Crab Shell Chitin Whisker Reinforced Natural ... - ACS Publications

In a previous work (part 1), nanocomposite materials were obtained using a latex of either unvulcanized or prevulcanized natural rubber as the matrix ...
0 downloads 0 Views 372KB Size
Biomacromolecules 2003, 4, 666-674

666

Crab Shell Chitin Whisker Reinforced Natural Rubber Nanocomposites. 2. Mechanical Behavior Kalaprasad Gopalan Nair and Alain Dufresne* Centre de Recherches sur les Macromole´ cules Ve´ ge´ tales (CERMAV-CNRS), Universite´ Joseph Fourier, BP 53, F 38041 Grenoble Cedex 9, France Received November 13, 2002; Revised Manuscript Received January 22, 2003

In a previous work (part 1), nanocomposite materials were obtained using a latex of either unvulcanized or prevulcanized natural rubber as the matrix and a colloidal suspension of crab chitin whiskers as the reinforcing phase. The mechanical behavior of the resulting nanocomposite films was analyzed in both the linear and the nonlinear range in the present study. The effects of the filler and processing technique were evaluated, and the results are discussed based on the knowledge of the structural morphology and swelling behavior reported in our previous work. The reinforcing effect of chitin whiskers strongly depended on their ability to form a rigid three-dimensional network, resulting from strong interactions such as hydrogen bonds between the whiskers. The results emanating from the successive tensile test experiments give clear evidence for the presence of a three-dimensional chitin network within the evaporated samples. Cross-linking of the matrix was found to interfere with the formation of this network. Introduction Natural rubber (NR) is one of the most important elastomers widely used in industrial and technological areas. The properties of NR can be tailored by the addition of fillers of varying surface chemistry and aggregate size/aspect ratio to suit the application concerned. Carbon black and silica are the main fillers used in the compounding recipes.1,2 Short fibers can also be used to reinforce polymers in order to improve or modify certain mechanical properties of the host matrix for specific applications.3 The reinforcement of rubber with fibers combines the elastic behavior of the rubber with the strength and stiffness of the reinforcing phase.4-6 The use of various natural fibers such as bamboo, coir, and oil palm as reinforcing agents in rubber matrixes were reported.7-12 There is currently a considerable interest in processing polymeric composite materials filled with nanosized rigid particles (essentially inorganic but can be organic also). This class of material is called “nanocomposites”. This growing interest in the field of nanocomposites originates from both the point of view of fundamental property determination and the development of new materials to meet a variety of applications. Indeed, because the building blocks of a nanocomposite are of nanoscale, they have an enormous interface area, and therefore there are a lot of interfaces between the two intermixed phases compared to usual microcomposites. The interfacial specific area can be of the order of a few 100 cm2 per gram of material, and the effect of interfacial phenomena are assumed to be noticeable on the bulk properties. In addition to this, the mean distance between particles is all the smaller since their size is small, * To whom correspondence [email protected]).

should

be

addressed

(e-mail:

favoring filler-filler interactions. Moreover the improvement of some properties, such as the modulus of the material, is often achieved without much lowering of properties such as impact strength or ability to plastic deformation. Nowadays the development of nanocomposites using variety of nanofillers is more restricted due to various reasons. These include, on one hand, the limited availability and cost of nanofillers and, on the other hand, the strong tendency of nanoparticles to aggregate, thus preventing a high level of dispersion within the polymer matrix, which is the key parameter required for superior mechanical properties. Cellulose whiskers were extensively used as model fillers in several kinds of polymeric matrixes, including synthetic13-18 and natural ones.19-24 Composites prepared by casting mixtures of tunicin (an animal cellulose) whisker suspensions and polymer lattices showed excellent mechanical properties, especially at T > Tg of the matrix and even at low whisker content by virtue of the formation of a cellulose whisker network, in the polymer matrix. The formation of this rigid network, resulting from strong interactions between whiskers, was assumed to be governed by a percolation mechanism.13,15 This hydrogen-bonded network induced a thermal stabilization of the composite up to 500 K, the temperature at which cellulose starts to decompose. Suspensions of chitin crystallites can be prepared as described by Revol and Marchessault25-28 by acid hydrolysis of chitin. In preliminary studies, we used chitin whiskers obtained from squid pen29 and Riftia tubes30 as the reinforcing phase in a host polymeric matrix. In the first part of the present paper,31 nanocomposite material was prepared by incorporation of chitin whiskers obtained from crab shells in a natural rubber (NR) matrix. Both unvulcanized and vulcanized NR systems were used as matrix. For unvulcanized NR latex based nanocomposites, two processing

10.1021/bm0201284 CCC: $25.00 © 2003 American Chemical Society Published on Web 03/06/2003

Chitin Reinforced Rubber

techniques, viz., water evaporation and freeze-drying followed by a hot-pressing step, were used to investigate the effect of the processing method on the material properties. The morphology and swelling behavior of the composite films in an organic medium were investigated. In the previous study, it was concluded that the whiskers form a rigid network which is assumed to be governed by a percolation mechanism in the evaporated samples only. Strong fillermatrix interactions were also reported. In the present study, the mechanical behavior of chitin whiskers/NR nanocomposites is analyzed in both the linear and the nonlinear range. The effect of the filler content and processing technique is evaluated. Results are discussed on the basis of the knowledge of the structural morphology reported in the first part of the paper.31 Experimental Section Materials. The details of the materials used such as chitin whiskers and NR latex, and their nanocomposite processing techniques, have been described in our previous paper.31 The chitin whiskers, prepared by acid hydrolysis of chitin from crab shells, consisted of slender parallelepiped rods with an average length around 240 nm and an aspect ratio close to 16. The specific area of this filler was estimated to be around 180 m2‚g-1. Both unvulcanized and prevulcanized natural rubber lattices were used as the matrix. After the aqueous suspensions of chitin whiskers and rubber were mixed and stirred, solid composite films were obtained by casting and evaporating methods. For unvulcanized systems, a freezedrying and subsequent hot-pressing processing technique was also used. Dynamic Mechanical Analysis. Dynamic mechanical tests were carried out with a spectrometer RSA2 from Rheometrics working in the tensile mode. The value of 0.05% for the strain magnitude was chosen so as to be in the domain of the linear viscoelasticity of the material. The samples were thin rectangular strips with dimensions about 25 × 4 × 1 mm3. The setup measures the complex tensile modulus E*, i.e., the storage component E′ and the loss component E′′. Measurements were performed in isochronal conditions at 1 Hz and the temperature was varied between -100 and 250 °C by steps of 3 °C. Tensile Tests. The nonlinear mechanical behavior of the chitin whisker filled composites was analyzed using an Instron 4301 testing machine in tensile mode, with a load cell of 100 N capacity. The specimen was a thin rectangular strip (∼30 × 5 × 1 mm). The gap between pneumatic jaws at the start of each test was adjusted to 20 mm. The stress-strain curves of the samples were obtained at room temperature at a strain rate d/dt ) 8.3 × 10-3 s-1 (cross-head speed ) 10 mm‚min-1). The true strain  can be determined by  ) ln(L/L0), where L and L0 are the length at the time of the test and the length at zero time, respectively. The true stress σ was calculated by σ ) F/S, where F is the applied load and S is the cross-sectional area. S was determined assuming that the total volume of the sample remained constant, so that S ) S0L0/L, where S0 is the initial cross-sectional area. Stress versus strain curves were plotted,

Biomacromolecules, Vol. 4, No. 3, 2003 667

and the tensile or Young’s modulus (E) was measured from the slope of the low strain region in the vicinity of σ )  ) 0 ([dσ/d]f0). The conventional modulus E100% was obtained from the slope of the straight line plotted between the origin σ )  ) 0 and the point corresponding to a true elongation of 100%. Ultimate mechanical properties were also characterized. The true ultimate stress, or true stress at break, σb ) Fb/S, where Fb is the applied load at break, was reported for each tested sample. Ultimate elongation was characterized by the true ultimate strain, or true strain at break, b ) ln[1 + (∆Lb/ L0)], where ∆Lb is the elongation at break. Mechanical tensile data were averaged over at least three specimens. Successive Tensile Tests. Successive tensile tests were performed to characterize the damage process occurring during tensile tests. They were carried out with the Instron 4301 testing machine previously described. At the beginning of each experiment, the sample was first stretched under a load of 2 N. Then, the experiment consisted of stretching the material, under the conditions described in the previous section (Tensile Tests), up to a certain elongation ∆L1 (cycle 1), then releasing the force down to 2 N, and stretching again the material up to a higher elongation ∆L2 ) 2∆L1 (cycle 2). This procedure was repeated with increasing elongation ∆Li (cycle i), until 30-40 mm elongation was obtained. Five to eight successive tensile tests were performed for each sample. The tensile modulus Ei for each successive cycle was determined as described in the previous section. Differential Scanning Calorimetry. Differential scanning calorimetry (DSC) was performed with a Perkin-Elmer DSC7 calorimeter, fitted with a cooler system using liquid nitrogen. Samples were placed in pressure-tight DSC cells, and at least two individual measurements were carried out at a heating rate of 10 °C/min to ensure perfect reliability of measurements. The glass transition temperature (Tg) was taken as the inflection point of the specific heat increment at the glass-rubber transition. Results and Discussions Dynamic Mechanical Analysis (Linear Range). Dynamic mechanical analysis was performed for all the samples. Both the effect of vulcanization on the properties of composite films prepared by evaporation and the effect of the processing technique on the properties of unvulcanized composite films are presented. UnVulcanized NR Matrix Based EVaporated Films. Figure 1 shows the plot of log(E′/Pa) (storage tensile modulus, Figure 1a) and tan δ (loss angle tangent, Figure 1b) at 1 Hz as a function of temperature for both unfilled and filled unvulcanized NR films prepared by evaporation. At low temperature, E′ remains roughly constant (Figure 1a). It is well-known that E′ values for any glassy polymer are constant around 3 × 109 Pa. It is due to the fact that in the glassy state, molecular motions are largely restricted to vibration and short-range rotational motions. Increasing the amount of chitin whiskers successively increases the values of E′, and the highest modulus is observed for the composite containing 20 wt % whiskers. This sort of enhancement in

668

Biomacromolecules, Vol. 4, No. 3, 2003

Gopalan Nair and Dufresne

Figure 2. Logarithm of the storage tensile modulus E′ vs temperature at 1 Hz for the vulcanized NR matrix obtained by evaporation: (b) first temperature scan corresponding to cast film; (O) second temperature scan corresponding to quenched film.

Figure 1. (a) Logarithm of the storage tensile modulus E′ and (b) loss angle tangent tan δ vs temperature at 1 Hz for chitin whiskers/ unvulcanized NR composites obtained by evaporation: NRev (b), NCH2ev (O), NCH5ev (9), NCH10ev (0), NCH15ev (2), and NCH20ev (4).

modulus even below glass transition temperature is good evidence for the strong reinforcing tendency of chitin whiskers in the NR matrix. A sharp modulus drop appears for all samples around -60 °C, i.e., in the glass-rubber transition zone determined from differential scanning calorimetry (DSC) measurements. This modulus drop corresponds to an energy dissipation phenomenon displayed in the concomitant relaxation process where tan δ passes through a maximum (Figure 1b). This relaxation process involves cooperative motions of long chain sequences. Above Tg the modulus becomes roughly constant over a wide temperature region called the rubbery plateau region (rubbery modulus) except in the case of unfilled NR and low filler content composites. The rubbery modulus is known to depend on the degree of crystallinity of the material. The crystalline regions act as physical cross-links for the elastomer. In this physically cross-linked system, the crystalline domains would also act as filler particles due to their finite size, which would increase the modulus substantially. No melting peak was observed in DSC experiments. However, the presence of a low degree of crystallinity in the NR matrix was evidenced by recording two consecutive DMA experiments (Figure 2). In the first set of experiments, the analysis was carried out by heating the sample from -125 to 150 °C (filled circles in Figure 2) under the conditions described in the Experimental Section. In the second set, the sample was first heated to 50 °C and then kept at this

temperature for 5 min followed by quenching to -125 °C. After that the experiment was carried out in the same manner as described above for the first set so as to get a second DMA curve (open circles in Figure 2). The figure clearly reveals that the rubbery modulus is different for the two curves in the temperature range between -60 and 0 °C. It is obvious that the rubbery modulus is much lower for the curve obtained after heat treatment at 50 °C (around 2 MPa, which is characteristic of fully amorphous polymers) than that for the one corresponding to the first experiment (around 20 MPa). This could be a clear indication of the low degree of crystallinity present in the NR film, and the drop of modulus around 0 °C in Figure 1a can be most probably attributed to the melting of the crystalline regions during the temperature scan and the retention of this amorphous state after subsequent quenching. In a nutshell, the two-step modulus drop observed in Figure 1a for the NR matrix and low filler content composites can be ascribed to the crystallinity of the matrix of the NR matrix. In the terminal zone of all curves, the elastic modulus becomes lower and lower with temperature and the experimental setup fails to measure it due to the flow of the material. For chitin whisker filled NR, a significant increase in rubbery modulus is observed with increasing chitin content. For instance, the relaxed modulus at Tg + 150 °C (∼90 °C) of a film containing only 2 wt % (around 1.34 vol %) of chitin is 7 times higher than the one of the unfilled matrix. For the 20 wt % chitin whisker filled composite, the rubbery modulus is more than 350 times higher than the one of the unfilled matrix. In addition to this high reinforcing effect, for filler contents higher than 5 wt %, a significant improvement in the thermal stability of the composite is also noticed up to 220-230 °C. Normally, chitin starts to degrade at this temperature. A similar effect was previously observed for cellulose whisker based nanocomposite materials.13-21 The reasons for both the high reinforcing effect and the improvement of the thermal stability of these materials were attributed to the formation of a rigid network of cellulose whiskers within the polymer matrix, formed as a result of hydrogen bonding. It was shown that the reinforcing effect strongly depended on the aspect ratio of the cellulose whiskers and thereby on its origin as well as on the processing technique.16 The

Chitin Reinforced Rubber

Biomacromolecules, Vol. 4, No. 3, 2003 669

Figure 3. (a) Logarithm of the storage tensile modulus E′ and (b) loss angle tangent tan δ vs temperature at 1 Hz for chitin whiskers/ vulcanized NR composites obtained by evaporation: PNRev (b), PCH2ev (O), PCH5ev (9), PCH10ev (0), PCH15ev (2), and PCH20ev (4).

Figure 4. (a) Logarithm of the storage tensile modulus E′ and (b) loss angle tangent tan δ vs temperature at 1 Hz for chitin whisker/ unvulcanized NR composites obtained by freeze-drying and hotpressing: NLR (b), NCH5L (O), NCH10L (9), NCH15L (0), and NCH20L (2).

formation of this rigid cellulose whisker network within the host matrix was found to be governed by a percolation mechanism. For chitin whiskers, similar possible interactions are also expected and were evidenced from swelling experiments.31 In this previous study, the percolation threshold was estimated around 4.4 vol %, i.e., around 6.4 wt %. This value is in the range 5-10 wt % for which the thermal stability is improved. Figure 1b shows the variation in tan δ as a function of temperature for composites with various concentrations of chitin whiskers. tan δ exhibits a maximum around -60 °C, a temperature which is not altered by any change in the whisker content. No significant broadening of the peak is observed either. Finally, the intensity of the relaxation process decreases with increasing amount of fillers, and it is directly linked to the drop of the elastic tensile modulus. It is also ascribed to the decrease of matrix material amount, which is responsible for damping properties. Vulcanized NR Matrix Based EVaporated Films. The evolutions of log(E′/Pa) and tan δ at 1 Hz as a function of temperature for vulcanized NR based composites prepared by evaporation are plotted in parts a and b of Figure 3, respectively. Contrary to what was reported for chitin whisker reinforced unvulcanized NR, a single-step modulus drop was reported at Tg for all compositions (Figure 3a). This is an indication of the fully amorphous state of the matrix. In addition, the rubbery modulus of the unfilled matrix around

2 MPa is typical of amorphous materials. This can be well understood since chemical cross-linking suppresses crystallization of the polymers. The rubbery modulus of the vulcanized matrix is increased by the incorporation of whiskers similarly as in the case of the unvulcanized one. The main reason for this similarity in rubbery modulus of the vulcanized and unvulcanized samples during the DMA test is discussed in the coming section. However, the thermal stability of the vulcanized composites is enhanced compared to that of the unvulcanized NR systems (250-300 °C instead of 220-230 °C). Figure 3b shows the evolution of tan δ vs temperature for vulcanized NR based composites. The behavior is similar to the one reported for unvulcanized systems except in the magnitude of the tan δ peak. For the same composition it is much higher for chitin whisker reinforced vulcanized NR. The magnitude of the relaxation process, which is related to the modulus drop, is known to depend on both the number of mobile entities and their contribution to the compliance. For a given composition, the increase of mobile units participating to the main relaxation process in the vulcanized system (amorphous material) should be responsible for the magnitude of the tan δ peak. UnVulcanized NR Matrix Based Hot-Pressed Films. Figure 4 displays the evolutions of log(E′/Pa) and tan δ at 1 Hz as a function of temperature for unvulcanized NR based composites prepared by the freeze-drying and subsequent hot-

670

Biomacromolecules, Vol. 4, No. 3, 2003

Gopalan Nair and Dufresne

Figure 5. Logarithm of the storage tensile modulus E′ at Tg + 160 °C ∼ 100 °C vs weight fraction of chitin whiskers for chitin whisker/ NR composites: (b) unvulcanized evaporated films, (O) vulcanized evaporated films, and (2) unvulcanized hot-pressed films.

Figure 6. Typical stress vs strain curves of chitin whisker/ unvulcanized NR composites obtained by evaporation (T ) 25 °C, d/dt ) 8.3 × 10-3 s-1). The crab shell chitin whisker contents are indicated in the figure.

pressing technique. The two-step modulus drop is much lower than that for the unvulcanized evaporated system (Figure 1a). This is because the chance of formation of crystalline phase in the hot-pressed matrix is negligible due to the effect of experimental conditions such as quenching prior to freeze-drying and the high-temperature heating (100 °C) during the followed hot-pressing step in this method. The curves themselves show the less crystalline state of the freeze-dried and hot-pressed NR matrix. It is seen that the rubbery modulus increases with increasing whisker content. However, for a given composition it seems that the rubbery modulus is higher for composites prepared by the evaporation method. This indicates that the reinforcing effect of chitin whiskers in the hot-pressed NR matrix is much lower than that in the evaporated NR matrix. The nonuniform dispersion of whiskers31 is most probably responsible for the lower reinforcing effect in freeze-dried and hot-pressed samples. This uneven dispersion of chitin whiskers in freeze-dried samples can be due to the absence of chitin network formation. Swelling and morphological analysis carried out in our previous study31 give good evidence and an additional insight for this fact. Higher magnitude of the tan δ peaks shown in Figure 4b also supports the lower reinforcing effect of chitin whiskers in the hot-pressed matrix. Effect of Vulcanization and Processing Technique. The variation of the relaxed modulus, taken at Tg + 160 °C ∼ 100 °C is plotted as a function of the whisker content in Figure 5 for the different types of matrixes. It is seen that the evolution is similar for films obtained by the evaporation technique, the matrix being cross-linked or not. This is an indication of the fact that the linear mechanical behavior is mainly governed by the possible chitin whiskers network, which is assumed to be formed in evaporated samples.31 The main experimental aspect that alters this behavior seems to be the processing technique. Indeed, for all compositions, the relaxed modulus is much lower for the freeze-dried and hot-pressed samples than for evaporated ones. For example for the 20 wt % filled nanocomposite, it is around 25 MPa for the former and around 500 MPa for the latter. This strong difference is most probably ascribed to the possible formation of a continuous percolating chitin whisker network within the NR matrix in the evaporated samples.

Tensile Tests (Nonlinear Range). The tensile mechanical behavior of the chitin whiskers/NR composite films was analyzed at room temperature. Typical stress vs strain curves (nominal data) for the chitin whiskers/unvulcanized NR evaporated composites are shown in Figure 6. For each measurement, it was observed that the strain was macroscopically homogeneous and uniform along the sample, until its break. The lack of any necking phenomenon confirms the homogeneous nature of these composites at the scale of a few hundred nm3. The samples exhibit an elastic nonlinear behavior typical of amorphous polymer at T > Tg. The stress continuously increases with the strain. The polymeric matrix is in the rubbery state and its elasticity from entropic origin is ascribed to the presence of numerous entanglements due to high molecular weight chains. The tensile modulus, tensile strength and elongation at break of the films were determined from the plot of the true stress versus true strain as described in the Experimental Section. The results are collected in Table 1. The tensile modulus and conventional modulus E100% of chitin whisker/ NR composite films are plotted in parts a and b of Figure 7, respectively, for the various matrixes. A clear hierarchy is observed. The modulus is systematically higher for unvulcanized evaporated samples than for vulcanized evaporated ones, which in turn is higher than that for unvulcanized hotpressed composites. The difference in mechanical stiffness between evaporated and hot-pressed unvulcanized samples was already reported in DMA experiments and explained on the basis of the formation of a chitin whisker network in evaporated films. Contrary to DMA experiments, tensile tests show a lower reinforcing effect for chitin whisker filled cross-linked NR matrix compared to unvulcanized systems. This discrepancy could originate from the fact that dynamic mechanical measurements involve weak stresses. The possible interactions between percolating chitin whiskers are not damaged under these weak stresses. Under the higher stress level, as applied in tensile tests, these interactions seem to be partially destroyed. This is an indication that the strength of the chitin network in the cross-linked NR matrix is lower than that in the unvulcanized one. The chemical cross-linking of the

Biomacromolecules, Vol. 4, No. 3, 2003 671

Chitin Reinforced Rubber

Table 1. Mechanical Properties of Chitin Whisker Filled Natural Rubber Using Data Obtained from Tensile Tests: Tensile Modulus (E), Conventional Modulus (E100%), Stress at Break (σB), and Elongation at Break (B) σB (MPa) sample

processing technique

E (MPa)

E100% (MPa)

true

NRev NCH2ev NCH5ev NCH10ev NCH15ev NCH20ev PNRev PCH2ev PCH5ev PCH10ev PCH15ev PCH20ev NRL NCH2L NCH5L NCH10L NCH15L NCH20L

water evaporation

1.7 5.6 17.8

1.8 2.5 5.0

25.5 21.7 25.0

127 229 1.6 2.5

17.0 23.9 1.6 2.1

25.9 52.8 111 1.1 1.4 2.1 4.6 8.7 10.2

6.8 11.4 15.3 0.71 0.99 1.7 3.4 7.2 9.4

water evaporation

freeze-drying and hot-pressing

B (%)

nominal

true

nominal

2.1 2.6 3.7

248 213 192

1099 740 583

29.7 9.8 395 388

7.9 126 11.8 13.7

133 126 351 334

276 252 3252 2728

178 128 20.5 41.6 5.9 15.5 21.7 62.7 83.9

10.5 9.4 5.6 3.4 1.0 2.7 4.4 10.5 14.4

283 261 130 252 176 175 160 179 176

1590 1265 268 1137 483 475 397 498 482

Figure 7. (a) Young’s modulus, (b) conventional modulus E100%, (c) tensile strength, and (d) elongation at break vs chitin content of unvulcanized evaporated (b), vulcanized evaporated (O), and unvulcanized hot-pressed (2) chitin whisker/natural rubber nanocomposites. The solid lines serve to guide the eye.

matrix most probably interferes with the formation of the chitin network. In our previous study,31 the average molecular weight between cross-links (Mc) and the average number of monomer units between cross-links were estimated from swelling experiments. For the unfilled vulcanized matrix, the average number of monomer units between cross-links was found to

be close to 190. From standard geometry, the monomer length is around 0.47 nm, and assuming that a rubber chain is in a straight, elongated fashion between two adjacent crosslinks, the distance between two adjacent cross-links should be around 90 nm. This value is of the same order of magnitude, but lower, as the average length of the chitin whiskers obtained from crab shell (around 240 nm). Hence

672

Biomacromolecules, Vol. 4, No. 3, 2003

Gopalan Nair and Dufresne

Figure 8. Evolution of (a) the force vs elongation and (b) true stress vs true strain for NRev during successive tensile tests.

it is reasonable to think that the chemical cross-linking during vulcanization can interfere with the formation of the chitin network in the NR matrix. The graphs showing tensile strength and ultimate strain versus whiskers content are plotted in parts c and d of Figure 7, respectively, for the different sets of chitin whisker/NR composites. Except at high whisker content, the ultimate properties of the materials are systematically higher for the cross-linked NR based composites than for unvulcanized NR based materials. This phenomenon is obviously ascribed to the high elastic behavior of NR induced by the chemical cross-linking of rubber chains during vulcanization. It is worth noting that the presence of chitin whiskers in the NR matrix results in a significant increase of the modulus as reported above, without much lowering the elongation at break. Successive Tensile Tests. Figure 8a shows the typical evolution of the force vs elongation for the unvulcanized NR matrix film obtained by the water evaporation technique during successive tensile tests carried out as described in the Experimental Section. A significant difference is observed between the curves obtained during stretching (positive elongation rate) and the ones obtained during recovery (negative elongation rate). Moreover after each cycle a residual elongation is reported. During stretching, short macromolecular chains can diffuse within the surrounding entangled lattice but do not tangle again immediately during the recovery step because of the viscous nature of the polymer. Therefore, the curves recorded during positive and negative elongation rate steps do not superimpose and a

Figure 9. Evolution of the relative tensile modulus, viz., the modulus of the composite measured during cycle i divided by the one measured for the same sample during the first stretching cycle, Ei / E1, as a function of i and of the whisker content for (a) unvulcanized hot-pressed, (b) unvulcanized evaporated, and (c) vulcanized evaporated materials.

permanent strain remains, at the time scale of the experiment. Figure 8b shows the evolution of the true stress vs true strain for the same sample. The global evolution is similar, but the stress level at the beginning of each cycle differs because of the permanent strain induced during each successive cycle. These successive tensile tests were performed for each composition, and the tensile modulus Ei for each successive cycle was determined. First we discuss the unvulcanized NR based composite films obtained by the freeze-drying and hotpressing techniques. The effects of the chitin whisker addition and successive tensile tests on Ei are shown in the threedimensional diagram of Figure 9a, and experimental data are collected in Table 2. Figure 9a displays the evolution of the relative tensile modulus, viz., the modulus of the composite measured during cycle i divided by the one measured for the same sample during the first stretching

Biomacromolecules, Vol. 4, No. 3, 2003 673

Chitin Reinforced Rubber Table 2. Tensile Modulus Ei Determined during the Successive Tensile Tests i sample

processing technique

E1 (MPa)

E2 (MPa)

E3 (MPa)

E4 (MPa)

E5 (MPa)

E6 (MPa)

E7 (MPa)

NRev NCH5ev NCH10ev NCH15ev NCH20ev PNRev PCH2ev PCH5ev PCH10ev PCH15ev PCH20ev NRL NCH5L NCH10L NCH15L NCH20L

water evaporation

1.9 18.4 48.1 129.7 231.5 1.6 2.2 6.2 14.3 31.4 73.4 1.4 2.6 3.9 8.5 8.3

4.1 13.3 16.9 48.9 83.4 2.1 1.9 5.3 12.2 21.1 35.2 2.1 3.2 5.5 11.2 11.4

4.1 12.9 16.7 38.6 69.3 2.1 2.5 4.5 9.5 18.0 30.2 2.1 3.7 5.8 15.1 13.7

4.1 11.8 16.5 40.5 68.1 2.2 2.5 4.2 8.6 18.1 26.1 2.1 4.0 6.6 19.6 17.2

4.3 12.7 17.0 46.7 77.3 2.3 2.6 4.2 8.4 17.4 24.4 2.2 4.1 6.9 23.1 21.4

4.7 13.7 18.0

5.1

water evaporation

freeze-drying and hot-pressing

cycle, Ei/E1. For the unfilled matrix, a continuous increase of Ei is observed for successive cycles. It increases from 1.4 MPa for the first cycle up to 2.2 MPa for the fifth one (Table 2). The relative tensile modulus increases from 1 to 1.57 during successive tests. This phenomenon is ascribed to the well-known strain-induced crystallization of rubbers. Recently a group of scientists put forward a new idea about strain-induced crystallization of natural rubber during uniaxial deformation.32 Contrary to the conventional concept, their studies revealed that the applied stress induces a network of microfibrillar crystals in NR and is responsible for the improved elastic properties. For a given cycle, an increase of the modulus with the chitin whiskers content is observed for all compositions (Table 2). This reinforcing effect agrees with the DMA and tensile tests experimental data. As for the unfilled matrix, a continuous increase in the relative modulus is observed for all the compositions (Figure 9a). It implies that the behavior of the composites during the successive tensile tests is mainly dominated by the matrix behavior. It could be due to the absence of the formation of a chitin network within the hot-pressed NR matrix. The successive moduli measured for unvulcanized NR based composite films obtained by the evaporation technique are reported as absolute values in Table 2 and as relative values in Figure 9b. For the unfilled matrix, a continuous increase of Ei is observed for successive cycles. It increases from 1.9 MPa for the first cycle up to 5.1 MPa for the seventh one (Table 2). The relative tensile modulus increases from 1 to 2.68 during successive tests. The sharper increase of the modulus during successive cycles for the evaporated matrix compared to the hot-pressed one is an indication of the higher strain-induced crystallization of the former. Similarly as in the case of hot-pressed samples, for a given cycle, the modulus of the composites increases with the chitin whiskers content. In agreement with the previous DMA and tensile tests experimental data, the reinforcing effect is much higher for the composite materials obtained via the water evaporation process compared to that for the materials obtained by freeze-drying and hot-pressing techniques. However, it is worth noting that for the evaporated composites the modulus Ei first decreases for the first three or four cycles and then increases. The initial decrease of Ei for

88.7 2.4 2.7 4.2 8.2 17.5 23.9

2.5 2.7 4.2 8.2 18.4

E8 (MPa)

2.5 2.7 4.2 8.2 18.6

5.0 7.3 27.2

composite materials can be ascribed to the progressive damaging of the chitin whisker network. The nonlinear decreasing tendency of modulus values of filled rubber with increasing strain amplitude (Payne effect) has been studied earlier.33,34 The payne effect increases with increasing fillerfiller network factor and decreases with increasing polymerfiller network.33,34 After the complete destruction of the chitin network, the tensile modulus starts to increase slowly as a result of the strain-induced crystallization already observed for the unfilled matrix. The continuous increase of Ei for successive cycles reported for hot-pressed nanocomposites is a clear indication of the absence of any chitin network within these materials. This continuous whiskers network is expected to govern the mechanical behavior of the evaporated composites. The same experiment was also performed for vulcanized NR based composites and results are reported in Table 2 and Figure 9c. It is observed that the film containing low percentages of whiskers (0 and 2 wt % filled material) exhibit a continuous increase of Ei for successive cycles. For instance, for the unfilled matrix, it increases from 1.6 MPa for the first cycle up to 2.5 MPa for the seventh one (Table 2). The relative tensile modulus increases from 1 to 1.59 during successive tests. This increase is much lower than the one observed for the unvulcanized evaporated matrix (Figure 9b). This can be well understood since for chemically cross-linked polymers the strain-induced crystallization is more restricted. In fact, the network of bridges in vulcanized rubber might have two opposite effects on strain-induced crystallization during stretching.32 The bridges can improve molecular orientation of chains in the vicinity of cross-links, thus inducing a crystallization of polymer chains. Conversely these bridges may also hinder the growth of the crystalline structure. The lower increase in modulus of the vulcanized matrix than that of the unvulcanized one is most probably due to the above-mentioned opposite effects which take place in the former. For composite materials, a clear decrease in the relative modulus is observed during the successive tensile tests up to the first four cycles and after that it remains almost unchanged. The first decrease in modulus is due to the disruption of the chitin whisker network that is expected to

674

Biomacromolecules, Vol. 4, No. 3, 2003

Gopalan Nair and Dufresne

be formed during the slow evaporation step of the composites. However, contrary to the increase in modulus of the unvulcanized evaporated composites, after the fourth cycle, it is found that the modulus of the vulcanized one remains unchanged after the fourth cycle. It is due to the limited extent of the strain-induced crystallization of cross-linked rubber as a result of the two opposite effects operating during stretching of the material as mentioned above.

result of hydrogen bonding within the evaporated samples. A clear decrease in modulus values of the evaporated samples up to first three or four cycles of the cyclic test indicates the progressive damaging of the chitin network, which was initially present in it. All the results reveal that the main aspect that governs the mechanical behavior of the chitin whisker reinforced NR nanocomposites is the processing technique.

Conclusions

Acknowledgment. K. Gopalan Nair gratefully acknowledges the French Ministry of Research for its financial support.

Linear and nonlinear mechanical behavior of chitin whiskers obtained from crab shell reinforced unvulcanized and vulcanized natural rubber (NR) nanocomposite materials were analyzed. For unvulcanized materials two processing methods were used, namely, (1) casting and evaporation and (2) freeze-drying and subsequent hot pressing whereas for vulcanized materials only the former method was used in this study. Both the effect of vulcanization on the mechanical properties of composite films prepared by evaporation method and the effect of processing technique on the properties of unvulcanized composite film were studied. Dynamic mechanical analysis revealed the presence of a small percentage of crystallinity in unvulcanized NR prepared by the evaporation method whereas no such evidence of crystallinity was detected either in vulcanized rubber prepared by evaporation or in unvulcanized natural rubber prepared by hot pressing methods. DMA also showed that the rubbery modulus of unvulcanized evaporated natural rubber is significantly improved by the incorporation of chitin whiskers. In addition to the excellent improvement in mechanical properties, for filler contents higher than 5 wt %, an improvement in the thermal stability of the composite is also noticed up to 220-230 °C. No significant effect of the crosslinking of the matrix is reported in DMA experiments, except an enhancement of the thermal stability of the composites up to 250-300 °C. For hot-pressed samples, the reinforcing effect is significantly lower. These observations agree with our previous swelling experiments31 and support the formation of a three-dimensional chitin whiskers network, governed by a percolation mechanism, in evaporated nanocomposites. The evaporation method is a slow step process, in which the whiskers get enough time and mobility to establish a rigid chitin-chitin network within the host matrix. In contrast to the results of DMA experiments, tensile test experiments produce a clear hierarchy in their results due to the higher stress level applied during the tensile testing. Tensile and conventional modulus values are higher for unvulcanized evaporated samples than for vulcanized evaporated ones which in turn are higher than that for unvulcanized hot pressed samples. Interference of chemical cross-linking reaction on the formation of chitin network during vulcanization is supposed to produce a lower reinforcing effect and thereby a lower tensile modulus in vulcanized samples. It is an indication that the weak stress level involved in DMA experiments is not sufficient to determine the extent of chitinchitin interactions. Successive tensile test experiments help to understand the existence of a three-dimensional chitin network formed as a

References and Notes (1) Blow, C. M.; Hepburn, C. In Rubber Technology and Manufacture, 2nd ed.; Butterworth-Heinemann: Oxford, 1982. (2) Liauw, C. M.; Allen, N. S.; Edge, M.; Lucchese, L. Polym. Degrad. Stab. 2001, 74, 159. (3) De, S. K.; White, J. R. In Short fibre-polymer composites; Woodhead Publishing: Cambridge, 1996. (4) Bounstany, K.; Hamed, P. Rubber World 1974, 171, 39. (5) Beatty, J. R.; Hamed, P. Elastomerics 1978, 110, 27. (6) Rogers J. W. Rubber World 1981, 183, 27. (7) Ismail, H.; Shuhelmy, S.; Edyham, M. R. Eur. Polym. J. 2002, 38, 39. (8) Ismail, H.; Edyham, M. R.; Wirjosentono, B. Polym. Test. 2002, 21, 139. (9) Geethamma, V. G.; Thomas Mathew, K.; Lakshminarayanan, R.; Thomas, S. Polymer 1998, 39, 1483. (10) Ismail, H.; Rozman, H. D.; Jaffri, R. M.; Mohd Ishak, Z. A. Eur. Polym. J. 1997, 33, 1627. (11) Ismail, H.; Jaffri, R. M.; Rozman, H. D. Polym. Int. 2000, 49, 618. (12) Ismail, H.; Jaffri, R. M.; Polym. Test. 1999, 18, 381. (13) Favier, V.; Canova, G. R.; Cavaille´, J. Y.; Chanzy, H.; Dufresne, A.; Gauthier, C. Polym. AdV. Technol. 1995, 6, 351. (14) Favier, V.; Cavaille´, J. Y.; Chanzy, H. Macromolecules 1995, 28, 18, 6365. (15) Helbert, W.; Cavaille´, J. Y.; Dufresne, A. Polym. Compos. 1996, 17, 604. (16) Dufresne, A.; Cavaille´, J. Y.; Helbert, W. Polym. Compos. 1997, 18, 198. (17) Favier, V.; Canova, G. R.; Shrivastava, S. C.; Cavaille´, J. Y. Polym. Eng. Sci. 1997, 37, 1732. (18) Chazeau, L.; Paillet, M.; cavaille´, J. Y. J. Polym. Sci., Polym. Phys. 1999, 37, 2151. (19) Dubief, D.; Samain, E.; Dufresne, A. Macromolecules 1999, 32, 5765. (20) Dufresne, A.; Kellerhals, M. B.; Witholt, B. Macromolecules 1999, 32, 7396. (21) Dufresne, A. Compos. Interfaces 2000, 7, 53. (22) Angle`s, M. N.; Dufresne, A. Macromolecules 2000, 33, 8344. (23) Angle`s, M. N.; Dufresne, A. Macromolecules 2001, 34, 2921. (24) Mathew, A. P.; Dufresne, A. Biomacromolecules 2002, 3, 609. (25) Marchessault, R. H.; Morehead, F. F.; Walter, N. M. Nature 1959, 184, 632. (26) Revol, J. F.; Marchessault, R. H. Int. J. Biol. Macromol. 1993, 15, 329. (27) Li, J.; Revol, J. F.; Naranjo, E.; Marchessault, R. H. Int. J. Biol. Macromol. 1996, 18, 177. (28) Revol, J. F.; Marchessault, R. H. J. Colloid Interface Sci. 1996, 183, 365. (29) Paillet, M.; Dufresne, A. Macromolecules 2001, 34, 6527. (30) Morin, A.; Dufresne, A. Macromolecules 2002, 35, 2190. (31) Gopalan Nair, K.; Dufresne, A. Biomacromolecules 2003, 4, 657. (32) Toki, S.; Sics, I.; Ran, S.; Liu, L.; Hsiao, B. S.; Murakami, S.; Senoo, K.; Kohjiya, S. Macromolecules 2002, 35, 6578. (33) Payne, A. R. J. Polym. Sci. 1962, 8, 57. (34) Payne, A. R.; Whittaker, W Rubber Chem. Technol. 1971, 44, 140.

BM0201284