glycol in Nylons using Small-Angle Neutron Scattering. Macromolecules, 1998. 31(1): 142-152. 148. Cser, F., Rasoul, F., and Kosior, E., Modulated differential scanning calorimetry. The effect of experimental variables. J. Therm. Anal., 1997. 50(5-6): 727-744. 149. Edwards, S., Discussion on Photoacoustic FTIR experimental conditions,, with J. Pockett, 2001: Adelaide, Aust. 150. McLelland, J.F., Jones, R.W., Luo, S., and Seaverson, L.M., A practical guide to FTIR Photoacoustic Spectroscopy, in Practical Sampling Techniques for Infrared Analysis, D.P.B. Coleman, Editor. 1992, CRC Press. 107-144. 151. Schultz, J.M., Polymer crystallization: The development of crystalline order in thermoplastic polymers. 2001, Washington, DC, USA: Oxford University Press. 152. Ramesh, C., New crystalline transitions in Nylons 4,6, 6,10, and 6,12 using high temperature X-ray diffraction studies. Macromolecules, 1999. 32(11): 3721-3726. 153. Yan, D., Li, Y., and Zhu, X., Brill transition in Nylon 10-12 investigated by variable temperature XRD and real time FTIR. Macromol. Rapid Commun., 2000. 21(15): 1040-1043. 154. Quintanilla, L., Rodriguez-Cabello, J.C., and Pastor, J.M., Structural analysis of injection-molded semicrystalline polymers by Fouriertransform infrared spectroscopy with photoacoustic detection and differential scanning calorimetry: 2. Polyamide-6,6. Polymer, 1994. 35(11): 2321-2328. 349

155. de Vries, K., Linssen, H., and van der Velden, G., End-group and mainchain resonance assignments for the aliphatic polyamide-4,6, -6,6, and 6 via two dimensional NMR methods. Macromolecules, 1989. 22(4): 1607-1610. 156. Howe, C., Sankar, S., and Tonelli, A.E., Carbon-13 NMR observation of poly(L-lactide) in the narrow channels of its Inclusion Compound with urea. Polymer, 1993. 34(12): 3674-3676. 157. Howe, C., Vasanthan, N., MacClamrock, C., Sankar, S., Shin, I.D., Simonsen, I.K., and Tonelli, A.E., Inclusion Compound formed between poly(L-lactic acid) and urea. Macromolecules, 1994. 27(25): 7433-7436. 158. Ratta, V., Crystallization, morphology, thermal stability and adhesive properties of novel high performance semicrystalline polyamides, in Department of Chemical Engineering. 1999, Virginia Tech: Blacksburg, Virginia. 159. Vincent, D.B., Discussion on reasons for weak peak in melting thermogram for polyamide-6,12/phenothiazine blend,, with J. Pockett, 2003: Adelaide, Aust. 160. Chen, H.-L. and Wang, S.-F., Crystallization induced microstructure of polymer blends consisting of two crystalline constituents. Polymer, 2000. 41(14): 5157-5164. 161. Wunderlich, B., Discussion with Mr P. Folland at Thermal Analysis conference Adelaide SA,, with P. Folland, 1995: Adelaide, Aust. 162. Simon, S.L. and McKenna, G.B., The effects of structural recovery and thermal lag in modulated DSC measurements. Thermochim. Acta, 1997. 307(1): 1-10. 163. Reading, M., Modulated differential scanning calorimetry - a new way forward in materials characterization. Trends in Polymer Science, 1993. 1(8): 248-253. 164. Buckman, N., Modulated DSC: the newest innovation in thermal analysis. Chem. Aust., 1993. 60(12): 666-667. 165. Cesaro, A., Navarini, L., and Pepi, R., A modulated DSC analysis of microbial biocompatible polyesters. Thermochim. Acta, 1993. 227(1-2): 157-166. 166. Gill, P.S., Sauerbrunn, S.R., and Reading, M., Modulated differential scanning calorimetry. J. Therm. Anal., 1993. 40(3): 931-939. 167. Wunderlich, B., Jin, Y., and Boller, A., Mathematical description of differential scanning calorimetry based on periodic temperature modulation. Thermochim. Acta, 1994. 238(1-2): 277-293.

350

168. Cao, J., Mathematical studies of modulated differential scanning calorimetry. I. Heat capacity measurements. Thermochim. Acta, 1999. 325(2): 101-109. 169. Ozawa, T. and Kanari, K., Linearity and non-linearity in DSC: A critique on modulated DSC. Thermochim. Acta, 1995. 253: 183-188. 170. Simon, S.L. and McKenna, G.B., Measurement of thermal conductivity using TMDSC: Solution to the heat flow problem. J. Reinf. Plast. Compos., 1999. 18(6): 559-571. 171. Xu, S.X., Li, Y., and Feng, Y.P., Temperature modulated differential scanning calorimetry: on system linearity and the effect of kinetic events on the observed sample specific heat. Thermochim. Acta, 2000. 359(1): 43-54. 172. Scherrenberg, R., Mathot, V., and Steeman, P., The applicability of TMDSC to polymeric systems. General theoretical description based on the full heat capacity formulation. J. Therm. Anal. & Calorim., 1998. 54(2): 477-499. 173. Wunderlich, B., Boller, A., Okazaki, I., and Ishikiriyama, K., Heatcapacity determination by temperature-modulated DSC and its separation from transition effects. Thermochim. Acta, 1997. 304(305): 125-136. 174. Simon, S.L. and McKenna, G.B., Interpretation of the dynamic heat capacity observed in glass-forming liquids. J. Chem. Phys., 1997. 107(20): 8678-8685. 175. Toda, A., Oda, T., Hikosaka, M., and Saruyama, Y., A new analyzing method of temperature modulated DSC of exo- or endo-thermic process:. Application to polyethylene crystallization. Thermochim. Acta, 1997. 293(1-2): 47-63. 176. Schick, C., Merzlyakov, M., and Wunderlich, B., Analysis of the reorganization of poly(ethylene terephthalate) in the melting range by temperature-modulated calorimetry. Polym. Bull. (Berlin), 1998. 40(23): 297-303. 177. Pyda, M. and Wunderlich, B. Analysis of Lissajous figures from TMDSC of melting and crystallization of polymers and small molecules. in Proc. Conf. North Am. Therm. Anal. Soc., 26th. 1998. 178. Scherrenberg, R., Mathot, V., and van Hemelrijck, A., The practical applicability of TMDSC to polymeric systems. Thermochim. Acta, 1999. 330(1-2): 3-19. 179. Wunderlich, B., Boller, A., Okazaki, I., Ishikiriyama, K., Chen, W., Pyda, M., Pak, J., Moon, I., and Androsch, R., Temperature-modulated differential scanning calorimetry of reversible and irreversible first-order transitions. Thermochim. Acta, 1999. 330(1-2): 21-38. 351

180. Simon, S.L. and McKenna, G.B., Quantitative analysis of errors in TMDSC in the glass transition region. Thermochim. Acta, 2000. 348(1-2): 77-89.

352