Synthesis and Characterization of New Copolymers from ... - umexpert

Fibers and Polymers 2012, Vol.13, No.5, 555-563

DOI 10.1007/s12221-012-0555-4

Synthesis and Characterization of New Copolymers from Glycidyl Methacrylate and Tetrahydrofurfuryl Acrylate: Determination of Reactivity Ratios Ahmad Danial Azzahari, Rosiyah Yahya*, Aziz Hassan, and Md. Rezaul Karim Sheikh Department of Chemistry, University of Malaya, 50603, Kuala Lumpur, Malaysia (Received October 3, 2011; Revised December 14, 2011; Accepted December 20, 2011) Abstract: The new copolymers from different feed compositions of glycidyl methacrylate (GMA) and tetrahydrofurfuryl acrylate (THFA) were synthesized using free radical polymerization in toluene at 70±1 oC using benzoyl peroxide (BPO) as initiator. The polymers were characterized by 1H NMR, 13C NMR and FTIR spectroscopic techniques. The polydispersities of the copolymers suggest a strong tendency for chain termination by disproportionation. The glass transition temperature of the copolymers increases with increase in GMA content. The thermal stability of the copolymers increases with increase in THFA content. The copolymer compositions were determined using 1H NMR analysis. Reactivity ratios for GMA and THFA as determined by the Mao-Huglin method were r1=0.379 and r2 =0.266. The results showed that all these copolymerizations are strictly linear systems describable by the Mayo-Lewis equation based on the terminal model and that accurate reactivity ratio data can be obtained. Keywords: Copolymerization, Glycidyl methacrylate, Tetrahydrofurfuryl acrylate, Reactivity ratios, Copolymer composition

with low molecular weight and narrow polydispersity. In this work, a new pair of monomers, GMA (hard segment) and tetrahydrofurfuryl acrylate (THFA) (soft segment) were copolymerised by free radical polymerization with different composition ratios of GMA and THFA. The copolymers’ composition and the reactivity ratios of this new pair of monomers were determined by FTIR, 1H NMR and 13C NMR spectroscopies. The solubility behaviour of the copolymers was observed using different solvents. Molecular weight and polydispersity of the copolymers were measured by GPC while glass transition temperature (Tg) and thermal gravimetric analysis were performed by DSC and TGA analyzers respectively.

Introduction Acrylic copolymers have achieved prime importance in various avenues of application [1-5]. Specific properties of polymers such as hydrophobic/hydrophilic character, and the presence of various functional groups, make these materials very attractive from the biomedical point of view. Crosslinked copolymeric systems based on derivatives of styrenedivinylbenzene, hydroxyalkyl methacrylate, glycidyl methacrylate (GMA), and acryloamide have generated considerable interest [6-8]. In the group of the mentioned monomers, GMA is of special attention because this monomer has not only a double bond but also an epoxy group. In recent years, copolymers based on GMA have attracted increasing interest since the presence of the epoxide groups can allow large number of chemical modifications on the initial polymer [9-13]. The first copolymers reported based on GMA have been used as supports for enzymes [14] and as sorbents for chromatography [15-17]. They were also used in removing metal ions [18,19] or even as biomaterials [20]. Recently Beta Podkoscielna [21] synthesized, characterized and modified the copolymers of GMA and bis[4(2hydroxy-3-methacryloyloxypropoxy)phenyl]sulphide (BES. DM) and found interesting properties of the modified copolymers. Aurel Wolf and co-workers [22] synthesized copolymers of styrene (S) and glycidyl methacrylate (GMA) at high temperature by free-radical copolymerization. They determined the reactivity ratios of the synthesized copolymers and the reactivity ratios in combination with rate coefficients of the homo polymerizations, which permit the development and modelling of a prepolymer for coatings

Experimental Materials Glycidyl methacrylate (GMA) (Merck) and tetrahydrofurfuryl acrylate (THFA) (Aldrich) were purified by distillation under reduced pressure. The purity of the monomers was confirmed by NMR spectra. 1 H NMR for GMA (CDCl3, ppm): 1.94 (3H) (–CH3), 2.66 and 2.84 (2H) (oxirane –OCH2–), 3.24 (1H) (–C*H–), 3.99 and 4.46 (2H) (ester –OCH2–), 5.59 and 6.15 (2H) (CH2=C). 13 C NMR for GMA (CDCl3, ppm): 17.94 (–CH3), 44.27 (oxirane –OCH2–), 49.08 (–C*H–), 64.88 (ester –OCH2–), 125.82 and 135.61 (CH2=C), 166.66 (C=O). 1 H NMR for THFA (CDCl3, ppm): 1.60 and 1.98 (2H) (cyclic –C*CH2–), 1.88 (2H) (cyclic –CH2–), 3.78 and 3.87 (2H) (cyclic –OCH2–), 4.14 (1H) (chiral –C*H–), 4.07 and 4.22 (2H) (ester –OCH2–), 5.81, 6.13 and 6.41 (3H) (CH2=CH). 13C NMR for THFA (CDCl3, ppm): 25.57(cyclic –CH2–), 27.93 (cyclic – C*CH2–), 66.49 (ester –OCH2–), 68.37 (cyclic –OCH2–), 76.39 (chiral –C*H–), 128.16 and

ê

*Corresponding author: [email protected] 555

556

Fibers and Polymers 2012, Vol.13, No.5

130.91 (CH2=CH), 166.05 (C=O). Benzoyl peroxide (BPO) (Merck) was recrystallized from ethanol and dried under vacuum at 40 oC. All the other solvents were purified by distillation prior to their use. Synthesis of Copolymers The copolymerization reactions were carried out with predetermined ratios of GMA and THFA monomers in toluene under N2 atmosphere at 70±1 oC for 20 min to keep the conversion low. The initiator BPO dissolved in a small amount of toluene was immediately charged into the reaction vessel to begin the reaction. The total concentrations of monomers and initiator were kept at 1.39 M and 0.03433 M respectively for each batch. The copolymer solution was precipitated in excess of n-hexane, purified by reprecipitation and dried in vacuum at ambient temperature for 12 h. Conversions were calculated gravimetrically. Measurements 1 H NMR and 13C NMR measurements were performed with a JEOL spectrometer (Lambda 400 MHz). The copolymers were dissolved in CDCl3 and concentrations were maintained at 2.5 % (w/v) and 15 % (w/v) for the 1H NMR and 13 C NMR analyses respectively. The chloroform peaks that were used as internal reference for the 1H NMR and 13C NMR were set at 7.26 ppm and 77.0 ppm respectively. The nature of each carbon atom was determined using the DEPT spectral editing technique, with proton pulses at θ =45 o, θ = 90 o and θ =135 o. The FTIR spectra of the copolymers were recorded with a PerkinElmer 400 spectrometer. Thermal data were obtained using a PerkinElmer HyperDSC instrument and PerkinElmer Pyris-Diamond TG/DTA thermobalance. Molecular weight determinations were performed using a gel permeation chromatography (GPC) instrument (Waters 2414 refractive index detector coupled with a Waters 717 plus Autosampler and Waters 600 Controller) with polystyrene standards as reference and tetrahydrofuran (THF) as the eluent. Determination of Solubility of Copolymers Solubility of the copolymers was tested in various polar and non-polar solvents. About 0.1~0.5 g of the copolymer was added to about 10 ml of different solvents in a test tube and kept overnight with the test tube tightly closed. The solubility of the copolymers was noted after 48 h.

Results and Discussion Synthesis of Copolymers The copolymers of GMA with THFA in toluene were synthesized with composition of mole fractions ranging from 0.10 to 0.85 as shown in Table 1. The copolymers obtained were colourless solids with intermediate properties of polyGMA (hard solid) and polyTHFA (soft adhesive).

Ahmad Danial Azzahari et al. Table 1. Composition data for free radical copolymerization of GMA (1) with THFA (2) in toluene solution at 70±1 oC f1 Conversion (wt%) 0.102 6.4 0.151 6.9 0.201 15.7 0.249 6.4 0.397 7.9 0.553 6.3 0.600 5.6 0.751 4.0 0.793 3.7 0.849 3.8

IGT 20.370 14.550 12.21 8.660 7.450 6.500 6.030 4.560 4.210 3.440

F1 0.214 0.285 0.329 0.429 0.478 0.526 0.554 0.661 0.693 0.776

µ1 1.043 1.068 1.095 1.125 1.249 1.468 1.568 2.139 2.449 3.124

µ2 3.354 2.491 2.057 1.803 1.405 1.215 1.177 1.088 1.069 1.047

Scheme 1. Synthesis of copolymer of GMA and THFA.

The synthesis of the copolymer is outlined in Scheme 1. Characterization of Copolymer Solubility Behaviour of Copolymer The copolymers are soluble in chloroform, acetone, dimethyl formamide, dimethyl sulfoxide, tetrahydrofuran, toluene and insoluble in n-hexane, petroleum ether and hydroxyl-group containing solvents such as methanol and ethanol. Structural Characterization and Composition of Copolymer 1 H NMR Spectra The 1H NMR spectra with increasing THFA compositions (from top to bottom) of poly(GMA-co-THFA) are presented in Figure 1. The proton assignments are based on their corresponding monomers as well as comparison with spectra of analogous chemical groups taken from the literature [2326]. Due to the tacticity of the α-CH3 from the GMA unit, a series of the corresponding resonance signals appear at 0.93, 1.02, and 1.09 ppm which have been assigned to syndiotactic (rr), heterotactic (mr+rm) and isotactic (mm), respectively. Three very well-defined peaks, belonging to the epoxy groups, appear between 2.5 and 3.4 ppm. The proton of the chiral carbon on the oxirane ring resonates at 3.22 ppm whereas protons on the –OCH2– of the oxirane ring resonates at 2.63 ppm and 2.83 ppm. These peaks were used as references to follow the GMA modification. The GMA unit’s protons on the –OCH2– of the ester group resonates at

New Copolymers and Reactivity Ratios


557

Figure 2. The 13C-NMR (top) and DEPT 135 spectra (bottom) of poly(GMA-co-THFA). Figure 1. From top to bottom, the 1H-NMR spectra of typical GMA-co-THFA polymers.

about 3.8 ppm and 4.3 ppm which overlaps with five other protons of the THFA unit in the range of 3.68 to 4.44 ppm. The peaks in the range of about 1.4 to 2.5 ppm consists of the methylene protons of the polymer chain backbone of both the GMA and THFA units, as well as the methine proton of the polymer chain backbone of the THFA unit along with four of the –CH2– protons on the THF ring. The intense peak at 2.16 ppm in most of the copolymer samples is due to protons of the trace acetone that could not be removed via the drying methods. 13 C NMR Spectra The proton decoupled 13C NMR and the DEPT 135 o spectrum of poly(GMA-co-THFA) are presented in Figure 2. The DEPT experiment differentiates between primary, secondary and tertiary carbon groups by variation of the selection angle parameter (the tip angle of the final 1H pulse): θ =45 o give all carbons with attached protons in phase; θ =90 o gives only methine groups, the others being suppressed; θ =135 o give all CH and CH3 in a phase opposite to CH2. The carbons assignments are based on their corresponding monomers as well as comparison with spectra of analogous chemical groups taken from literature [24,26]. The α-CH3 from the GMA unit resonates at 16.65, 16.80 and

17.83 ppm due to tacticity (C1). Carbons on the THF ring show resonance signals at 25.52, 28.16, 68.08 and 76.06 ppm (C9, C2, C3 and C10 respectively). The peak at 76.06 ppm which is almost hidden by the nearby CDCl3 signals in the 13C NMR spectra becomes much more noticeable in the DEPT spectra since signals from quaternary carbons and other carbons with no attached protons are always absent in the DEPT spectra. The methine carbon on the polymer backbone of the THFA unit shows a signal at 41.14 ppm (C5). Other methylene carbons belonging to the polymer backbone of both the GMA and THFA units show less pronounced peaks between 35 to 54 ppm (C4 and C4’). On the epoxy group of the GMA unit, the methylene carbon shows a signal at 44.56 and 44.68 ppm (C6) whereas the methine carbon shows a signal at 48.69 and 48.88 ppm (C7). A series of resonance signals between 65.54 to 66.63 ppm corresponds to the methyleneoxy group on both the GMA and THFA units (C8 and C8’). The carbonyl carbon of the THFA unit resonates at 174.14 and 174.23 ppm and (C11’) whereas the GMA unit resonates between 175.04 to 177.20 ppm (C11). Carbons of the trace acetone resonate at 30.81 and 206.84 ppm for the methyl and carbonyl carbons respectively. Infrared Spectra The IR spectrum of poly(GMA-co-THFA) (0.554:0.446) is shown in Figure 3. All other copolymers exhibit the same

558


Ahmad Danial Azzahari et al. Table 3. TGA and DSC data for poly(GMA-co-THFA) system (oC)

Figure 3. FTIR spectra of poly(GMA-co-THFA).

Table 2. Molecular weight data for copolymers of GMA and THFA Mw × 104 Mn × 104 F1a 0.214 7.12 3.54 0.285 7.23 3.56 0.329 8.31 4.12 0.429 7.31 3.66 0.478 9.34 4.59 0.526 9.23 4.52 0.554 9.97 4.80 0.661 9.96 3.56 0.693 7.78 3.72 0.776 7.41 3.14 a F1 is the mole fraction of GMA in the copolymer.

PDI 2.01 2.03 2.02 2.00 2.03 2.04 2.08 2.80 2.09 2.36

characteristic IR peaks since the functional groups are similar. The peaks observed at 2991 and 2874 cm-1 are attributed to the asymmetrical and symmetrical C–H stretching of methylene and methyl groups. The carbonyl stretching of the two ester groups are observed at 1728 cm-1. The asymmetrical and symmetrical bending vibrations of methyl group are seen at 1448 and 1391 cm-1, respectively. The peaks at 1250 and 1160 cm are due to C–O stretching. The C–O–C stretching of the epoxide occurs at 848 cm-1. Molecular Weights The number and weight-average molecular weights for copolymers of GMA and THFA determined by GPC as presented in Table 2 shows that the polydispersity indices generally have a value of 2.0 irrespective of the copolymer composition. The theoretical value of Mw/Mn for polymers produced via disproportionation is 2.0 [27]. Thus, the polydispersity indices of poly(GMA-co-THFA) obtained suggest that the chain termination takes place mainly by disproportionation. Glass Transition Temperature The DSC analysis procedure used consisted of first heating the polymer from –50 to 150 oC, then cooling it to –50 oC at a rate of 100 oC/min, and holding at –50 oC for 5 minutes before the next heating cycle. Polymer glass transition

Temperature at weight lossa Tg at heating cycleb F1 10 % 30 % 50 % 70 % 90 % first second third 0.214 338 374 398 420 448 3.63 –3.08 –3.08 0.285 326 364 389 415 444 –1.27 –3.20 –3.11 0.329 315 359 387 415 446 0.21 –3.29 –2.71 0.429 311 358 381 411 441 0.18 –2.10 –1.23 0.478 311 351 374 408 440 15.23 15.70 15.98 0.526 307 345 370 408 440 42.74 41.72 42.01 0.554 307 342 368 409 446 46.81 38.25 38.55 0.661 306 340 363 407 441 66.56 58.79 59.15 0.693 297 338 361 400 441 71.44 63.30 63.84 0.776 294 333 356 401 441 70.61 65.71 63.40 a As determined from TGA and bas determined from DSC.

Figure 4. Variation of Tg with composition of poly(GMA-coTHFA).

temperature, which represents the molecular mobility of polymer chains, is an important phenomenon that influences the material properties and potential applications of a given polymer. Table 3 gives the glass transitions as determined by DSC. Figure 4 portrays the Tg values determined on all heating cycles. In general, Tg values measured on the second and third heating cycles were slightly lower than that of the first heating cycle and has an S-shaped Tg versus composition relation. The THFA unit does not significantly influence the Tg of the copolymer until it reaches about 43 % mol GMA. From that point onwards, Tg are seen to increase in a relatively linear manner up to about 66 % mol GMA, where a plateau is achieved. This behavior can be described



559

Figure 5. TGA thermogram of poly(GMA-co-THFA).

by a recent equation proposed by Liu et al. [28] which takes into consideration the contribution of triad sequences to the Tg of the copolymer: Tg = n111Tg1 + n222Tg2 + 2(n121−n112)Tg[12] + 3n112Tg[112] + 3n221Tg[221] (1) where the triad fractions of n111, n222, n121, n112, and n221 are given as: n111=F1P11P11, n222=F2P22P22, n121=F1P12P21, n112= F1P11P12, and n221=F2P22P21. The diad probabilities P11, P12, P22, and P21 are calculated from the reactivity ratios (described in the later section using values obtained from the MH method) and are given as [29]: P11=r1/(r1+f2/f1), P12=1−P11, P22=r2/(r2+f1/f2), and P21=1−P22. Tg1 and Tg2 corresponds to the homopolymers of PGMA and PTHFA with values of −8.07 and 74 oC [30], respectively. The Tg versus composition relation equation above has three intractable parameters: Tg[12], Tg[112] and Tg[221] which were determined by fitting Tg experimental data at the third heating cycle with the equation above by regression methods and the fitted values obtained were −21.85, 101.89, and −3.99 oC, respectively. High Tg of the copolymers indicates substantial decrease of chain mobility of the copolymer due to α-CH3 from the GMA unit. Thermal Stability Thermogravimetric analysis was performed in N2 at a heating rate of 20 oC/min. The results presented in Table 3 shows the thermal decomposition data of the copolymer series. Figure 6 shows the first derivative plots corresponding to the specified samples, i.e., the bottom most plot corresponds to the lowest GMA content and the incremental plots in between lead up to the highest GMA content. It can be deduced from both Table 3 and Figure 6 that a general

Figure 6. The first derivative plots of the thermograms of GMAco-THFA polymers.

trend does exist when one sample is compared to the next increment, i.e., when the GMA content is higher, the bulk of the decomposition occurs at the lower temperature range. The following mechanism is assumed to explain the decomposition behavior: The first step of the decomposition is due the loss of CO2 and the rupture of weak linkages in the pendant group and thereby volatilization of low molecular weight species for the first 10 % weight loss. This process is

560


Ahmad Danial Azzahari et al.

followed by the breakage of main chains and volatilization of the cleaved products for the remaining 90 % weight loss. The similarities of both the pendant groups (epoxy and furfuryl) of the copolymers in their thermal response are reflected in Figure 5 by a single major stage of decomposition, and this observation is typical for all copolymers in the series. However, the first derivative plots of the various compositions of poly(GMA-co-THFA), as shown in Figure 6, clearly indicates that the single major stage of decomposition in Figure 5, involves at least two major steps. The first derivative peak that appears at the lower temperature range corresponds to GMA moiety whereas the second derivative peak at the higher temperature range corresponds to THFA moiety. This is most likely due to the THFA moiety being more stable and having a slightly larger molecular weight than GMA. Reactivity Ratios of Copolymers The precise determination of monomer reactivity ratios (MRR) would serve as a useful tool towards the accurate estimation of copolymer composition, understanding their properties and utility for tailoring copolymers with desired physicomechanical properties. However, due to the similarity in the constituent units of copolymers containing acrylate and methacrylate monomers, it is difficult to determine their compositions by normal analytical techniques [23,31]. Although it is possible to determine the varying compositions of the copolymers via FTIR spectra by monitoring the C–O–C stretching of the epoxide at 847 cm-1, the NMR offers a less ambiguous, yet more simpler and rapid evaluation technique based on the distinctive absorption of the lone proton of the chiral carbon on the oxirane ring of the GMA unit which is then compared against the overlapping protons of the GMA unit (2 protons) and THFA unit (5 protons) located downfield from the lone proton of the chiral carbon signal on the GMA unit. The integrated area of the lone proton signal is set to a normalized value of 1, hence the simplified equation to determine the copolymer composition becomes: F1=5/(IGT +3) and F2=1−F1, where IGT is the integrated area of the ester protons in the GMA unit plus the ester, methine and methylene protons in the THFA unit, F1 and F2 are the mole fractions of GMA and THFA in the copolymer, respectively. The results of the copolymer compositions from initial GMA feed f1 (and THFA feed f2= 1–f1) are given in Table 1. The behaviour of the GMA/THFA system was evaluated through a plot of copolymer composition versus feed composition (Figure 7). Both reactivity ratios are less than unity as shown by an azeotropic composition at approximately 55% GMA feed. As such, a tendency towards alternation and no long homopolymeric blocks is expected [29]. The experimental results presented in Table 3 have been treated with several methods to calculate the monomer

Figure 7. Copolymer composition of GMA vs feed composition of GMA.

reactivity ratios (MRR) for the GMA/THFA system. All these methods are based on the terminal model, viz. the Mayo-Lewis equation [32]. d [ M1 ] [ M1 ] ⎛ r1 [ M1 ] + [ M2 ]⎞ -------------- = ----------- -------------------------------d [ M1 ] [ M2 ] ⎝ r2 [ M2 ] + [ M1 ]⎠

(2)

Where [M1]/[M2] is the ratio of the molar concentrations of monomers M1 and M2 in the feed or, equivalently, the ratio of their mole fractions fl and f2. This ratio is usually denoted by f, where f = f1/f2 = [M1]/[M2] and d[M1]/d[M2] is the ratio of the instantaneous rates of consumption of the monomers and also expresses the instantaneous composition of the copolymer produced. If the conversion is sufficiently low, the average copolymer composition (usually expressed by F) produced from tinitial =0 to tfinal =t is approximated by F=F1 /F2=∆[M1]/∆[M2]≈d[M1]0/d[M2]0 where F1 and F2 are the average values of the mole fractions of the monomer units in the copolymer. One of the well-known linearization methods to determine the MRR is given by the Kelen-Tudos method [33,34]: r2 η = r1 ξ – --α- ( 1 – ξ )

(3)

Where η=(G/α+H); ξ=(H/α+H). The G and H are parameters from the Fineman-Ross method [35], defined as: G=f(F−1)/F and H=f 2/F. The Fineman-Ross method however, has the unfortunate consequence of having certain experimental points with abnormal and inappropriate weights in the plot. The Kelen-Tudos method overcomes this by introducing a symmetry parameter defined as α=(Hmin × Hmax)0.5.



561

Figure 9. Joint confidence region (95 %) for the evaluated values of r1 and r2 by KT, EKT and MH method.

Figure 8. Kelen-Tudos plot, Extended Kelen-Tudos plot and MaoHuglin plot at the last iteration.

A plot of η vs ξ gives a straight line, which on extrapolation gives the intercept equalling to –r2 /α and the slope as r1+r2/α. Kelen and Tudos modified this equation for higher conversions [36] by redefining G and H into G=(F−1)/z and H=F/z2 where z=log10(1−τ1)/log10(1−τ2); τ2=(w/100)(µ+f)/ (µ+F); τ1 =τ2(F/f ); µ=(mol. wt monomer2)/(mol. wt monomer1) and w is the weight percent conversion of total monomers. This procedure does not suffer from reindexing errors and it gives data symmetrically located along the interval of the independent variable. The Mao-Huglin method [37,38] is a more recent technique for determination of MRR using the integrated form of equation (1) [39] and the Kelen-Tudos method in each iteration. The η vs ξ plots of all the three methods used are shown in Figure 8. According to the confidence interval estimation technique of the linear-least squares method, [40] the 95 % individual confidence intervals of the r1 and r2 for all three methods can be calculated by the following unified equations: 2

SR Σ( 1 – ξ i) - ---------------------∆r1 = ±t0.025, (n – 2) --------n–2 D

(4)

2

SR Σξ i - -------∆r2 = ±α × t0.025, (n – 2) --------n–2 D

(5)

Where t0.025, (n−2) is the Student’s t distribution with (n–2) degrees of freedom and with each tail area probability equaling 0.025. n is the number of experimental points. All summations are carried out from i=1 to i=n. The quantities SR and D are as follows: r2 SR = ∑ ηi – r1 ξi + --α- ( 1 – ξ i)

2

(6)

2

2

D = Σξi Σ ( 1 – ξi ) – [ Σξi ( 1 – ξ i) ]

2

(7)

Since r1 and r2 are not independent of one another, it would be more appropriate to use the 95 % joint confidence region, which is enclosed by the ellipse, according to the following equation: 2 2 2 ( r1 – R1 ) Σξi + --- ( r1 – R1 ) ( r2 – R2 )Σξi ( ξi – 1 ) α

r2 – R2⎞ 2 SR 2 -F + ⎛ -------------Σ ( ξ i – 1 ) = 2 --------⎝ α ⎠ n – 2 0.05;2, (n – 2)

(8)

Where F0.05; 2, (n−2) is the F distribution having 2 and n–2 degrees of freedom and R1, R2 are the least-squares estimation of r1, r2 respectively. The joint confidence region of all the three methods used, as shown in Figure 9, reveals that the calculated results from the extended Kelen-Tudos and Mao-Huglin methods are in close agreement since these two methods take into account the extent of conversion of the copolymers produced from the reactions, which causes compositional drift of F. Kelen and Tudos introduced a quantitative measure [41,42] of the confidence δ , which is defined by the relative values of confidence intervals for systems consistent with the simple binary copolymerization model of equation (1). δ =|(∆r1 ∆r2)/(r1r2)|. A class 1 system is defined by Kelen and Tudos as one that is strictly linear and describable by the two-parameter binary system and has a δ value less than 0.1. Class 1(!) systems are linear and still describable by the two-parameter model but giving poorer parameter estimates because of a greater dispersion of experimental points. The borderline between class 1 and class 1(!) systems is not very sharp, and the η vs ξ plot must be considered to judge the classification of intermediate systems. Kelen and Tudos proposed that a well-planned experiment has ∆r1=∆r2/α, α exptl =α theor, and ∆r1/r1=∆r2/r2 where α theor =r2 /r1. With these

562


Ahmad Danial Azzahari et al.

Table 4. Summary of monomer reactivity ratios Method KT EKT MH

r1 ± ∆r1 r2 ± ∆r2 0.388 ± 0.092 0.282 ± 0.061 0.378 ± 0.090 0.266 ± 0.058 0.379 ± 0.088 0.266 ± 0.057

αexptl 0.653 0.632 0.630

αtheor 0.727 0.704 0.703

∆r1/r1 0.236 0.237 0.232

∆r2/r2 0.216 0.218 0.213

Q 0.898 0.898 0.897

δ 0.051 0.052 0.049

Copolymer Microstructure The probabilities of finding the sequences of x GMA and x THFA units, respectively, are calculated as follows [29]: x–1 x–1 NGMA(x) = P11 ·(1−P11) and NTHFA(x) = P22 ·(1−P22). The values of NGMA(x) and NTHFA(x), shown in Figure 10, were calculated for the copolymers obtained at an equimolar feed. It is observed that poly(GMA-co-THFA) contains moderately alternating sequences of both monomers when at equimolar feeds. The mean sequence length µ1 and µ2 corresponding to GMA and THFA respectively, was also calculated using the following relations [29]: µ1 = 1 + r1(f1/f2) and µ2 = 1 + r2(f2/f1). The values of µ1 and µ2 are presented in Table 1. The µ1 values increase from 1.043 to 3.124 in the GMA/THFA system with increasing concentration of GMA in feed. The almost similar µ values for the GMA/THFA system at near equimolar feeds indicate a moderately alternating distribution of monomeric units in that particular range.

Conclusion

Figure 10. The probabilities of finding the sequences of x GMA units (white columns) and x THFA units (black columns) in poly(GMA-co-THFA) at an equimolar feed of f1=f2=0.5.

considerations in mind, Kelen and Tudos introduced a parameter Q that allows a quantitative evaluation of the design of an experiment. r Q = exp – ln ⎛ α ----1⎞ ⎝ r2⎠

(9)

Q has a maximum value of 1 since it was shown that in ideally planned experiments the auxiliary parameter α is equal to the ratio of the copolymer parameters r2/r1. A low Q value, obtained when α exptl is significantly different from the ideal, indicates an improperly designed experiment. Although general conclusions could not be reached by Kelen and Tudos, they reported that class 1 determinations are associated with high values of Q. The following parameters were calculated by using reactivity ratios and confidence intervals for the GMA/THFA system that were determined by the extended Kelen-Tudos procedure as shown in Table 4. Considering all of the parameter data in Table 4, it can be concluded that the experimental design of the GMA/THFA system is qualitatively a good one.

A series of new copolymers of poly(GMA-co-THFA) have successfully been synthesized and characterized by FTIR, 1H and 13C-NMR spectroscopies. Values of the polydispersity index obtained for poly(GMA-co-THFA) suggest a strong tendency for chain termination by disproportionation in all cases. Glass transition temperatures of the copolymers were found to increase with increase in the mol fraction of GMA. Thermal behaviour of these copolymers showed at least two major stages of decomposition. The reactivity ratios were obtained by the KT, EKT and MH methods. The values of r1 and r2 are less than unity indicating that the system gives rise to an azeotropic polymerization and a strong tendency to alternation.

Acknowledgements The authors gratefully acknowledge the financial support for this project from the University of Malaya under the University Research Grant number PS371-2010B.

References 1. B. Adhikari and S. Majumdar, Prog. Polym. Sci., 29, 699 (2004). 2. M. Y. ArIca, G. Bayramoglu, and N. Biçak, Proc. Biochem., 39, 2007 (2004). 3. J. P. H. Pérez, E. López-Cabarcos, and B. López-Ruiz,



Biomol. Eng., 23, 233 (2006). 4. G. Bayramoglu, S. Akgöl, A. Bulut, A. Denizli, and M. Y. ArIca, Biochem. Eng. J., 14, 117 (2003). 5. G. D. Nino, M. Turacchio, A. A. D’Archivio, S. Lora, B. Corain, and G. Antonini, React. Funct. Polym., 61, 411 (2004). 6. M. Malmsten and A. Larsson A, Colloid. Surf. B, 18, 277 (2000). 7. S. Slomkowski, Prog. Polym. Sci., 23, 815 (1998). 8. X.-Z. Zhang, P. J. Lewis, and C.-C. Chu, Biomater., 26, 3299 (2005). 9. J. G. Park, J. W. Kim, and K. D. Suh, Colloid. Polym. Sci., 279, 638 (2001). 10. K. D. Safa and M. H. NNasirtabrizi, Polym. Bull., 57, 293 (2006). 11. A. Trzeciak, E. Mieczyñska, J. Ziólkowski, W. Bukowski, A Bukowsa, J. Noworól, and J. Okal, New J. Chem., 32, 1124 (2008). 12. D. Horák, E. Pollert, and H. J. Macková, Mater. Sci., 43, 5845 (2008). 13. T. Caykara, S. S. Alaskan, M. Gürü, H. Bodugöz, and O, Güven, Radit. Phys. Chem., 76, 1569 (2007). 14. R. Prodanoviéc, S. Jovanoviéc, and Z. Vuzc iéc, Biotechnol. Lett., 23, 1171 (2001). 15. A. Nastasoviéc, S. Jovanoviéc, D. D ord eviéc, A. Onjia, D. Jakovljeviéc, and T. Novakoviéc, React. Funct. Polym., 58, 139 (2004). 16. D. Blondieau, M. Bigan, and P. Desres, React. Funct. Polym., 27, 175 (1995). 17. S. Zhang, X. huang, N. Yao, and C. Horváth, J. Chromatogr. A, 948, 193 (2002). 18. N. Bicak, D. C. Sherrington, S. Sungur, and N. Tan, React. Funct. Polym., 58, 43 (2004). 19. Y. Bondar, H. J. Kim, S. H. Yoon, and Y. J. Lim, J. React. Funct. Polym., 54, 141 (2003). 20. C. Soykan, M. Ahmedzade, and M. Coskum, Eur. Polym. J., 36, 1667 (2000). 21. B. Podkoéscielna, J. Appl. Polym. Sci., 120, 3920 (2011). 22. A. Wolf, F. Bandermann, and C. Schwede, Macromol.

563

Chem. Phys., 203, 393 (2002). 23. H. Bakhshi, M. J. Zohuriaan-Mehr, H. Bouhendi, and K. Kabiri, Polym. Test., 28, 730 (2009). 24. M. H. Espinosa, P. J. O. del Toro, and D. Z. Silva, Polymer, 42, 3393 (2001). 25. P. Y. Ghi, D. J. T. Hill, and A. K. Whittaker, J. Polym. Sci. Pol. Chem., 37, 3730 (1999). 26. K. Rajendrakumar and R. Dhamodharan, Eur. Polym. J., 45, 2685 (2009). 27. S. Teramachi, A. Hasegawa, M. Akatsuka, A. Yamashita, and N. Takemoto, Macromolecules, 11, 1206 (1978). 28. G. Liu, Z. Meng, W. Wang, Y. Zhou, and L. Zhang, J. Phys. Chem. B, 112, 93 (2008). 29. G. G. Odian, “Principles of Polymerization”, pp.1-55, 4th ed., Wiley-Interscience, New Jersey, 2004. 30. J. Brandrup, E. H. Immergut, and W. McDowell, Polymer Handbook, 2nd ed., Wiley Interscience, New York, 1975. 31. N. Grassie, B. J. D. Torrance, J. D. Fortune, and J. D. Gemmell, Polymer, 6, 653 (1965). 32. F. R. Mayo and F. M. Lewis, J. Am. Chem. Soc., 66, 1594 (1944). 33. T. Kelen and F. Tüdós, J. Macromol. Sci. A, 9, 1 (1975). 34. T. Kelen and F. Tüdós, React. Kinet. Catal. L., 1, 487 (1974). 35. M. Fineman and S. D. Ross, J. Polym. Sci., 5, 259 (1950). 36. F. Tüdos, T. Kelen, T. Földes-berezsnich, and B. Turcsányi, J. Macromol. Sci. A., 10, 1513 (1976). 37. R. Mao and M. B. Huglin, Polymer, 34, 1709 (1993). 38. R. Mao, M. B. Huglin, T. P. Davis, and A. S. Overend, Polym. Int., 31, 375 (1993). 39. I. Skeist, J. Am. Chem. Soc., 68, 1781 (1946). 40. G. E. P. Box, W. G. Hunter, and J. S. Hunter, Statistics for Experiments, John Wiley & Sons, New York, 1978. 41. F. Tüdos, T. Kelen, B. Turcsányi, and J. P. Kennedy, J. Polym. Sci. Pol. Chem., 19, 1119 (1981). 42. D. Braun, W. Czerwinski, G. Disselhoff, F. Tüdós, T. Kelen, B. Turcsányi, and D. Angew, Makromol. Chem., 125, 161 (1984).

ê