Composites: Part B 57 (2014) 126–135
Contents lists available at ScienceDirect
Composites: Part B journal homepage: www.elsevier.com/locate/compositesb
Effect of coupling agent on natural fibre in natural fibre/polypropylene composites on mechanical and thermal behaviour A. El-Sabbagh ⇑ Institute of Polymer Materials and Plastics Engineering, Clausthal University of Technology, Agricola Str. 6, Clausthal-Zellerfeld D-38678, Germany Design and Production Engineering Department, Ain Shams University, Cairo, Egypt
a r t i c l e
i n f o
Article history: Received 29 January 2013 Received in revised form 19 September 2013 Accepted 23 September 2013 Available online 1 October 2013 Keywords: A. Fibres A. Polymer-matrix composites (PMCs) B. Adhesion B. Mechanical properties E. Thermoplastic resin
a b s t r a c t To enhance the adhesion between the natural fibre and the thermoplastic matrix, a coupling agent of maleic anhydride grafted polypropylene MAPP is applied. In literature, there are different guidelines of the optimum percentage required of MAPP. Therefore, a systematic work is carried out to optimise the MAPP percent with respect to the type of the natural fibre. Different parameters are investigated namely; Coupling agent ratio to the fibre (0%, 6.67%, 10%, 13.3%, 16.67%), coupling agent source, fibre type (flax, hemp, sisal), and fibre content (30%, 50%). Composite is produced using a kneader and the resulting material is assessed mechanically, thermally, microscopically and for water absorption. For different MAPP source and the natural fibre type, optimum MAPP to fibre ratio is found in average to range between 10% and 13.3% according to the investigated property (stiffness, strength and impact). Increase of MAPP is found to decrease the melting temperature. The thermal behaviour is also linked to the copolymer molecular weight. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Natural fibres polymer composites occupy a significant position in the engineering market. A lot of automobile interior parts are already manufactured using these natural fibre composites in order to replace as much as possible the synthetic fibre products. It is already known that natural fibre benefits are attractive [1–7]: their light weight, high mechanical properties, drop of environmental problems concerning manufacturing, and recycling and longer life for tooling specially with high shear rate processes. Thermoplastic composites are earning a lot of attention to have both the functions of natural fibres and the production easiness of thermoplastics. However, NFTC still suffer from many problems that hinder the growth of their markets. Namely, processing problems like the fibres’ limited processing temperature range [3,4], fibre hydrophilic nature [4–7], bacterial degradation [8], dimensional instability because of the water absorption [8–10], inconsistent mechanical properties [3,8], and processing difficulties [11]. The main problem which arises during the manufacturing phase is the hydrophobic nature of the polymer and the hydrophilic nature of the thermoplastic polymer [12–14]. Lots of methods are adopted to treat either the surface of the fibre or the polymeric matrix in order to enhance the adhesion and ensure an acceptable load transfer [12,15,16]. Physical treatment ⇑ Tel.: +49 5323722487; fax: +49 5323722324. E-mail address:
[email protected] 1359-8368/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesb.2013.09.047
includes treatments by heat [6], plasma [17], corona [18], and surface fibrillation [6]. Chemical treatment includes different methods. Alkalinisation for example depends on the breaking of the hydrogen bond in cellulose and removes some of the lignin and wax covering the natural fibres [19]. Free radical formation in cellulose accompanied by graft copolymerisation is another route as in acrylation, acetylation [20], silanisation [21] and maleic anhydride MA treatment [13]. Of the previous methods, Alkalisation and chemical grafting using MA treatment are most popular. Alkalisation is even considered a standard step to increase the efficiency of any further treatment. MA grafting is not only used to modify fibre surface but also the thermoplastic matrix to achieve better interfacial bonding and mechanical properties in composites. It exploits the hydroxyl groups which are abundantly available in the cellulosic natural fibre from one side and grafts to the matrix polymer from the second side. Additionally, unlike acrylic acid, MA does not react with itself under typical industrial grafting conditions. MA grafting improves the mechanical properties, water absorption resistance and dimensional stability of the composites. Besides, it is suitable to be used cost-wise. Also, grafted MA, like maleic anhydride grafted polypropylene MAPP, can be used directly with the available compounding machine without inserting a new processing step to the production cycle. An issue, however, arises during the use of the MA copolymers like MAPP. It is the presence of many rules to follow to get best results of grafting. Table 1 illustrates some of the common practices
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135 Table 1 Plan of investigated parameters. #
Parameter
Variation
1 2 3
Fibre type MAPP: NF NaOh conc. pretreatment Fibre content % MAPP source
Flax, Hemp, Sisal 0%, 6.7%, 10.0%, 13.3%, 16.7%* 0%, 5%, 10%
4 5
30%, 50% Type A from Sigma Aldrich Type B from Kometra
*
MAPP% in 30% fibre loading is [0%, 6.7%, 10%, 13.3%, 16.7%]⁄30% = 0%, 2%, 3%, 4%, 5%, while in 50% fibre loading the MAPP% is [0%, 6.7%, 10%, 13.3%, 16.7%]⁄50% = 0%, 3.3%, 5.0%, 6.7%, 8.3%.
followed. Bos [8] reported that 3.5% of MAPP powder (Hostaprime HC5) is optimum in kneaded PP/Flax. Keener et al. [13] similarly said that 3% is optimum with MAPP Epolene for extruded PP/Agrofibre. Sain et al. [22] found that 2.5% of E43 MAPP is the optimum percent in kneading PP/Wood fibre. Arbelaiz [14] stated that E43 MAPP should be dependent on the retted flax fibre amount in 1:20 ratio. Zampaloni et al. [23] used compression moulding for PP/Kenaf fibres and reported that 3% MAPP Epolene wax G-3015 is optimum. Bos [24] again in 2006 reported that 1:10 of MAPP (Epolene TM G-3015) to flax is optimum by kneading with PP. Pimenta et al. [25] found that 6% of MAPP G-3015 is needed in the extrusion of PP/Sisal. As seen, the amount of MAPP changes from 2.5% to 6.0% regardless the natural fibre content. The fibre type may have an effect on the optimum MAPP% due to the cellulose content. The ratio of the copolymer to natural fibre is another suggested method to define the amount of the required copolymer [14,24] regardless of the fibre type or the MAPP properties. A study is required to see the effect of different parameters and find out the range of optimality for using the coupling agent. MAPP is taken as the studied copolymer for different natural fibres with polypropylene PP matrix. To compare with the literature results regarding the effect of MAPP effect, weight ratios are used because MAPP is not evaluated in volume fraction. 2. Experimental work The objective of this work is to investigate the optimum MAPP content at different parameters. Table 1 illustrates the factors investigated in this study. These factors are the copolymer to natural fibre weight ratio, hereafter called MAPP:NF, the copolymer molecular weight or graft%, natural fibre type and content. As previously mentioned, pretreatment with alkalinisation is essential. Therefore a set of preliminary experiments is added to the plan of tests in order to find out the optimum NaOH concentration. Three types of fibres are selected due to their common use. Bast fibres represented in flax and hemp. Leaf fibres which are characterised with its coarseness and relatively high stiffness are also represented in sisal. Fibres were supplied by Sachsenleinen-Germany. Flax, hemp and sisal constitute respectively of 63%, 65%, 63% cellulose, 16%, 16%, 12% hemicellulose, 4%, 4%, 11% pectin/lignin, 1%, 1%, 1% fats/waxes, 4%, 2%, 1% proteins/ashs and 12%, 12%, 12% water. Cellulose and moisture content of the fibres’ three types are almost the same. This means that the numbers of OH groups able to react are almost equal. Hemicellulose of the leaf sisal fibres is 4% less than flax and hemp whereas the pectin/lignin is 10% more. Lignin has a positive effect on the adhesion behaviour between the natural fibres and the thermoplastic polymer as reported [26]. Percentages of proteins and ashes are insignificant in sisal compared to the flax and hemp bast fibres. Homopolymer PP developed for thin wall high speed injection moulding is used
127
as the host matrix, and is supplied by DOW GmbH (MFR = 52 at 230 °C/2.16 kg and specific density of 0.9). Fibres are washed with different concentrations of sodium hydroxide solution for different times at room temperature to remove possible surface impurities. The fibres are washed with water then with acetone to reach neutralisation. Fibres are then left to dry at 90 °C for 24 h then left in open air for 3–4 days. Before compounding these alkalinised fibres with PP, they are re-dried again at 90 °C. Two levels only of fibre contents are selected for this plan, due to their wide market application, namely 30% and 50%. The source and the form of the coupling agent are also considered due to the effect of molecular weight, acid number, outer surface area and the probable sites of reaction and grafting [13]. Kim [27] reported that neither low nor high molecular weight copolymers help in the proper entanglement between MAPP and PP matrix. Also neither low nor high MA grafting provide balanced interaction with PP and the fibres. Molecular weight affects the resulting composite’s mechanical properties in a direct proportion way [27]. Two types are studied. Type A is the granulate form MAPP from Sigma Aldrich with molecular weight of 9100 GPC and thermosel viscosity 4000 poise (190 °C). Type B is powder form MAPP supplied by Kometra (SCONA TPPP 8112) with molecular weight of 119850 g/mol. Grafting increases with MA content [28]. From another side; higher molecular weight MAPP has normally lower acid number and hence lower MA content. This results in higher melt flow index MFI, better mixing, and better diffusion [28]. Therefore, both types of MAPP are studied to see the effect of different shapes and molecular weights of the coupling agent. Compounding mechanism is carried out in the batch kneader Haake with roller type. High speed is used firstly when PP and MAPP are added. Torque is observed to ensure total melting of PP and MAPP. Then when the natural fibres are inserted, the speed is lessened to help in smoothing the acquisition of fibres for 5 min. This time has allowed also good impregnation. After the end of fibres insertion, speed is increased again for a while. Speed is then lessened again to 50 rpm and again increased to 100 rpm for 5 min each. This cycle is not the optimum for fibres’ length but it helps to reach homogeneous well impregnated fibres using these rollers from one and relatively short rotation cycle. The kneaded material is then shredded. The granulates are injection-moulded at a temperature pattern of 185–190–195–200 °C using Allrounder 220C 600–250, Arburg, Lossburg, Germany. Samples of mechanical testing are conditioned at 23 °C/50% relative humidity for at least 88 h according to ISO 291 for test room conditions. Tension tests were made using Zwick 0.25 ton tensile machine according to DIN EN ISO 527-1. Test sample is injection-moulded according to ISO 527-2. Test is conducted and evaluated according to ISO 527-1. Water absorption test is conducted also according to DIN 53495-method 3 where the weight of the sample is observed after 30 min at 100 °C distilled water followed by 15 min at 23 °C in distilled water and finally drying of the sample. [1] Thermal gravimetric analysis TGA is also conducted on the 50% flax PP composites to cover the MAPP:NF spectrum for both types of the studied MAPP. TGA is conducted using Q5000 IR from TA instruments in nitrogen atmosphere with 10 k/min rate. [2] Differential scanning calorimetry DSC test is applied using DSC Q2000 by cycle heating from 100 to 250 °C then recooling to 100 °C with 2 k/min to observe the melting temperature, recrystallization temperature, melting enthalpy and recrystallization energy of the composites of different MAPP:NF ratios and MAPP types.
128
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
Effect of time
6
4
2
(a)
0 0
Effect of concentration
8
Loss in weight [%]
Loss in weight [%]
8
6
4
2
(b)
0 5
10
15
20
0
25
2
Time [hr]
4
6
8
10
Concentration [%]
Fig. 1. Effect of (a) Time (b) NaOH Concentration on natural fibre weight loss after alkalinisation.
3. Results 3.1. Effect of pretreatment course Fig. 1 shows the effect of alkalinisation time and concentration of NaOH on the natural fibre weight. Loss in weight is taken as an indication for the removal of impurities, waxes, hemicelluloses and delignification. Logarithmic plateau form can be depicted from both curves. 24 h treatment is taken as an appropriate treatment period. Actually starting from 5 h, the weight loss in Fig. 2a seems to be enough and the increase in loss is not so significant. The weight loss in Fig. 2b points out that 2% is significant in comparison with no treatment. This result matches previous work [29]. However, 5% treatment shows a good improvement even in comparison with 2%. The change from 5% to 10% cannot be considered insignificant. Therefore fibres’ treatments with 0%, 5% and 10% concentration values are carried out and then compounded into composites to investigate the final mechanical properties.
Table 2 presents the effect of fibre pre-treatment (alkalinisation) on the mechanical properties of the manufactured composites. Fibre content is kept at 30 wt%. The ratio of MAPP:NF is kept constant at 1:10 in the test samples of NaOH effect. The used MAPP was that of the granulate form out of Sigma Aldrich. Stiffness improves with 6–9% for the three types of fibres. Only sisal shows a decrease at 10% NaOH treatment. This may be contributed to the excess removal not only of surface impurities but also of lignin [30]. Therefore sisal, which has already higher content of lignin playing a positive coupling rule, was the most influenced one. Excessive NaOH in a way or another decreases the binding between the fibrils and make them move more freely and hence decreases the overall stiffness. Another tentative explanation is due to the excessive loss of weight of fibres after alkalinisation which in turn lessens the load transfer efficiency. This suggests that there are two influencing factors on the mechanical effect. First is the increasing number of sites ready for coupling and adhesion with the natural fibres with the increase of NaOH concentration%.
60
10000 9000
(a)
(b) 50
Strength [MPa]
7000 6000 5000 4000 3000 2000
40 30 20 10
1000 0
0 30
50
30
50
[Wt. %] Flax
Hemp
[Wt. %] Flax
Sisal
6
Hemp
Sisal
9
(c)
8
(d)
5
Impact [kJ/m²]
7
Elongation [%]
Stiffness [MPa]
8000
4 3 2
6 5 4 3 2
1 1 0
0 30
50
30
50
[Wt. %] Flax
Hemp
[Wt. %] Sisal
Flax
Hemp
Sisal
Fig. 2. Effect of fibre type and content at 5% NaOH on (a) Stiffness (b) Strength (c) Elongation % (d) Impact.
129
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135 Table 2 Effect of NaOH concentration on the mechanical properties of PP/30% flax/3% MAPP (Type A). Fibre
Fibre [wt%]
NaOH [%]
Stiffness [MPa]
Dev. [MPa]
UTS [MPa]
Dev. [MPa]
Elongation [%]
Dev. [%]
Impact [kJ/m2]
Dev. [kJ/m2]
Flax Flax Flax Hemp Hemp Hemp Sisal Sisal Sisal
30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0
0.0 5.0 10.0 0.0 5.0 10.0 0.0 5.0 10.0
5178.0 5631.4 5701.4 5231.7 5608.7 5702.7 4928.4 5205.9 4794.9
344.4 502.8 304.1 203.5 180.1 158.8 417.7 247.6 147.3
40.1 43.1 44.1 41.1 42.9 46.0 38.0 38.9 40.8
1.3 0.9 0.4 0.9 1.1 1.0 0.5 1.4 1.0
3.6 4.3 2.8 4.7 3.1 3.2 4.3 3.5 4.5
1.1 0.7 0.5 0.7 0.3 0.2 0.4 0.6 1.0
7.1 7.4 8.1 6.9 7.0 6.9 5.6 6.0 7.3
0.9 0.6 0.9 0.7 0.6 0.8 1.0 0.7 0.4
Natural Fibres 30%
6500 6000
Stiffness [MPa]
Natural Fibres 50%
10000 9000
Stiffness [MPa]
5500 5000 4500 4000 3500
(a)
8000 7000 6000 5000 4000
3000 0
(a)
2
4
6
8
10
12
14
16
3000
18
0
2
4
6
MAPP/Fibre [%] Flax
Hemp
Sisal
Strength [MPa]
Strength [MPa]
14
16
18
40 35 30
(b)
Hemp
Sisal
(b)
55
0
12
Natural Fibres 50%
60
45
25
10
Flax
Natural Fibres 30%
50
8
MAPP/Fibre [%]
50 45 40 35 30
2
4
6
8
10
12
14
16
18
25 0
2
4
6
8
10
12
14
16
18
MAPP/Fibre [%] Flax
Hemp
MAPP/Fibre [%]
Sisal
Flax
Natural Fibres 30%
9
(c)
7
6
impact [kJ/m²]
Impact [kJ/m²]
8
7
Sisal
Natural Fibres 50%
9
8
Hemp
5 4 3 2
4 3
1
0 0
5
2
(c)
1
6
2
4
6
8
10
12
14
16
18
Flax
Hemp
0 0
MAPP/Fibre [%]
2
4
6
8
10
12
14
16
18
MAPP/Fibre [%]
Sisal Flax
Hemp
Sisal
Fig. 3. Effect of MAPP:NF ratio on (a) Stiffness (b) Strength (c) Impact 30% NF. Fig. 4. Effect of MAPP:NF ratio on (a) Stiffness (b) Strength (c) Impact for 50% NF composite.
Second is the negative effect of delignification caused by alkalinisation. Both factors act against each other. Therefore plateau form is found for both flax and hemp while sisal reaches its saturation level of NaOH treatment at 5% concentration. 5% NaOH pre-treatment seems to be sufficient for the three types of fibres. On the other side, strength behaviour is still positively affected with NaOH% compared to the stiffness behaviour. As mentioned,
sisal strength improves 8% at 10% NaOH treatment whereas sisal stiffness decreases 8% between 5% and 10% NaOH treatment. Flax and hemp still show an obvious enhancement in strength. Hemp has a linear trend increase while flax shows a plateau form. Stiffness behaviour reaches its optimum before that of the strength. This difference in the optimum values between strength and
130
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
Table 3 Optimum MAPP:NF ratios (Improvement %) for 30% and 50% NF. 30 wt%
Stiffness Strength Impact
50 wt%
Flax
Hemp
Sisal
Flax
Hemp
Sisal
10 (3.7) 6.7 (19.1) 6.7 (23.8)
6.7 (11.7) 6.7 (13.5) 10 (12.9)
6.7 (11.5) 6.7 (8.4) 6.7 (34.1)
13.3 (20.6) 6.7 (66.1) 6.7 (60.1)
10 (17.0) 13.3 (28.5) 6.7 (39.5)
16.7 (6.3) 6.7 (12.8) 6.7 (19.2)
30 wt% 18 16
(a)
30 wt%
70
(a) 60
Improvement [%]
14
MAPP:NF [%]
12 10 8 6
50 40 30 20
4
10
2
0 Flax
0 Flax
Hemp E-modulus
Strength
Sisal
(b) 60
Improvement [%]
(b)
14
MAPP:NF [%]
Sisal Impact
50 wt%
70
16
Strength
Impact
50 wt%
18
Hemp E-modulus
12 10 8
50 40 30 20
6 4
10
2
0 Flax
0 Flax
Hemp E-modulus
Strength
Sisal Impact
Fig. 5. Optimum MAPP:NF ratios on stiffness, strength and impact for (a) 30 wt% (b) 50 wt%.
stiffness can be related to the nature of fibres’ orientation at the injection-moulded samples’ surfaces [30]. The fibres are better oriented and shows almost no sign of fibres agglomeration. Thus the strength values compared to those of stiffness are already way behind and hence NaOH or similar treatments affect the strength more positively. The impact values of the un-notched specimens. As expected from the stiffness and strength results; the impact values show also an improvement of 13% and 28% for flax and sisal. Hemp fibres, however, has no change. This can be attributed to the difference in the fibrils orientation. Flax fibrils are almost 10° to the axis [31]. This gives flax some sort of elasticity compared to hemp. Sisal has no orientation angle. However it has a little bit elongation because of the increased quotient of strength to the stiffness. From the results presented in Table 2, a concentration of 5–10% NaOH is selected as an almost optimum range to be adopted. However, the improvement attained by the increase of NaOH concentration from 5% to 10% is not significant especially in stiffness
Hemp
E-modulus
Strength
Sisal
Impact
Fig. 6. Maximum mechanical properties at optimum MAPP:NF ratios for (a) 30 wt% (b) 50 wt%.
results for all types of fibres or strength results for flax. The improvement of strength in flax at 10% is not accompanied with an improvement in impact. Therefore 10% pretreatment is not so attractive and 5% is enough to be selected as a constant NaOH concentration pretreatment. Also from Table 2, it can be inferred that sisal out of the three fibre types is the least sensitive fibre to the alkalinisation treatment in comparison to the flax and hemp fibres. The significance of the results listed in Table 2 was tested by two tailed t-test with 95% confidence. T-test proved the significance for the pairs (0% vs 5%), (5% vs 10%) and (0% vs 10%). Only the impact results of hemp seem totally insignificant. 3.2. Effect of fibre type and content Fig. 2 presents the mechanical properties spectrum of the 30% and 50% natural fibre polypropylene composites with the use of type (A) granulated form MAPP of 1:10 ratio pre-treated with 5% NaOH. Fig. 2a presents the stiffness results. Addition of fibres enhanced the stiffness from almost 1.5 GPa to 5.2–5.6 GPa for the
131
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
Fig. 7. Effect of MAPP:NF ratio on impact surface (a) 0.0%; (b) 6.7%; (c) 16.7%.
3.3. Effect of copolymer to fibre ratio This section deals with the main problem claimed in this work which is the diversity in the optimisation rules for implementing
the MAPP as a coupling agent in the natural fibre thermoplastic composites. Figs. 3 and 4 present the effect of different MAPP:NF ratio on the mechanical properties of natural fibre composites at 30% and 50% fibre content respectively. Maximum value of the curve plateau is considered the optimum value. For 30% natural fibre as shown in Fig. 3a, MAPP imparts almost 8%, 10% and 12% improvement in stiffness for flax, hemp and sisal respectively. Regarding strength improvement after the addition of MAPP shown in Fig. 3b; 15%, 20% and 4% are attained for the flax, hemp and sisal. Sisal shows the least improvement with the use of the MAPP copolymer. This can be attributed to the less number of 120
(a)
0%
6,7%
10,0%
13,3%
16,7%
Weight [%]
100 80 MAPP:NF increases in the arrow direction
60 40 20 0 100
150
200
250
300
350
400
450
500
550
600
Temperature [°C] Derivative Weight [%/°C]
30% natural fibre loading and 7.6–8.5 GPa for the 50% loading. Flax and hemp have close stiffness results for both fibre loadings 30% and 50% with a slight superiority of 7–12% in comparison to the sisal behaviour. Fibre content has obviously a positive effect where the 50% loading has stiffness 42%, 53% and 46% more than those of the 30% loading for flax, hemp and sisal respectively. Flax has relatively lower stiffness than hemp because of the 10° inclination of the fibrils to the stem axis [31]. Fig. 2b shows the strength results. Again the coarser sisal fibres shows the least strength compared to the bast fibres flax and hemp. However sisal has the largest jump in strength, where strength of the 50% group is 18%, 19% and 25% more than the 30% group for flax, hemp and sisal respectively. Close strength results for flax and hemp because both fibres’ constituents and size are similar. This significant increase of strength specially with the 50% flax can be contributed to the surface matrix adhesion efficiency and its impact on the load transfer mechanism [32–34]. Similar strength and different stiffness values between flax and hemp results, as expected, in lower elongation for hemp. Flax fibrils have the chance to elongate along the stem axis. Thus results in the composite elongation shown in Fig. 2c. The elongation of the 50% composites decreased to almost half of the values attained by 30% composites. This results in the non-consistent behaviour of the impact results shown in Fig. 2d. The difference between the averages of both the 30% and the 50% is insignificant compared to the inner deviation in each group. 0.4 kJ/m2 difference exists between average values of 6.8 and 6.4 kJ/m2 for the 30% and 50% respectively. While the ranges of the impact results are 1.1 and 1.4 kJ/ m2 for both fibre contents. The significance of the results shown in Fig. 2 was also tested by two tailed t-test with 95% confidence. T-test proved the significance for the pairs (flax vs hemp), (hemp vs sisal) and (flax vs sisal). Two thirds of the results proved significance at 95%.
1,6 1,4
(b)
0%
6,7%
10,0%
13,3%
16,7%
1,2 1 0,8 0,6 0,4 0,2 0 250
300
350
400
450
500
550
Temperature [°C] Fig. 8. Effect of MAPP:NF ratio for 50% Flax PP composites coupled with MAPP (Type A) on (a) TGA; (b) DTG.
132
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
Table 4 Effect of MAPP:NF% and type on melt temperature and enthalpy of 50% flax PP composites. MAPP type
MAPP:NF [%]
Melt temperature [°C]
Melt enthalpy [j/g]
Recrystallisation temp. [°C]
Recrystallisation enthalpy [j/g]
Type A
0 6.7 10 13.3 16.7
167.31 167.00 166.86 166.28 166.62
51.46 54.55 58.66 56.00 44.97
132.26 128.87 129.18 128.99 128.19
67.36 70.94 75.65 69.36 61.01
Type B
0 6.7 10 13.3 16.7
167.81 166.42 164.51 165.69 164.29
51.46 48.75 58.87 50.48 50.84
available sites for copolymerisation on the fibre’s surface due to their relative bigger size. Impact results shown in Fig. 3c ensures the efficiency of MAPP copolymerisation compared to the non MAPP composites. Flax has its optimum results at 10%, 6.7%, 6.7% for stiffness, strength and impact respectively. Hemp optimum values are at 6.7%, 6.7% and 10% while sisal values are 6.7%, 6.7%, 6.7%. Table 3 lists the aforementioned values. Flax has almost 7% improvement for stiffness at its optimum values while hemp and sisal have 11% improvement. For strength, sisal has the least improvement of 5.4% while flax has the best value of 19% and hemp has 16% increase. Finally for impact, the sisal has the best improvement more than the other fibres. It is also noted that higher MAPP:NF ratio than the optimum induces slightly a negative effect on the corresponding property. For 50% natural fibre, flax has optimum values of 13.3%, 6.7% and 6.7% for stiffness, strength and impact respectively. Hemp values are 10%, 13.3% and 6.7% while sisal values are 16.7%, 6.7% and 6.7%. Fig. 4a illustrates the stiffness results. Flax stiffness improves from 6.8 GPa to 8.3 GPa at MAPP:NF = 13.3%. Hemp, similar to its behaviour at 30% loading, improves from 7.3 GPa to 8.5 GPa at lower MAPP:NF = 10%. Sisal slightly improves from 7.6 GPa to 7.6 GPa. Fig. 4b shows the strength development with the increase of MAPP:NF ratio. 50% Flax PP composites without coupling agent show a poor strength of 30.7 MPa which is only 10% more than the pure PP. However implementing MAPP enhances it immediately. Therefore at the first investigated point, namely MAPP:NF = 10%, strength increases of the flax PP composite to 49.6 MPa. Then more MAPP does not result in an improvement rather a slight strength reduction to 48.6 MPa. Hemp PP composite without MAPP has a higher strength from the beginning of 41.6 MPa and then improves to 50.8 MPa at 6.7% MAPP:NF. The same goes for sisal which improves from 43.9 MPa to 51.4 MPa at 6.7% MAPP:NF. The same behaviour of slight strength reduction at higher MAPP is observed for the three types of fibres. Finally the impact behaviour is shown in Fig. 4c. The ratio of 6.7% is enough for all fibre types to reach the maximum values. Statistical t-test analysis is applied and confidence of more than 95% is found for the results of stiffness, strength and elongation. Fig. 5 summarises the previous section for the 30% and 50% fibre loadings. It is obvious that 6.7% of MAPP:NF is a suitable solution. This 6.7% ratio is more consistent with the 30% fibre loading. In case of 50% fibre loading, the hemp strength is the only deviation where its optimum at 13.3%. However, the difference in hemp strength between 6.7% and 13.3% is less than 3 MPa. Only stiffness which requires 10% for 30% flax and 10–16.7% for all fibres in case of the 50%. But how much is needed to increase the MAPP: NF ratio? Fig. 6 shows the improvement percentage at the optimum MAPP:NF. Improvement% is calculated according to Eq. (1).
In general, Stiffness improvement has the least significance compared to the strength and impact improvements. In other words; Stiffness shows improvement at higher MAPP:NF where the tensile and impact strengths are getting maximised at lower MAPP:NF ratios. This remark can be explained in terms of the more coupling sites which are reflected in better fibre/matrix adhesion and hence higher stiffness. On the other side, the tensile and impact strengths are properties related to post-elastic zone range where the failure either by pull-out or debonding or even fibre failure promotes the failure regardless how stiff is the composite. Another interesting notice is noticed in Fig. 6a where the improvement of the three fibres at 30% lies in a close range relatively when compared to the deviation between flax and sisal at 50%, Fig. 6b.This is an evidence that natural fibre type and shape play an important role in the composite behaviour and cannot be considered as fillers. Fig. 7 shows SEM fracture surfaces at different MAPP:NF ratios. Composite free from MAPP copolymer shows long pulled out fibres. At 6.7% MAPP:NF for the 30% flax, the fracture shows less pulled out fibre length and some fibre debonding sites of transversal fibres. At more MAPP:NF ratio of 16.7%, the fibres are well adhered to the PP matrix and the polymer remnants are present on the broken pulled fibre surface. Fig. 8 shows the TGA and the differential thermal gravimetric DTG of the flax PP composites copolymerised with granulate MAPP of Type A. Decomposition proceeds by losing the absorbed water then hemicelluloses followed by alpha cellulose and finally by lignin [35]. The DTG curve in Fig. 8 shows no charring evidence around 500 °C. This matches the finding of Doan [36] and against what reported by Bhaduri [37]. Composites show two peaks’ with maxima ranging in 369.4– 374.8 °C and 432.4–481 °C corresponding to the decomposition
Improvement% ¼ ½ðP optimum MAPP : NF PMAPP : NF ¼ 0Þ= P MAPP : NF ¼ 0 100 where P is the property (stiffness, strength, impact).
ð1Þ Fig. 9. DSC results of MAPP:NF ratio for 30% Flax PP composites coupled with MAPP (Type A).
133
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
of flax and PP respectively. The recorded decomposition behaviour is similar to that of Doan [36]. The temperatures at the peaks’ maxima are decreasing slightly as the MAPP:NF ratio increases. This note is valid for 0%, 6.7% and 10% of MAPP:NF ratios. For higher ratios of MAPP:NF with 13.3% and 16.7%; the decomposition takes place at significant lower temperatures. This is an evidence that the excess MAPP induces a dispersion effect rather than a coupling
Flax 30%
7000
(b)
9000
Stiffness [MPa]
6000 5500 5000 4500 4000
8000 7000 6000 5000 4000
3500 3000
3000 0
5
10
15
20
0
Type (A)
Type (B)
Strength [MPa]
43 41
Type (B)
(d)
55
45
20
Flax 50%
60
50 45 40 35
37 35
30 0
5
10
15
20
0
5
MAPP/Fibre [%] Type (A)
Type (B)
Type (A)
15
20
(f)
3,5
Elongation [%]
8 7 6 5 4 3 2
Type (B)
Flax 50%
4
(e)
9
10
MAPP/Fibre [%]
Flax 30%
10
Elongation [%]
15
Type (A)
39
3 2,5 2 1,5 1 0,5
1
0
0 0
5
10
15
0
20
10
15
MAPP/Fibre [%]
Type (A)
Type (A)
Type (B)
20
Type (B)
Flax 50%
10
(g)
9
5
MAPP/Fibre [%]
Flax 30%
10
(h)
9 8
8
Impact [kJ/m²]
Impact [kJ/m²]
10
MAPP/Fibre [%]
(c)
47
5
MAPP/Fibre [%]
Flax 30% 49
Strength [MPa]
Flax 50%
10000
(a)
6500
Stiffness [MPa]
effect on the fibre matrix interface as reported by [38]. This also may explain why the wetting seems better although the mechanical properties are decreasing when the MAPP:NF ratio is more than optimum. DSC results are given in Table 4. Results show similar behaviour reported by Lee [39]. The melting temperature decrease with the increase of MAPP:NF ratio. The reduction in temperature is too
7 6 5 4 3
7 6 5 4 3
2
2
1
1
0
0 0
5
10
15
MAPP/Fibre [%] Type (A)
Type (B)
20
0
5
10
15
MAPP/Fibre [%] Type (A)
Type (B)
Fig. 10. Effect of MAPP:NF on (a,b) Stiffness (c,d) Strength (e,f) Elongation (g,h) Impact.
20
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
small to be felt as it is almost 0.5%. While in Fig. 9, the melt enthalpy shows an increasing trend or even a plateau effect and a maximum value at 10% MAPP:NF then with a steeper trend goes down. The reduction in enthalpy is greater as it drops to 25% for the MAPP of Type A and 14% for MAPP of type B. This assures the TGA result of the dispersion phenomenon above 10% MAPP:NF. Interesting is also the similarity in behaviour of recrystallization where 10% MAPP:NF has the maximum recrystallization energy. This can be correlated to the maximum efficiency of coupling between the fibre and the polymer. Higher MAPP:NF decreases the recrystallization because the excess MAPP does not react with itself when do not undergo grafting [13].
120
0%
6,7%
10,0%
13,3%
16,7%
80 60 40 20 0 100
150
200
250
300
350
400
450
500
550
600
Temperature [°C] 1,8
Derivative Weight [%/°C]
3.4. Effect of copolymer type For the sake of comparison between MAPP types A and B, Fig. 10 shows the mechanical properties corresponding to both 30% and 50% flax with the different forms of the MAPP. Granulate MAPP, Type A, always shows near parabolic forms with optimum values while the powder MAPP, Type B, shows that even after reaching the optimum value, constant or slightly increasing trend is still going on. Fig. 10a and b show the stiffness behaviour for both fibre loadings of 30% and 50%. Type B does not impart improvement for the 30% composite stiffness whereas Type A imparts almost 7% improvement at the 6.7% MAPP:NF ratio. For the 50% fibre loading, both types A and B result in an improvement namely 20% and 15%. It is important to note that the optimum MAPP:NF ratio is the same for the 30% and 50% fibre loadings. Fig. 10c and d show the strength behaviour. Type B has a better effect than Type A on the composite strength. Type A has a faster strengthening effect than Type B. Type A has its optimum MAPP:NF ratio of 6.7% in the 30% composite whereas Type B has a slower response and reached its maximum at 13.3%. In case of 50% composite, Type A again has a maximum point at 6.7% of MAPP:NF. The Type B shows an increasing linear trend of composite strength in the range of 6.7–16.7% without reaching an optimality point. Fig. 10e shows an interesting observation. Type A composite has a decreasing elongation with more MAPP:NF whereas the Type B has a softer effect where the elongation slightly changes in the 30% and 50% fibre loading composites. The same behaviour for Type A and B is observed in Fig. 10f with lower elongation values. Fig. 10g shows the impact results of the un-notched specimens for the composites filled with 30% flax. Both types A and B are coincident in the 0–6.7% range then Type B continues positively in comparison to Type A. Again the same behaviour is repeated in Fig. 10h for the 50% filled composites. However, the change is less positively for Type B and negatively for Type A. The better behaviour of Type B in general can be attributed to the better thermal stability of the composites with higher molecular weight of the powder form MAPP. As mentioned earlier; Type B has normally lower acid number and hence higher melt flow index MFI, better mixing, and better diffusion [28]. This explains the good impregnation of the coupling inside the fibres. Type A shows a drop in strength after reaching the maximum value because excess MAPP does not result in more coupling with fibres but results in self reaction leading to a weak layers instead. Type A has faster thermal degradation as shown in Fig. 8. This is due to more MAPP: NF compared to the more thermal stability of MAPP Type B, as shown in Fig. 11. The DTG curve, shown in Fig. 11b, shows two peaks with ranges of 357.3–374.8 °C and 480–486 °C. The increase of MAPP:NF ratio to 16.7% leads to 48 °C degradation of the thermal stability which is less than 6 °C degradation for the Type A composites. The result, that Type B has more thermal stability, matches the results of Kim [27] where the lower molecular weight copolymer has lower thermal stability behaviour. excessive MAPP:NF does not impart enhancement of composite strength
(a)
100
Weight [%]
134
1,6
(b)
0%
6,7%
10,0%
13,3%
16,7%
1,4 1,2 1
0,8 0,6 0,4 0,2 0 250
300
350
400
450
500
550
Temperature [°C] Fig. 11. Effect of MAPP:NF ratio for 50% Flax PP composites coupled with MAPP (Type B) on (a) TGA (b) DTG.
and impact. It is also noted that the change in the first peak around the 370 °C is more affected in case of Type B. This is also matching with the results of Lee [25] who worked with bamboo fibre PP composites. DSC results are given in Table 4. Results show similar behaviour reported by Lee [38]. The decrease in melting temperature and enthalpy is slightly larger in MAPP of Type B compared to Type A. This is attributed to the more reaction and higher diffusivity of Type B [31]. Finally the effect of copolymer type on the water absorption level by natural fibre thermoplastic composite. Fig. 12 shows better behaviour for the granulated Type A MAPP where the water absorption of Type B composites is about 16% more than those of Type A composites. A decreasing trend is noticed with the increase of MAPP:NF ratio. This points out to sufficient interface reaction due the Type A copolymer compared to the Type B. This is attributed to the high MA grafting of Type A [27].
Fig. 12. Effect of MAPP:NF ratio and MAPP type on water absorption.
A. El-Sabbagh / Composites: Part B 57 (2014) 126–135
4. Conclusions The behaviour of the NFTC mechanical properties is dependent on the fibre type. In other words; fibre type plays an important role in the strengthening and cannot be considered as fillers only. Pretreatment of 5% NaOH for 24 days at room temperature is sufficient to have more efficient fibre surface for coupling with the thermoplastic matrix and hence further processing. The optimum MAPP:NF is searched for and the following points are deduced: (1) MAPP: NF ratio of 6.7% is enough for having a high improvement in mechanical properties. However for reaching Maximum value in the studied property (Stiffness, strength, impact), more MAPP:NF ratio is needed. This result suggests that lower ratio tan 6.67% can result in an acceptable improvement in strengthening. (2) Optimality for each property (stiffness, strength and impact) are defined for flax, hemp and sisal natural fibres as shown in Table 3. (3) In general, it is noticed that the required MAPP:NF ratio for maximising the Stiffness is less than that needed for maximising both tensile strength and impact strength properties. This is explained in terms of the increasing probability that failure scenarios are most likely to happen in the fibre or the matrix/fibre interfaces. These failure favourite places for failure are increasing with the increase of fibres. (4) Granulated Type A MAPP is better in the range of [0–6.7%] for stiffness and strength and vice versa. For impact, both types go on coincidently then type B composites result in higher impact properties. Composites of high percentage Type B MAPP has more thermal stability than the Type A composites. This is related to the more molecular weight of Type B. Water absorption of the Type A MAPP natural fibre composites are better than the Type B MAPP composites.
[11]
[12]
[13] [14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
[22]
[23]
[24] [25] [26] [27]
Acknowledgements [28]
Thanks are given to the Fachagentur Nachwachsende Rohstoffe e.V. (FNR) agency for financing. Also the help of Dr Leif Steuernagel, Dr Dieter Meiners and Prof Ziegmann through the mutual deep discussions.
[29]
[30]
References [31] [1] Torres FG, Arroyo OH, Gomez C. Processing and mechanical properties of natural fiber reinforced thermoplastic starch biocomposites. J Thermoplast Compos Mater 2007;4:207–23. [2] Garcia M, Garmendia I, Garcia J. Influence of natural fiber type in ecocomposites. J Appl Polym Sci 2008;107:2994–3004. [3] Bledzki AK, Sperber VE, Faruk O. Natural and wood fibre reinforcement in polymers. Shrewsbury, UK: Rapra Technology Limited; 2002. [4] Bogoeva-Gaceva G, Avella M, Malinconico M, Buzarovska A, Grozdanov A, Gentile G, et al. Natural fiber eco-composites. Polym Compos 2007;28:98–107. [5] Van de Velde K, Kiekens P. Development of a flax/polypropylene composite with optimal mechanical characteristics by fiber and matrix modification. J Thermoplast Compos Mater 2002;15:281–300. [6] Saheb DN, Jog JP. Natural fiber polymer composites: a review. Adv Polym Tech 1999;18:351–63. [7] Panthapulakkal S, Sain M. Injection-molded short hemp fiber/glass fiberreinforced polypropylene hybrid composites—mechanical, water absorption and thermal properties. J Appl Polym Sci 2007;103:2432–41. [8] Bos, H. The Potential Of Flax Fibers As Reinforcement For Composite Materials. PhD Thesis. Technische Universiteit Eindhoven: Eindhoven, The Netherland, 2004. [9] Taha I, Ziegmann G. A comparison of mechanical properties of natural fiber filled biodegradable and polyolefin Polymers. J Compos Mater 2006;40:1933–46. [10] Espert A, Vilaplana F, Karlsson S. Comparison of water absorption in natural cellulosic fibres from wood and one-year crops in polypropylene composites
[32]
[33]
[34]
[35] [36]
[37] [38]
135
and its influence on their mechanical properties. Compos Part A 2004;35:1267–76. Moran J, Alvarez V, Petrucci R, Kenny J, Vazquez A. Mechanical properties of polypropylene composites based on natural fibres subjected to multiple extrusion cycles. J Appl Polym Sci 2007;103:228–37. Li X, Tabil LG, Panigrahi S. Chemical treatments of natural fiber for use in natural fiber-reinforced composites: a review. J Polym Environ 2007;15:25–33. Keener TJ, Stuart RK, Brown TK. Maleated coupling agents for natural fibre composites. Compos Part A 2004;35:357–62. Arbelaiz A, Fernandez B, Cantero G, Llano-Ponte R, Valea A, Mondragon I. Mechanical properties of flax fibre/polypropylene composites. Influence of fibre/matrix modification and glass fibre hybridization. Compos Part A 2005;36:1637–44. George J, Sreekala MS, Thomas S. A review on interfacial modification and characterization of natural fiber reinforced plastic composites. Polym Eng Sci 2001;41:1471–85. Eichhorn SJ, Baillie CA, Zafeiropoulos N, Mwaikambo LY, Ansell MP, Dufresne A, et al. Review current international research into cellulosic fibres and composites. J Mater Sci 2001;36:2107–31. Yuan X, Jayaraman K, Bhattacharyya D. Effects of plasma treatment in enhancing the performance of woodfibre-polypropylene composites. Compos Part A 2004;35:1363–74. Ragoubi M, Bienaimé D, Molina S, George B, Merlin A. Impact of corona treated hemp fibres onto mechanical properties of polypropylene composites. Ind Crops Prod 2010;31:344–9. Mohanty AK, Misra M, Drzal LT. Surface modifications of natural fibers and the performance of the resulting biocomposites. Compos Interface 2001;8:313–43. Hill ASC, Abdul Khalil HPS, Hale MD. A study of the potential of acetylation to improve the properties of plant fibres. Ind Crops Prod 1998;8:40–53. Agrawal R, Saxena NS, Sharma KB, Thomas S, Sreekala MS. Activation and crystallization kinetics of untreated and treated oil palm fiber reinforced phenol formaldehyde composites. Mater Sci Eng 2000;277:77–82. Sain M, Park SH, Suhara F, Law S. Flame retardant and mechanical properties of natural fibre-pp composites containing magnesium hydroxide. Polym Degrad Stab 2004;83:363–7. Zampaloni M, Pourboghrat F, Yankovich SA, Rodgers BN, Moore J, Drzal LT, et al. Kenaf Natural fiber reinforced polypropylene composites: a discussion on manufacturing problems and solutions. Compos Part A 2007;38:1569–80. Bos HL, Muessig J, Van den Oever. Mechanical properties of short-flax-fibre reinforced compounds. Compos Part A 2006;37:1591–604. Pimenta MTB, Carvalho AJF, Vilaseca F, Girones J, Lopez JP, Mutje M, et al. Sodatreated sisal/polypropylene composites. J Polym Environ 2008;16:35–9. Graupner N. Application of lignin as natural adhesion promoter in cotton fibrereinforced poly(lactic acid) (PLA) composites. J Mater Sci 2008;43:5222–9. Kim H, Lee B, Choi S, Kim S, Kim H. The effect of types of maleic anhydridegrafted polypropylene (MAPP) on the interfacial adhesion properties of bioflour-filled polypropylene composites. Compos Part A 2007;38:1473–82. Gueldogan Y, Egri S, Rzaev Z, Piskin E. Comparison of maleic anhydride grafting onto powder and granular polypropylene in the melt by reactive extrusion. J Appl Polym Sci 2004;92:3675–84. El-Sabbagh A, Steuernagel L, Ziegmann G. Effect of fibre/matrix chemical modification on the mechanical properties & water absorption of extruded flax/polypropylene composite. J Appl Polym Sci 2009;111:2279–89. Taha I, Steuernagel L, Ziegmann G. Optimization of the alkali treatment process of date palm fibres for polymeric composites. Compos Interfaces 2007;14:669–84. Charlet K, Jernot J, Eve S, Gomina M, Breard J. Multi-scale morphological characterisation of flax: from the stem to the fibrils. Carbohydrate Polym 2010;82:54–61. Elsabbagh A, Steuernagel L, Ziegmann G. Modelling of strength and stiffness behaviour of natural fiber reinforced polypropylene composites. Polym Compos 2009;30:510–9. Taha I, El-Sabbagh A, Ziegmann G. Modelling of strength and stiffness behaviour of natural fiber reinforced polypropylene composites. Polym Polym Compos 2008;16:295–302. El-Sabbagh A, Steuernagel L, Ziegmann G. Modelling of the mechanical properties of flax fibre/polypropylene composite. In Proceedings of the Polymer Processing Society 24th Annual Meeting, Salerno, Italy, 15–19 June 2008; PPS-24. Santos F, Mauler R, Nachtigall S. Effectiveness of maleated- and silanized-pp for coir fiber-filled composites. J Reinf Plast Polym 2009;28:2119–29. Doan T, Brodowsky H, Maeder E. Jute fibre/polypropylene composites II. Thermal, hydrothermal and dynamic mechanical behaviour. Compos Sci Technol 2007;67:2707–14. Bhaduri S, Mathew A, Day M, Pandey S. Thermal behaviour of jute fibre and its components Part-1, D.S.C. studies. Cellulose Chem Technol 1994;28:391–9. Lee S, Kang I, Park B, Doh G. Effects of filler and coupling agent on the properties of bamboo fiber-reinforced polypropylene composites. J Reinf Plast Comp 2009;28:2589–604.
