Selective Localization of Multiwalled Carbon Nanotubes in Poly(Îµ

Biomacromolecules 2009, 10, 417–424

417

Selective Localization of Multiwalled Carbon Nanotubes in Poly(ε-caprolactone)/Polylactide Blend Defeng Wu,*,†,‡ Yisheng Zhang,†,‡ Ming Zhang,†,‡ and Wei Yu§ School of Chemistry and Chemical Engineering, Yangzhou University, Jiangsu 225002, People’s Republic of China, Provincial Key Laboratories of Environmental Material and Engineering, Jiangsu 225002, People’s Republic of China, and Department of Polymer Science and Engineering, Shanghai Jiaotong University, Shanghai 200240, People’s Republic of China Received October 20, 2008; Revised Manuscript Received December 11, 2008

Poly(ε-caprolactone)/polylactide blend (PCL/PLA) is an interesting biomaterial because PCL and PLA present good complementarity in their physical properties and biodegradability. However, the thermodynamic incompatibility between two component polymers restricts further applications of their blend. In this work, we used functionalized multiwalled carbon nanotube (MWCNT) to control the morphology of immiscible PCL/PLA blend. The ternary PCL/PLA/MWCNTs composites were hence prepared by melt mixing for the morphology and the properties investigation. It is interesting to find that the functionalized MWCNTs are selectively dispersed in the matrix PCL phase and on the interface between two polymer phases, leading to simultaneous occurrence of thermodynamically and kinetically driven compatibility. Those interface-localized MWCNTs prevent coalescence of the discrete domains and enhance the phase interfacial adhesion as well. As a result, the phase morphology of the ternary composites is improved remarkably in contrast to that of the blank PCL/PLA blend. Owing to that unique selective interface-localization and improved phase morphology, the ternary composites present far lower rheological and conductive percolation thresholds than those of the binary composites, and also present extraordinary mechanical properties even at very low loading levels of the MWCNTs. Therefore, the amphiphilic MWCNTs are believed to act as the reinforcements as well as the compatibilizer in the immiscible PCL/PLA blend.

1. Introduction Recently, environmental concerns and a shortage of petroleum resources have driven efforts aimed at bulk production of biodegradable polymers such as aliphatic polyesters, polysaccharides, unsaturated polyester, polyvinylalcohols, and modified polyolefins.1 Among them, poly(ε-caprolactone) (PCL) has received much attention as a new aliphatic polyester being developed for a wide range of applications due to its thermoplastic, biodegradable, and biocompatible properties.2,3 Increasing realization of the intrinsic properties of PCL, coupled with knowledge of how such properties can be improved to achieve the compatibility with processing, manufacturing, and end-use requirements of thermoplastics, has hence fueled technological and commercial interest. Several approaches such as copolymerization4,5 and mixing with the nanofiller6 have been used to improve and/or control the properties of PCL. Another convenient strategy is to blend PCL with many other polymers to obtain new biocompatible materials with high performance.7 Among those blends, poly(εcaprolactone)/polylactide blend material (PCL-blend-PLA) is very interesting due to the large difference in the physical properties and biodegradability between two polymers, where the glassy PLA with high degradation rate presents better tensile strength, while the rubbery PCL with much slower degradation rate shows better toughness. This property complementarity is quite important for the PCL-blend-PLA material because one can control or even design the performance by adjusting the * To whom correspondence should be addressed. Tel.: +86-51487975590, ext. 9115. Fax: +86-514-87975244. E-mail: [email protected]. † Yangzhou University. ‡ Provincial Key Laboratories of Environmental Material and Engineering. § Shanghai Jiaotong University.

molecular characteristics of the polymers and blending ratio as well as processing conditions to meet the requirements of various applications. Therefore, much research work8-21 has hitherto been reported on the preparation and the applications of this kind of blend material. The performance such as enzymatic and nonenzymatic degradation, thermal and mechanical properties as well as drug release behavior has been studied extensively, aiming at relating those properties to the compositions and the immiscible phase structures.8-13 The applications of such blend system as a new type of microporous substrate or temporary scaffold for tissue engineering and drug delivery have also been explored.14-16 It is well-known that the performance of the polymer blends depends not only on the properties of the matrix polymers, but also highly on the phase morphology.22,23 However, PCL and PLA are incompatible thermodynamically with each other and can only form multiphase structure in their blend system with poor interfacial adhesion, which restricts its further applications. It is well accepted that the phase structure of the immiscible polymer blends is mainly governed by the composition and rheological properties of the component polymers, which is hence assumed to be a unique function of the flow history and the properties of the polymers. Thus, the relations between morphology and rheological properties within the PCL/PLA blend have been studied,17,18 aiming at exploring appropriate processing conditions to improve phase structure kinetically. Much more research work, however, is concentrated on using amphiphilic diblock or multiblock copolymers as compatibilizer to improve the miscibility the PCL/PLA blend.19-21 It has been reported that the third component, namely, those well-defined copolymers whose chemical nature is identical to that of the main components, can act as emulsifying agents, reducing the

10.1021/bm801183f CCC: $40.75  2009 American Chemical Society Published on Web 01/13/2009

418

Biomacromolecules, Vol. 10, No. 2, 2009

coalescence effect by lowering the interfacial tension between the two components and finally leading to a well-dispersed morphology. In recent years, many researchers24-33 have found that the anisotropic nanofiller, such as organoclay, can also be used as a third component to improve phase morphology of the immiscible polymer blends. It has been reported that addition of the nanofiller may affect the interfacial properties in an immiscible blend system with results similar to that of addition of a compatibilizer. Two possible mechanisms33 have been proposed on this morphological improvement. Mechanism I (thermodynamic compatibility): The organic modifier of the nanofiller is miscible or at least compatible with both phases, thus the overall free-energy of mixing (∆Gmix) becomes negative and thermodynamically driven compatibility is likely to occur between the immiscible components. Mechanism II (dynamic compatibility): The polymer pairs present large difference in their polarity or rheological properties. In this case, the nanofiller is mainly dispersed in the component with stronger polarity or with lower viscosity. This selective localization not only changes the viscosity ratio of two components but also prevents coalescence of the domains during melt mixing, improving compatibility between two phases kinetically. Therefore, the addition of the anisotropic nanofiller may bring reinforcing and compatible effects into an immiscible blend system simultaneously, which is a new approach to obtain polymer blend nanocomposites with high performance. As a new anisotropic one-dimensional nanomaterial, the carbon nanotube (CNT) has become the next-generation reinforcements for nanostructured polymeric composite materials, increasingly owing to its extraordinarily high elastic modulus, strength, and resilience.34,35 Many conducting polymer/CNT composites have been prepared successfully in recent years via the approaches of melt mixing, film casting and/or polymerization.36-46 As expected, small addition of CNTs can improve mechanical properties and electrical conductivity of the matrix polymers remarkably. Moreover, in vivo studies44-46 have confirmed good biocompatibilities of CNTs with various cells, indicating that the CNTs can be used as new type of nanomodifier to prepare biopolymeric composite materials. In this work, therefore, we used CNTs as both the nanoreinforcements and the compatibilizer to modify the immiscible PCL/PLA blend. The weight ratio of PCL and PLA is 70/30, in which the PCL component is continuous phase.18 The distribution of the CNTs and the immiscible morphology of the blend matrix were systematically investigated, aiming at relating the morphological improvement to the selective localization of the CNTs. The physical and mechanical properties were then studied to further confirm the micro- and mescoscopic structure-property relations proposed in the PCL/PLA/CNT ternary nanocomposites.

2. Experimental Section 2.1. Materials Preparation. The poly(ε-caprolactone) (CAPA6500) used in this study is a commercial product of Solvay Co. Ltd., Belgium. Its Melt Index (MI) is about 7 g/10 min (160 °C/2.16 Kg, ASTM D1238), and the -OH value is lower than 2 mg KOH/g. The polylactide (2002D) is also a commercial product of NatureWorks Co. Ltd., U.S.A. Its residual monomer content is less than 0.3 wt %, and MI is about 8 g/10 min (190 °C/2.16 Kg, ASTM D1238). The multiwalled carbon nanotubes (MWCNT) were supplied by Chengdu Organic Chemistry Co. Ltd., Chinese Academy of Sciences. The purified MWCNT (M1203, purity > 95%) is a chemical vapor deposition material with outside diameter of 10-20 nm, inside diameter of 5-10 nm and length

Wu et al. of 10-30 µm. Its special surface area is higher than 200 m2/g. The carboxylic MWCNT (MS1223, purity > 95%), which is functionalized on M1203 by -COOH, has the identical dimension parameter and special surface area with those of M1203. The rate of surface carbon atom on MS1223 is about 8-10 mol % and the -COOH weight percent is about 1-6 wt % (measured by XPS). The PCL/PLA/MWCNT ternary composites (PCL/PLAs, where s denotes the weight of the MWCNT per hundreds weight of the blend resin (phr)) were prepared by direct melt compounding various loadings of carboxylic MWCNT with the blend matrix (the two matrix components of PCL and PLA are in the proportion of 70/30 (w/w)) in a HAAKE polylab rheometer (Thermo Electron Co., U.S.A.) at 170 °C and 50 rpm for 6 min. All materials were dried at predetermined temperature under vacuum before using. The sheet specimens in thicknesses of about 1 mm for the rheological, morphological, and electrical conductivity measurements were prepared by compression molding at 180 °C and 10 MPa. The dog-bone shaped specimens for the tensile and the dynamic mechanical property characterizations were prepared by injection molding at 180 °C and 13 MPa using a RR/ PSMP2 test sample injection molding apparatus (Ray-ran Co., England). The binary PCL nanocomposites with 1 wt % of carboxylic MWCNT (PCL1) and the ternary composites with 1 wt % of purified MWCNT (PCL/PLA1′) were also prepared under the same processing conditions for the properties comparison. 2.2. Characterization. Microstructure and Morphology. The dispersion of MWCNT was observed using a Tecnai 12 transmission electron microscope (TEM) (PHILIPS Co., Netherlands) with 120 kV accelerating voltage. The transmission electron micrographs were taken from microtomed sections in thickness of 80-100 nm. The morphologies of the fractured surfaces of the tensile samples and the brittlely fractured samples were observed using a XL-30ESEM scanning electron microscope (SEM) (PHILIPS Co., Netherlands) with 20 kV accelerating voltage and a S-4800 field-emission scanning electron microscope (FESEM; Hitachi Co., Japan) with 15 kV accelerating voltage. The numberaverage diameter of the domains for sea-island morphology (a typical phase separation structure, in which one polymer phase is dispersed in another polymer phase as discrete domains) was determined according to the following relation

Dn )

∑ND ⁄∑N i

i

i

(1)

where Ni is the number of dispersed domains with a diameter of Di counted from the SEM images. The total number of particles is about 100 in the analysis. Mechanical Properties. The tensile properties of the blank blend and the nanocomposite samples were determined by an Instron Mechanical Tester (ASTM D638) at a crosshead speed of 50 mm/min at room temperature using the dog-bone shaped specimens. Property values reported here represent an average of the results for tests run on six specimens. The dynamic mechanical properties of the samples were characterized using a DMA-242C dynamic thermal mechanical analyzer (NETZSCH Co., U.S.A.). The testing was performed in threepoint bending mode at the vibration frequency of 5 Hz in a N2 atmosphere. The heating rate is predetermined as 5 °C/min and the temperature ranges from -100 to 100 °C. Rheological Properties. Rheological measurements were carried out on a rheometer (HAAKE RS600, Thermo Electron Co., U.S.A.) using a parallel plate geometry with 20 mm diameter plates. The samples about 1.0 mm in thickness were melted in the parallel plate fixture at 180 °C for 5 min to eliminate residual thermal history, and then carry out the dynamic shear measurements immediately. The dynamic strain sweep was first carried out to determine a common linear region, strain level of 1%. Then, the small amplitude oscillatory shear (SAOS) was applied and the dynamic frequency sweep was carried out. Electrical Properties. The volume resistance of sheet samples was measured by a four-point probe apparatus, ZC36 high resistance meter

Localization of Multiwalled Carbon Nanotubes


419

Figure 1. SEM images of the brittlely fractured (a) blank PCL/PLA sample, (b-e) PCL/PLA samples with various carboxylic MWCNT loadings of 0.2, 0.5, 1, and 2 wt %, and (f) PCL/PLA1′ sample with purified MWCNT loadings of 1 wt %. The PCL and PLA are in the weight ratio of 70/30.

(Shanghai Precision and Scientific Co. Ltd., P. R. China) at room temperature. Three specimens of each sample were tested taking four data points.

3. Results and Discussion 3.1. Micro- and Mescoscopic Structure of the PCL/PLA/ MWCNTs Ternary Nanocomposites. Figure 1a-e gives the SEM images of the fracture surface, respectively, for the blank PCL/PLA blend and its ternary composite samples with various carboxylic MWCNT loadings. Clearly, all samples show typical sea-island morphologies, where the discrete PLA spherical domains are dispersed in the PCL matrix. With addition of MWCNTs, the size of the PLA domains reduces remarkably. The average diameter of the domains decreases from 21.5 to 6.3 µm as the MWCNT loadings achieve up to 1 phr. Moreover, the blend matrix presents better interfacial adhesion between two phases in the presence of the carboxylic MWCNTs. This indicates that addition of the carboxylic MWCNTs improves

interfacial properties of the immiscible PCL/PLA blend with similar results of adding a compatibilizer. However, it is seen that the size of the discrete PLA domains also reduces evidently with addition of the MWCNTs without functionalization, whereas the interfacial adhesion between the two matrix phases is poor in this case. (Figure 1f). Thus, it is interesting to explore the mechanism of the morphological change with addition of the MWCNTs. Figure 2a-c gives the TEM images of the PCL/PLA1 sample. The typical two-phase structure can be seen in Figure 2a, in which the light and dark gray parts correspond to discrete PLA and matrix PCL phases, respectively. On the amplified images of the parts marked in Figure 2a, it is seen clearly that the carboxylic MWCNTs are mainly dispersed in the PCL phase and on the phase interface (Figure 2b,c). The selective localization of the CNTs in one polymer phase has also been observed on some ternary nanosystems such as PA6/PPS/CNTs47 and PC/ PP/CNTs composites48 in which the two matrix polymers show

420


Wu et al.

Figure 2. TEM images for the samples of (a) PCL/PLA1 and (d) PCL/PLA1′. The parts indicated in (a) are amplified and shown in (b) and (c).

Figure 3. FE-SEM images of (a) tensile failure and (b) brittle fracture surface of the PCL/PLA1 sample.

a large difference in their polarity and in their affinity to the CNTs. Thus, the CNTs are mainly located in one polymer phase with stronger polarity. For the ternary systems in this work, however, there is no remarkable difference in the affinity of two polymers to the carboxylic MWCNTs because both the PCL/MWCNTs42 and the PLA/MWCNTs binary composites43 present a good dispersion state of the MWCNTs. Actually, the PLA shows more or less better affinity to the MWCNTs than that of the PCL. The selective localization of MWCNTs, therefore, is attributed to a large difference in the rheological properties of two matrix polymers in which the PLA shows far higher viscosity than that of the PCL (ηPLA/ηPCL ≈ 16). In this case, the PCL chain can diffuse around and into the MWCNT aggregates more easily compared with that of the PLA at the

initial stage of melt mixing and, as a result, the MWCNTs prefer to be detached and dispersed in the lower viscous PCL continuous phase rather than the discrete PLA phase. With rapid reduction in the viscosity ratio between two matrix polymers, the higher viscous PLA phase can be broken into smaller droplets more easily. On the other hand, many MWCNTs have a tendency to be further dispersed on the phase interface driven by the mixing flow because the carboxylic group on their surface has nice affinity to both the PCL and the PLA phases. Finally, they are distributed in the phase interface layer and ranged more or less ordered along the surface of the PLA droplets, acting as the emulsifier49 to enwrap the discrete domains well, as can be seen in Figure 2c.



421

Figure 6. Schematic diagrams of the percolated MWCNTs network in (a) homogeneous matrix and (b) immiscible blend matrix. The white part is matrix PCL phase, the gray part is discrete PLA phase and the short black curve is MWCNTs. Figure 4. DTMA thermogram for the PCL/PLAs samples with various carboxylic MWCNT loadings.

Figure 7. Plots of electrical conductivity vs weight fraction of the carboxylic MWCNTs.

Figure 5. Plots of (a) dynamic storage modulus (G′) and (b) loss modulus (G′′) vs frequency and (c) ν-GP plots of phase angle (δ) vs complex modulus (|G*|) obtained from dynamic frequency sweep on the PCL/PLAs samples with various carboxylic MWCNT loadings.

The mechanism proposed above on that selective localization is further confirmed by the results of the TEM measurements

on the PCL/PLA1′ sample that has identical MWCNT loadings with that of the PCL/PLA1 sample, while the filled MWCNTs in this ternary system are only purified without further functionalized. As can be seen in Figure 2d, the purified MWCNTs also present selective localization in the PCL phase, which confirms that the MWCNTs, as the filled component, show a general tendency to be dispersed in the less viscous phase during melt mixing. However, most of MWCNTs are still presented as small aggregates or bundles in the PCL/PLA1′ sample, showing poor dispersion in contrast to that of the PCL/PLA1 sample. This is due to the poor affinity between the matrix polymers and those MWCNTs without functionalization. The migration of the MWCNTs in this case is hard to occur and thus the selective interface distribution is nearly not observable in the PCL/PLA1′ sample. As a result, the immiscible blend matrix of the PCL/PLA1′ sample presents rather weak phase interfacial adhesion compared with that of the PCL/PLA1 sample, although both samples have comparable size of discrete PLA phase, as shown in Figure 1d and f. This confirms that besides the rheological properties of the blend components, the surface functionalization is also vital to the selective interface localization of the MWCNTs. On the one hand, the surface carboxylic group of the MWCNTs is compatible with both the PCL and the PLA, which reduces the overall free energy of mixing and increases the thermodynamical compatibility. On the other hand, interfacial localization of the MWCNTs could prevent the coalescence of the PLA domains effectively, which helps compatibilization during melt mixing. Therefore, both the thermodynamically and kinetically driven compatibility can occur.25,26 Such interface localization of the carboxylic MWCNTs, as a result, improves the interfacial adhesion of the PCL/PLA blend matrix evidently, which is further confirmed by the FESEM images shown in Figure 3.

422


Figure 8. (a) Stress/strain curves for the neat PCL, blank PCL/PLA, ternary PCL/PLAs, and PCL/PLA1′ samples; (b) plots of tensile yield strength vs weight fraction of the carboxylic MWCNTs.

To further evaluate the compatible effect of the carboxylic MWCNTs on the PCL/PLA blend system, the dynamic thermal mechanical analysis was performed. Figure 4 gives the DTMA thermogram for the neat matrix polymers and PCL/PLAs samples. It is seen that with the increase of MWCNT loadings, the glass transition temperatures (Tg) of the two component polymers shift to each other gradually. This indicates that emulsification occurs at the phase interface in the presence of the amphiphilic carboxylic MWCNTs, leading to interface stabilization thermodynamically. However, Both the PCL and the PLA still present evident glass transition behavior, respectively, in the ternary system and the shift degree of the two Tg is not remarkable, suggesting that thermodynamically driven compatibility by the carboxylic MWCNTs is not as strong as that by the well-defined copolymers.19,20 This is mainly due to large difference of chemical structure between the MWCNTs and the copolymers. In general, the copolymers used as compatibilizer present the blocks structure similar to that of the parent homopolymers and, the length of the blocks should exceed or at lest close to that of the homopolymers, while the MWCNTs only have amphiphilic carboxylic group, which may be too small to make mother-MWCNTs fully reptate at the phase interface.50 In addition, the presence of MWCNTs restricts motion of the matrix PCL chain, also counteracting the shift of Tg somewhat. However, the overall compatible effect of the carboxylic MWCNTs is still good, indicating that kinetically driven compatibility is dominant in the melt mixing process. In other words, the improvement of the phase morphology is mainly due to the obstructing effect of those interfacial-located MWCNTs on the coalescence of those discrete domains. 3.2. Percolation Behavior of the PCL/PLA/MWCNTs Ternary Nanocomposites. Because the selective interface localization of the MWCNTs improves the phase morphology

Wu et al.

evidently, it may influence the final performance of the ternary nanocomposites. The rheological behavior was first characterized and the dynamic storage modulus (G′) and loss modulus (G′′) obtained from the dynamic frequency sweep are shown in Figure 5a,b. The shoulder presented on the modulus curve of the blank PCL/PLA sample (see the arrow in Figure 5a) is attributed to the shape relaxation of the discrete PLA phase in the PCL matrix.51 During oscillatory shear flow, the total area of the interface as well as the interfacial energy are changing periodically but with a much longer relaxation time than that of component polymers.52 This kind of long relaxation due to the presence of interface leads to an additional transition shoulder on the modulus curves. It is seen that the relaxation shoulder shifts to low-frequency region with small addition of the MWCNTs (0.2 phr). This indicates that the presence of MWCNTs retards shape relaxation of the discrete PLA domains. Because the interfacial relaxation time is proportional to the ratio between droplet size and the interfacial tension, longer interfacial relaxation process but with decreasing droplet size means a large decrease of interfacial tension, which is also an indication of the selective localization of MWCNTs on the interface. It is notable that further addition of the MWCNTs enhances the low-frequency modulus sharply. As the MWCNT loadings achieving up to 0.5 phr, the low-frequency G′ increases even by about 4 orders as compared with that of the blank PCL/ PLA sample and the frequency dependence nearly disappears. This nonterminal behavior is due to formation of the percolated MWCNTs network,36-43 which highly restrains the long-range relaxation of the matrix PCL chains. van Gurp-Palmen plot53 is usually used to detect the rheological percolation of the filled polymeric composites, also including polymer/CNTs systems.37,38,41 Figure 5(c) gives the v-GP plots of phase angle (δ) versus complex modulus (|G*|) for the PCL/PLAs ternary systems. The low-frequency δ of the blank PCL/PLA and the PCL/PLA0.2 samples are close to 80°, which is indicative of a flow behavior presented by the viscoelastic fluid. As the MWCNT loadings increase to 0.5 phr, the low-frequency δ decreases remarkably to lower than 45°, indicating a rheological fluid-solid transition in that ternary system. Accordingly, the rheological percolation threshold for the PCL/PLAs ternary systems is lower than 0.5 phr. It is interesting that the threshold of the ternary systems is far lower than those reported on the PCL/MWCNTs and PLA/MWCNTs binary systems prepared also by melt mixing, in which present the critical values of about 2-3 wt %.42,43 This is due to the selective localization of the MWCNTs consequentially. As discussed in the section of morphology, the interface-localized MWCNTs tend to be arranged more or less ordered along the surface of the PLA droplets during melt mixing. In contrast to the disorderly dispersion in the binary systems, the self-assembly like behavior of the MWCNTs can enhance the particle-particle interactions more effectively in the ternary systems, promoting formation of the percolated MWCNTs network structure even at lower volume concentration, as can be seen in the schematic illustration shown in Figure 6. As a result, the PCL/PLAs ternary systems present very low percolation threshold. It is well-known that nonlinear electrical properties of the polymeric composites filled with conductive filler, including polymer/CNTs systems,36,37,39,40 also shows percolation behavior.54 Owing to the selective localization of the MWCNTs, it can be expected that the ternary systems may show lower conductive percolation threshold than that of the binary ones. Figure 7 gives the electrical conductivity as a function of MWCNT loadings for the ternary composites. The curve clearly



423

Figure 9. SEM images of the tensile samples of (a) blank PCL/PLA, (b) PCL/PLA0.2, (c) PCL/PLA1, and (d) PCL/PLA1′ systems.

presents three regions (dielectric, transition, and conductive), which is indicative of a typical percolation behavior. At identical MWCNT loadings of 1 phr, the conductivity of the ternary composites is far higher than that of the binary ones by about 3-4 orders. This indicates that those well-ranged MWCNTs improve structure of the conductive network. On the one hand, selective interface-localization inosculates the MWCNTs well with one another, allowing direct electron transport more smoothly. On the other hand, the increase of the phase interface due to the decrease of the discrete phase size makes the conductive network form more easily, even at relative lower loading levels of the MWCNTs. Thus, the ternary systems present higher conductivity and lower conductive percolation threshold than those of the binary ones. 3.3. Mechanical Properties of the PCL/PLA/MWCNTs Ternary Nanocomposites. The results above show that the performance of the PCL/PLAs ternary composites highly depends on the phase morphology of the blend matrix and the dispersion state of the MWCNTs. The enhancement in the mechanical properties, therefore, is foreseeable on those ternary composites owing to the unique localization of the MWCNTs and the improvement of phase structure as compared with that of the blank PCL/PLA blend. Figure 8 gives the strength and stress/strain curves for all samples. The tensile yield strength is used here for better comparison because the PCL and its blends are generally ductile.55,56 Although the PLA presents higher strength than that of the PCL,2 their blend shows the strength even lower than that of the neat PCL (Figure 8a). This is due to the poor interfacial adhesion between matrix PCL and minor PLA phase.11 Small addition of the MWCNTs, as expected, enhances the blend materials remarkably, and the strength of the ternary composites increases monotonically with increase of the MWCNTs loadings within experimental loading ranges, as can be seen in Figure 8b.

Figure 9 gives the SEM images of those tensile samples. The blank PCL/PLA sample clearly shows interface break (Figure 9a), while those ternary samples present both the interface and bulk break (see the parts marked in Figure 9c). Addition of the amphiphilic carboxylic MWCNTs improves the phase adhesion due to the selective interface localization, resulting in good load transfer between two matrix phases. With increasing loading levels, those well-dispersed MWCNTs selectively in the matrix PCL phase begin to play an important role in bearing an extra load. As a result, the tensile yield strength of the ternary composites further increases, accompanied by an evident reduction of the elongation at break because the presence of MWCNTs impedes movement of the matrix molecule chain. It is notable that the PCL/PLA1′ sample (Figure 9d) shows poor phase adhesion between PCL and PLA in contrast to that of PCL/PLA1 sample (Figure 9b), although both present similar sea-island morphologies and almost identical radius of discrete domains. On the one hand, those MWCNTs without functionalization can not be dispersed well in the matrix polymer and merely act as a stress concentration point, leading to a decrease of the strength of bulk PCL. The interfacial adhesion between two polymers is very weak if the MWCNTs are not functionalized and not selectively distributed on the interface. Thus, the PCL/PLA1′ sample presents a characteristic of brittle fracture (most of domains are pulled out from fractured surface in the tensile process). As a result, the breaking strength of the PCL/ PLA1′ sample is far lower than the yield strength of the PCL/ PLA1 sample, even lower than the blank PCL/PLA blend (Figure 8a). This also confirms that the surface functionalization is a key point to the selective localization of the MWCNTs and, only in the case of interface localization, addition of the amphiphilic MWCNTs can bring reinforcing and compatible effects together into an immiscible blend system.

424


4. Conclusions PCL/PLA/MWCNTs ternary composites with various MWCNT loadings were prepared by direct melt mixing. The weight ratio of PCL and PLA is fixed in the proportion of 70/30 for all samples in which PCL is the matrix phase. With addition of the MWCNTs, the size of the discrete PLA domains reduces remarkably due to the selective localization of the MWCNTs in the blend matrix. The ternary systems containing the carboxylic MWCNTs show the selective localization of MWCNTs both in the matrix PCL phase and on the phase interface, while the ternary systems containing the MWCNTs without functionalization only show the selective localization in the matrix PCL phase. This is attributed to a large difference in the affinity of two MWCNTs to the polymers. Only in the case of interface localization, the addition of the MWCNTs can bring reinforcing and compatible effects together into the immiscible PCL/PLA blend. The ternary systems containing carboxylic MWCNTs hence present high improvement of the performance in terms of rheological, conductive, and mechanical properties as compared with those of PCL/PLA blend and binary composites. Acknowledgment. This work was supported by the research grants from the National Natural Science Foundation of China (No. 50803052), the Startup Program of Innovative Talent of Jiangsu Province (No. BK2007559), and the Key Program of Jiangsu Province (No. 06KJA15011).

References and Notes (1) Vert, M.; Feijen, J.; Albertsson, A. C.; Scott, G.; Chiellini, E. Biodegradable Polymer and Plastics; Royal Society of Chemistry: London, 1992. (2) Smith, R. Biodegradable Polymers for Industrial Applications; CRC Press: Boca Raton, FL, 2005. (3) Pitt, C. G.; Schindler, T. A. Biodegradable Drug DeliVery Systems Based on Aliphatic Polymers: Application of ContraceptiVe and Narcotic Antagonists in Controlled Release of BioactiVe Materials; Academic Press: New York, 1980. (4) Xie, W. H.; Zhu, W. P.; Shen, Z. Q. Polymer 2007, 48, 6791–6798. (5) Bogdanov, B.; Vidts, A.; Schacht, E.; Berghmans, H. Macromolecules 1999, 32, 726–731. (6) Ray, S. S.; Bousmina, M. Prog. Mater. Sci. 2005, 50, 962. (7) Eastmond, G. C. AdV. Polym. Sci. 2000, 149, 59–223. (8) Gan, Z.; Yu, D.; Zhong, Z.; Liang, Q.; Jing, X. Polymer 1999, 40, 2859–2862. (9) Liu, L.; Li, S.; Garreau, H.; Vert, M. Biomacromolecules 2000, 1, 350–359. (10) Broz, M. E.; VanderHart, D. L.; Washburn, N. R. Biomaterials 2003, 24, 4181–4190. (11) Sarazin, P.; Li, G.; Orts, W. J.; Favis, B. D. Polymer 2008, 49, 599– 609. (12) Lopez-Rodriguez, N.; Lopez-Arraiza, A.; Meaurio, E.; Sarasua, J. R. Polym. Eng. Sci. 2006, 46, 1299–1308. (13) Dell Erba, R.; Groeninckx, G.; Maglio, G.; Malinconico, M.; Migliozzi, A. Polymer 2001, 42, 7831–7840. (14) Tsuji, H.; Ishizaka, T. Macromol. Biosci. 2001, 1, 59–65. (15) Sarazin, P.; Roy, X.; Favis, B. D. Biomaterials 2004, 25, 5965–5978. (16) Roy, X.; Sarazin, P.; Favis, B. D. AdV. Mater. 2006, 18, 1015–1019. (17) Han, S.; Moon, T. J.; Bae, Y. C.; Yi, S. J.; Lee, S. H. Polymer 1998, 39, 1113–1117. (18) Wu, D. F.; Zhang, Y. S.; Zhang, M.; Zhou, W. D. Eur. Polym. J. 2008, 44, 2171–2183. (19) Maglio, G.; Migliozzi, A.; Palumbo, R.; Immirzi, B.; Volpe, M. G. Macromol. Rapid Commun. 1999, 20, 236–238. (20) Na, Y. H.; He, Y.; Shuai, X. T.; Kikkawa, Y.; Doi, Y.; Inoue, Y. Biomacromolecules 2002, 3, 1179–1186.

Wu et al. (21) Li, S. M.; Liu, L. J.; Garreau, H.; Vert, M. Biomacromolecules 2003, 4, 372–377. (22) Paul, D. R.; Bucknall, C. B. Polymer Blends: Formulation & Performance; John Wiley & Sons: New York, 2000. (23) Utracki, L. A. Polymer Alloys and Blends; Hanser Publishers: Munich, 1989. (24) Khatua, B. B.; Lee, D. J.; Kim, H. Y.; Kim, J. K. Macromolecules 2004, 37, 2454–2459. (25) Li, Y. J.; Shimizu, H. Polymer 2004, 45, 7381–7388. (26) Ray, S. S.; Bousmina, M. Macromol. Rapid Commun. 2005, 26, 1639– 1646. (27) Ray, S. S.; Pouliot, S.; Bousmina, M.; Utracki, L. A. Polymer 2004, 45, 8403–8413. (28) Wang, Y.; Zhang, Q.; Fu, Q. Macromol. Rapid Commun. 2003, 24, 231–235. (29) Li, Y. J.; Shimizu, H. Macromol. Rapid Commun. 2005, 26, 710– 715. (30) Zou, H.; Zhang, Q.; Tan, H.; Wang, K.; Du, R. N.; Fu, Q. Polymer 2006, 47, 6–11. (31) Chow, W. S.; Mohd Ishak, Z. A.; Karger-Kocsis, J.; Apostolov, A. A.; Ishiaku, U. S. Polymer 2003, 44, 7427–7440. (32) Hong, J. S.; Kim, Y. K.; Ahn, K. H.; Lee, S. J.; Kim, C. Y. Rheol. Acta 2007, 46, 469–478. (33) Wu, D. F.; Wu, L. F.; Zhang, M.; Zhou, W. D.; Zhang, Y. S. J. Polym. Sci., Part B: Polym. Phys. 2008, 46, 1265–1279. (34) Subramoney, S. AdV. Mater. 1998, 10, 1157. (35) Moniruzzaman, M.; Winey, K. I. Macromolecules 2006, 39, 5194– 5205. (36) Du, F. M.; Scogna, R. C.; Zhou, W.; Brand, S.; Fischer, J. E.; Winey, K. I. Macromolecules 2004, 37, 9048–9055. (37) Potschke, P.; Fornes, T. D.; Paul, D. R. Polymer 2002, 43, 3247– 3255. (38) Potschke, P.; Abdel-Goad, M.; Alig, I.; Dudkin, S.; Lellinger, D. Polymer 2004, 45, 8863–8870. (39) Nogales, A.; Broza, G.; Roslaniec, Z.; Schulte, K.; Sics, I.; Hsiao, B. S.; Sanz, A.; Garcia-Gutierrez, M. C.; Rueda, D. R.; Domingo, C.; Ezquerra, T. A. Macromolecules 2004, 37, 7669–7672. (40) Hu, G. G.; Zhao, C. G.; Zhang, S. M.; Yang, M. S.; Wang, Z. G. Polymer 2006, 47, 480–488. (41) Wu, D. F.; Wu, L.; Zhang, M. J. Polym. Sci., Part B: Polym. Phys. 2007, 45, 2239–2251. (42) Wu, D. F.; Wu, L.; Sun, Y. R.; Zhang, M. J. Polym. Sci., Part B: Polym. Phys. 2007, 45, 3137–3147. (43) Wu, D. F.; Wu, L.; Zhang, M.; Zhao, Y. L. Polym. Degrad. Stab. 2008, 93, 1577–1584. (44) Shi, X. F.; Hudson, J. L.; Spicer, P. P.; Tour, J. M.; Krishnamoorti, R.; Mikos, A. G. Biomacromolecules 2006, 7, 2237–2242. (45) Shi, X. F.; Hudson, J. L.; Spicer, P. P.; Tour, J. M.; Krishnamoorti, R.; Mikos, A. G. Nanotechnology 2005, 16, S531–S538. (46) Mei, F.; Zhong, J. S.; Yang, X. P.; Ouyang, X. Y.; Zhang, S.; Hu, X. Y.; Ma, Q.; Lu, G. J.; Ryu, S. K.; Deng, X. L. Biomacromolecules 2007, 8, 3729–3735. (47) Zou, H.; Wang, K.; Zhang, Q.; Fu, Q. A. Polymer 2006, 47, 7821– 7826. (48) Potschke, P.; Kretzschmar, B.; Janke, A. Compos. Sci. Technol. 2007, 67, 855–860. (49) Jung, R.; Park, W. I.; Kwon, S. M.; Kim, H. S.; Jin, H. J. Polymer 2008, 49, 2071–2076. (50) Liang, H. J. Macromolecules 1999, 32, 8204–8209. (51) Lee, H. M.; Park, O. O. J. Rheol. 1994, 38, 1405–1425. (52) Yu, W.; Bousmina, M.; Grmela, M.; Zhou, C. X. J. Rheol. 2002, 46, 1401–1418. (53) Van Gurp, M.; Palmen, J. Rheol. Bull. 1998, 67, 5–8. (54) Bhattacharya, S. K. Metal Filled Polymers; Dekker: New York, 1986. (55) Ishiaku, U. S.; Pang, K. W.; Lee, W. S.; Mohd. Ishak, Z. A. Eur. Polym. J. 2002, 38, 393–401. (56) Rezguil, F.; Swistek, M.; Hiver, J. M.; G’Sell, C.; Sadoun, T. Polymer 2005, 46, 7370–7385.

BM801183F

Selective Localization of Multiwalled Carbon Nanotubes in Poly(Îµ

Recommend Documents