- Email: [email protected]
Changes of morphology and properties of block copolymers induced by carbon nanotubes
Polymer 54 (2013) 2285e2291
Contents lists available at SciVerse ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
Changes of morphology and properties of block copolymers induced by carbon nanotubes Karell Saint-Aubin, Philippe Poulin, Christèle Jaillet, Maryse Maugey, Cécile Zakri* Université de Bordeaux, CNRS, Centre de Recherche Paul Pascal, 115 Avenue Schweitzer, 33600 Pessac, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 January 2013 Received in revised form 1 March 2013 Accepted 4 March 2013 Available online 13 March 2013
Carbon nanotubes have been extensively used in isotropic polymer media as mechanical reinforcements or as conductive fillers. New phenomena arise when the polymer matrix is made of an ordered block copolymer. Dispersions of nanotubes stabilized by block copolymers in selective solvents can be used to cast composite films in which the nanotubes are segregated in microdomains of the structured polymer. This concept is here investigated for the case of carbon nanotubes in a poly(styrene)-b-poly(butadiene)b-poly(methylmethacrylate) terpolymer (SBM). It is observed that casting films of SBM from different solvents in the presence or absence of nanotubes can lead to different morphologies with distinct mechanical and electrical properties. In particular it was found that neat SBM cast from a mixture of cyclohexane and acetone exhibits a cylindrical microstructure. This metastable form is mechanically weak and brittle. But the polymer adopts a stable lamellar morphology in the presence of nanotubes. This phase exhibits much better mechanical properties. The capability of nanotubes to alter the morphology of a block copolymer and to lead to large improvements of mechanical properties appears therefore as a new mechanism of mechanical reinforcement beyond the already reported mechanisms of direct reinforcement and network formation. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Carbon nanotube Block copolymer Nanocomposite
1. Introduction Block copolymers exhibit a rich phase behavior and the capability to adsorb at interfaces. They have been used over the last years as stabilizers of various colloids, including carbon nanotubes in aqueous or organic media [1e11]. Block copolymers can undergo microphase separations [12e17] and be used as self-organized matrices for nanoparticles; offering thereby the possibility to develop novel nanocomposite structures. Depending on the relative size of the constituents, block copolymers can segregate into microphase lamellae, cylinders and spheres, with several other microstructures between these boundaries. In addition to this rich phase behavior at equilibrium, it is also known that the properties of block copolymers depend on their processing. In particular, metastable microstructures can be quenched via different thermal or casting treatments, resulting in materials with potentially different properties in spite of a similar chemical composition. Embedding colloids in such matrices is reminiscent of the approach which consists in dispersing carbon nanotubes in thermotropic or lyotropic liquid crystals [18,19] to achieve anisotropic functional
* Corresponding author. E-mail address: [email protected] (C. Zakri). 0032-3861/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymer.2013.03.006
materials. Several examples of carbon nanotube-block copolymer composites have been reported in the last years. This class of materials holds a great promise for potential applications in organic electronics, optics, biotechnologies and reinforced polymers [9,20e 28]. Most studies in the field have focussed on diblock copolymers or ABA type triblock copolymers. More recently, Pieré et al. reported pioneering studies of the inclusion of multiwall carbon nanotubes (MWNTs) in ABC triblock copolymers [8,9]. Such polymers offer a greater degree of complexity associated to a large potential of properties and possible functionalities. Pieré et al. investigated poly(styrene)-b-poly-(butadiene)-b-poly(methyl methacrylate) systems (SBM). The properties of such polymers depend on their microstructure, in their neat state [13e17,29] and also on the dispersion of the inclusions [9]. In particular, it was observed that well dispersed MWNTs can induce a direct mechanical reinforcement due to the stress transfer from the polymer to the nanotubes. This type of reinforcement can be described by micromechanical models of particle reinforced composites, such as the HalpineTsaï equations [9,30]. Instead at high temperature the observed increase of Young’s modulus with the particle fraction was ascribed to the formation of a nanotube network in the softened polymer. Regardless the physical mechanisms of reinforcement it was observed that the presence of MWNTs did not affect the SBM polymer morphology in the conditions of investigation. The
2286
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
addition of carbon nanotubes to block copolymers, and to polymers in general, is an effective approach to provide electrical conductivity. Conductive composites can find applications from simple antistatic materials to more sophisticated components of organic electronics. Pieré et al. demonstrate that the addition of MWNTs in SBM terpolymers led to conductive composites [9]. The materials were found to be conductive when the weight fraction of nanotubes was above 1 wt%. Actually the so-called percolation threshold was estimated just below 1 wt%. The same electrical behavior and percolation threshold were observed for different SBM materials [9]. The experimentally measured percolation threshold was consistent with the percolation threshold theoretically expected considering the aspect ratio of the used MWNTs [31,32]. In this work, we study the behavior of single wall carbon nanotubes (SWNTs) or MWNTs stabilized by SBM polymers in two distinct selective solvents. The achieved dispersions are used to produce nanotube-SBM composites by solvent casting. Different morphologies are achieved depending on the solvent. In particular non equilibrium structures can be quenched during solvent evaporation [29]. Surprisingly it is observed in the present work that carbon nanotubes strongly alter the casting behavior. The block copolymers indeed exhibit distinct microstructures in the absence or presence of nanotubes. A related phenomenon was reported for poly(styrene-b-isoprene-b-styrene) triblock copolymer [28]. However, in this case, the nanotubes were functionalized and a surfactant was added to stabilize the system. Changes of morphologies were ascribed to interactions of the nanotubes with the polystyrene (PS) blocks. Here different solvents are used to modify their selectivity towards the blocks of the terpolymer. In particular mixing solvents that exhibit different rate of evaporation allows the selectivity of the medium to vary during film casting. As a result different morphologies can be obtained. A cylindrical morphology is achieved in the absence of nanotubes or any other additive when a mixture of acetone and cyclohexane is used. This metastable morphology results from the segregation of poly-(methyl methacrylate) (PMMA) blocks in the cores of cylindrical micelles. Mechanical characterizations show that the cylindrical morphology is particularly brittle. By contrast the same polymer forms a stiff and thermodynamically lamellar morphology in presence of unmodified nanotubes. The role of the nanotubes is discussed by considering their capability to act as nucleating agent for the most stable phase of the block copolymer. Indeed the nanotubes onto which polybutadiene (PB) blocks are anchored hinder the segregation of the PMMA blocks as observed in the absence of nanotubes. The capability of nanotubes to alter the morphology of a block copolymer and to lead to large improvements of mechanical properties appears therefore as a new mechanism of mechanical reinforcement beyond the already reported mechanisms of direct reinforcement and network formation. The electrical properties were also found to strongly depend on the solvent used to cast the composites. Systems that are expected to be well above the percolation threshold were found to be insulating if the films were cast from the solvent mixture of acetone and cyclohexane (A/C). By contrast systems with exactly the same final chemical composition are conductive if they are cast from solutions in dichloromethane (DCM). This strong difference is qualitatively discussed by considering the variable degree of ordering in systems cast from distinct solvents. 2. Experimental section Raw block copolymer material was provided by ARKEMA. The product contains approximately 70% wt of triblock SBM copolymer, and 30% wt of SB diblock copolymer impurity. The product was purified following a method developed by Di Cola et al. [33]. Raw
material was put in a 60 wt%/40 wt% mixture of cyclohexane/nheptane at 90 C during 2 h. By contrast to SB, SBM molecules are not soluble in this solvent mixture which is a bad solvent for the PMMA block. The non-dissolved fraction which is solely comprised of SBM was collected after centrifugation. The process was repeated twice in order to optimize the purification of the material. Purity was checked using 1H NMR and exclusion chromatography. The purified terpolymer has the following composition: 32% wt of PS (molar fraction 24%), 36% wt of PB (molar fraction 51%) and 32% wt of PMMA (molar fraction 25%). Its average molecular weight is 64,000 g mol1. Chromatography indicates a polydispersity of 1.4. The glass transition temperature of each block is respectively 96 C, 86 C and 144 C for PS, PB and PMMA. Multiwalled carbon nanotubes were provided by ARKEMA (GraphistrengthÒ C100 batch 6068). The nanotubes are synthesized by a catalyzed chemical vapor deposition process. Their external diameter lies in the 10e15 nm range. Their length exceeds several microns before sonication. Raw nanotubes contain catalytic iron nanoparticles supported by alumina particles. The total weight fraction of impurities is about 15 wt%. The nanotubes were purified via the following treatment: 9 wt% of MWNTs were added to a 15 wt% sulfuric acid aqueous solution. After 5 h under reflux at T ¼ 105 C, the materials were washed with deionized water, filtered and kept in water under the form of a wet-cake that contains 10 wt% of carbon nanotubes. Lastly, this cake was freeze-dried. Freeze-drying leads to a fluffy powder that can be easily processed and mixed with liquid and polymer materials [34]. Thermal gravimetric analyzes showed that the weight fraction of impurities in purified and dried nanotubes was decreased down to 3 wt%. Single walled carbon nanotubes (SWNTs) were provided by Thomas Swan (ElicarbÒ batch K3772). These tubes are also made by a catalyzed chemical vapor deposition process. They are directly provided as purified materials by the provider. Nanotube dispersions were prepared as follows for both MWNTs and SWNTs: purified nanotubes were put in a given organic solvent and the solution was sonicated at a power of 20 W during 60 min. A Branson sonifier, model S-250D operating at a 20 kHz and associated to a 13 mm disruptor horn and a 3 mm tapered microtip was used. Two different solvents were used during this study: dichloromethane (DCM) and a mixture of acetone and cyclohexane 50 vol%/50 vol% (A/C). The volume of the sonicated dispersion was 10 mL. Cooling the samples in a water-ice bath prevented the suspension from overheating during sonication. Purified SBM was then added to the dispersion which was again sonicated during 60 min in the same conditions. The amount of added SBM is chosen in order to achieve a desired composite composition after evaporation of the solvent. Table 1 shows the Hildebrand solubility parameters d of the three SBM blocks, compared to that of DCM, acetone and cyclohexane. The Hildebrand value of a solvent mixture can be determined by averaging the Hildebrand values of the individual solvents by volume. The used mixture of acetone and cyclohexane has thus a solubility parameter of about 9.1 (cal cm3)1/2. The components are compatible when their solubility parameters are similar. By contrast dissimilar values yield limited solubility.
Table 1 Hildebrand solubility parameters d, in (cal cm3)1/2 of the PS, PB and PMMA blocks and of DCM, acetone and cyclohexane (values extracted from the Handbook of Polymers). SBM blocks
d
Solvents
PS
PB
PMMA
DCM
Acetone
Cyclohexane
9.1
8.4
9.3
9.8
9.8
8.3
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
Homogeneity of the nanotube dispersions was first assessed by optical microscopy. The nanotube-SBM dispersions were placed in rectangular Teflon@ trough 40 mm in length, 10 mm in width and 0.5 mm deep. Film casting of nanotube-SBM composites was achieved by solvent evaporation. The dilute solutions and composites have been characterized by transmission electronic microscopy (TEM) with a Hitachi 120 kV microscope. Ultrathin slices less than 100 nm in thickness were prepared by microtomy at 100 C. The slices were deposited on Butvar@ or Formar@ membranes and stained by exposure to osmium tetraoxide (OsO4) during 30 min. By contrast to PMMA, the PS and PB phases are stained by OsO4 and respectively appear as light gray and black. The same coloration method was used for the observation of the nanotubes dispersions. In this case, a droplet of solution was deposited on a carbon/nickel grid. TEM observation was realized after the almost instantaneous evaporation of the solvent. The structure of the films was also characterized by small angle X-ray scattering (SAXS) experiments, realized on a Bruker Nanostar diffractometer. This set-up allowed investigations in a scattering wave-vector range from 0.008 to 1 Å1. Mechanical characterizations of neat SBM and nanotube-SBM composite films were realized under tensile load. These characterizations were performed with a Zwick Z2.5/TN1S mechanical testing instrument. The tested samples were cut in the cast films and were 30 mm in length, 2 mm in width and approximately 30 mm thick. The samples were fixed in the jaws of the sample holder in order to have a gage length of 15 mm. Similar samples were used for electrical resistivity measurements. Such measurements were performed via two probes made of silver paste [35]. The resistance was measured with a Keithley 2000 instrument and the resistivity was deduced from normalization of the resistance value by the dimensions of the sample. 3. Results and discussion 1 Solvent effects on nanotube-SBM mixtures in solution e Solutions containing 0.2 wt% of SBM and 0.025 wt% of carbon nanotubes have been prepared following the procedure described in the experimental section. SWNTs and MWNTs were used. Both types of nanotubes could be well dispersed in DCM and A/C mixture. Macroscopic dispersions were black without aggregates visible by naked eye. A control by optical microscopy allowed us to check the absence of aggregates at a smaller scale. As shown in Fig. 1a, there are no visible aggregates in the dispersion, proving that SBM is an efficient dispersant of nanotubes in the investigated organic solvents. Similar observations were done for SWNTs and MWNTs in DCM and A/C mixture. Fig. 1b and c is electron micrographs of dried solutions of MWNTs in DCM. The MWNTs appear well dispersed; reflecting that they remained efficiently stabilized by the SBM polymers during evaporation of the solvent on the grids for electron microscopy experiments. Similar observations were
2287
achieved for nanotubes in the A/C mixture in spite of its different solubility parameters. It is also observed for both systems that the nanotube interfaces are surrounded by a black coating. This coating is more easily seen in Fig. 1c at a greater magnification. Considering the coloration method used, the coating is likely due to the presence of PB blocks adsorbed at the surface of the nanotubes. The fact that PB blocks preferentially adsorb at the surface of nanotubes in both DCM and A/C mixture can be understood by considering its large difference of solubility parameter with the used solvents. The nanotubes are stabilized against aggregation by steric repulsions provided by the PS and PMMA blocks for which DCM and the A/C mixture are good solvents. The present stabilization mechanism in selective organic solvents is similar to that observed in other nanotube-block copolymer dispersions [3,8,9]. 2 Pure SBM films e SBM films were cast from solutions in the two investigated different solvents. The structure of the obtained films was characterized by TEM and SAXS. The solutions, initially containing 5 wt% of SBM, were put into Petri dishes during several hours to let the solvent slowly evaporate. The films were then annealed at high temperature. We tested different speeds of evaporation by opening or closing the Petri dishes and different annealing temperature, from 50 C to 120 C, all of them led to the same structuration of the films. The following results and mechanical characterizations were performed for films dried in closed Petri dishes for 48 h, and then annealed during 2 h at 100 C. Fig. 2a shows an electron micrograph of an SBM film cast from an SBM solution in DCM and Fig. 2b the SAXS spectrum of the material. The X-ray intensity exhibits three broad peaks on top of a large scattered intensity background at low scattering wave vector. The positions of the three peaks can be assigned to the three first orders of a lamellar structure with a smectic distance D of about 60 nm. The large width of the peaks is probably due to a small size of the monodomains and to the presence of a large density of defects. Nevertheless, TEM experiments confirm the lamellar microstructure of the SBM cast from solutions in DCM. This result is in good agreement with previous studies of different but still symmetric SBM [13,29]. Such kind of polymers is indeed expected to form a lamellar microstructure at thermodynamic equilibrium [36]. In addition, it has been shown that the smectic spacing of the lamellar structure should scale as function of the molecular weight of the polymer Mn as D(nm) ¼ 0.063.Mn0:619 [29]. Mn being of about 64,000 g mol1, the D spacing is expected to be 59.5 nm, in very good agreement with the X-ray characterizations. The scenario seems fundamentally different for films cast from A/C solutions. The TEM image of Fig. 2c shows a structure that differs from the equilibrium structure identified with DCM. This structure looks like a hexagonal, yet poorly ordered, phase. The
Fig. 1. Micrographs of a dispersion containing MWNTs and SBM in DCM. (a) optical micrograph (scale bar 200 mm). (b) and (c) TEM micrographs (scale bars: (b)200 nm, (c)100 nm).
2288
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
Fig. 2. (a)-TEM image of an ultrathin film of SBM prepared from a solution in DCM (scale bar 100 nm). (b)-SAXS spectrum of the (a) film (I ¼ scattered intensity and q ¼ scattering wave vector) (c)-TEM image of an ultrathin film of SBM prepared from a solution in A/C solvent (scale bar 100 nm). (d)-SAXS spectrum of the (c) film. The first, second and third order diffraction peaks are labeled by numbers 1, 2 and 3 in (b) and (d).
SAXS spectrum of Fig. 2d confirms this difference. The X-ray data show three large peaks positioned at relative wave vectors which are characteristic of a hexagonal packing, with a lattice distance of about 61 nm. The large widths of the peaks presumably reflect the poor ordering observed by TEM. The possibility to achieve metastable structures with SBM materials cast from different solvents or mixtures of solvent was already reported [29]. Nevertheless, solvents in previously investigated mixtures had approximately similar vapor pressures and rates of evaporation. In the present case, acetone is more volatile and has a greater rate of evaporation than cyclohexane. As a consequence the system becomes temporarily enriched with cyclohexane when the material dries. Cyclohexane is a bad solvent for the PMMA and PS blocks and a good solvent for the PB blocks. This selectivity promotes the segregation of PMMA blocks and the swelling of the PB blocks. This effect can explain the arrangement of white domains of PMMA surrounded by a dark continuous matrix of PB as observed in the TEM image of Fig. 2c. The PMMA blocks arrange in cylindrical domains that form an hexagonal packing. This temporary packing cannot revert towards a lamellar equilibrium structure when the acetone and cyclohexane have been fully evaporated. The PB and PS blocks are maintained at the outer part of the micelles with frozen cores of PMMA; leading to a quenched metastable hexagonal structure. 3 Nanotube-SBM composite films e Composite films have been obtained from casting dispersions of SWNTs or MWNTs and SBM in different solvents. The weight fraction of nanotubes after solvent evaporation is 5 wt%. TEM images and SAXS data of the achieved composites are presented in Fig. 3a and b. The electron micrographs of the polymer structure and X-ray data are similar for SWNTs and MWNTs. Nevertheless, because of their small diameter SWNTs can be hardly seen in the present experimental conditions.
data of Fig. 3b. Nevertheless, the scattering intensity profile differs from that of neat SBM. This difference can be ascribed to the scattering from the nanotubes and from some disorder introduced by the nanotubes. The composite films cast from the A/C mixture exhibits a less ordered structure shown in Fig. 3c. But this structure still clearly contains large domains with lamellar ordering. The morphology is very different from the metastable hexagonal structure of neat SBM. The SAXS experiments confirm this difference and reveal the lamellar ordering of composite films cast from A/C mixture. The same behavior is observed for SWNTs and MWNTs. It can be deduced that the presence of nanotubes promotes the formation of the equilibrium microstructure, i.e. the lamellar phase. This surprising behavior can be explained by considering the evaporation process and the adsorption of polymer at the interface of the nanotubes. Before evaporation the PB blocks are strongly anchored at the nanotube interface, the A/C mixture being a bad solvent for PB and a good solvent for PS and PMMA. The presence of the nanotubes artificially induces the packing of the PB blocks and the segregation of PMMA and PS at the outer part of the micelles. Nevertheless, indicated above, the system becomes temporarily enriched with cyclohexane during drying. In these conditions, the solvent mixture is becoming a bad solvent for PMMA blocks and a better solvent for PB. In an unconstrained material this would lead to the segregation of PMMA blocks in cores of micelles as previously observed. This segregation is here impeded by the trapping and adsorption of the PB blocks at the interface of the nanotubes. Desorption of the polymer from the nanotubes and polymer restructuration require too much energy or too much time to take place before the system fully dries. As a result, the metastable cylindrical morphology achieved in the absence of nanotubes phase cannot form in the present conditions. Instead the nanotubes can be viewed as nucleating agents of the lamellar equilibrium structure.
A lamellar structure is visible in the TEM image of Fig. 3a for films cast from solutions in DCM. This observation is confirmed by the diffraction first and second order peaks observed in the X-ray
4 Mechanical characterizations e Tensile load experiments have been performed on the different films described above. Fig. 4 the stress vs strain curves for neat SBM and composites made
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
2289
Fig. 3. (a)-TEM image of an ultrathin film of MWNT-SBM composite cast from a DCM solution (scale bar 200 nm). (b)-SAXS spectrum of the (a) film (I ¼ scattered intensity and q ¼ scattering wave vector) (c)-TEM image of an ultrathin film of MWNT-SBM prepared from an A/C solution (scale bar 200 nm). (d)-SAXS spectrum of the (c) film. The dashed lines in (b) and (d) represent the X-ray data of the neat SBM films (data shown in Fig. 2).
from DCM and A/C solutions for SWNTs and MWNTs. The Young’s modulus, Y, the strength at break and the failure strain deduced from these experiments are listed in Table 2. As indicated above, the nanotubes do not alter the morphology of the SBM matrix when the materials are cast from DCM solutions. The polymer has always the stable lamellar morphology. In these conditions, the nanotubes are expected to act as direct mechanical reinforcements. This effect of reinforcement is actually observed. The Young’s modulus of composites is increased from 560 MPa to respectively 1350 and 1810 MPa in the presence of MWNTs or SWNTs. The better efficiency of SWNTs, compared to MWNTs, can be understood by their greater aspect ratio and better intrinsic mechanical properties [37]. The strength at break is also clearly improved from 12 MPa for the neat SBM to respectively 32 MPa and 50 MPa in the presence of MWNTs or SWNTs. More surprising effects are observed in the case of films cast from A/C solvents. In this case the neat SBM, which has a cylindrical morphology, is weak and brittle. Its Young’s modulus is about 250 MPa and the strength at break is of only 4 MPa. The Young’s modulus is increased by a factor 3.4 and 11 in the presence of MWNTs or SWNTs. Significant improvements are also observed for the strength at break which is respectively multiplied by 3 and 10 in the presence of MWNTs or SWNTs. These strong variations of mechanical properties are likely due to the changes of morphology induced by the nanotubes. Indeed, as indicated above, the nanotubes promote the formation of
Table 2 Young’s modulus Y, strength and strain to failure of pure SBM, MWNT-SBM and SWNT-SBM composites prepared from DCM or A/C solvents. Solvent
Fig. 4. Stress vs strain curves of neat SBM and nanotube-SBM composites. (a)-films made from DCM solutions, (b)-films made from A/C solutions.
Y (MPa) Strength (MPa) Strain to failure (%)
SBM
MWNT-SBM
SWNT-SBM
DCM
A/C
DCM
A/C
DCM
A/C
560 12 22
250 4 2
1350 32 24
851 12 3
1810 50 14
2720 41 11
2290
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
Table 3 Electrical resistivity MWNT-SBM and SWNT-SBM composites prepared from DCM or A/C solvents.
DCM A/C
MWNT-SBM
SWNT-SBM
450 U cm >108 U cm
190 U cm >108 U cm
the stable lamellar morphology. The latter has better mechanical properties than the cylindrical microstructure. Therefore the effect of the carbon nanotubes is double: they act as direct stiff reinforcements and stabilize the microstructure that exhibits the better mechanical properties. 5 Electrical conductivity of nanotube-SBM composites e Values of the electrical resistivity of nanotube-SBM composites cast from DCM or A/C mixtures are listed in Table 3. All the materials contain the same fraction of carbon nanotubes which is 5 wt%. This weight fraction is above the expected percolation threshold of the composites. This explains why materials cast from DCM solutions are electrically conductive. By contrast, the resistivity of systems cast from the A/C solvent mixture could not be measured via the present experimental procedure. This means that the resistivity for both MWNT-SBM and SWNT-SBM composites is above 108 U cm when they are cast from the A/C mixture. The composites can thus be considered still insulating in spite of their relatively large fraction of carbon nanotubes. The difference can be explained by considering the distinct degree of ordering of composites cast from different solvents. Indeed, materials obtained after evaporation of DCM exhibit a well ordered lamellar morphology with large monodomains. The nanotubes are segregated in the PB domains and can form a continuous conductive network throughout the samples. Composites obtained after evaporation of the A/C mixture exhibit also a lamellar morphology. But, as indicated above, the materials are poorly ordered. The domains are small and the materials contain a large density of defects with even possibly some domains with a cylindrical morphology. In such conditions the nanotubes which are selectively segregated in the PB phase cannot form a continuous conductive network even at relatively high concentration. The present results show how the structure of block copolymer can strongly alter the electrical conductivity of composite in comparison to a conventional isotropic matrix. 4. Conclusion The inclusion of carbon nanotubes in a block copolymer matrix leads to original phenomena and requires a delicate control of processing to achieve desired properties. It was already known that carbon nanotubes could alter the morphology of a block copolymer. It was shown in this work that this alteration of morphology can be associated to large changes of electrical and mechanical properties. These changes are essentially dominated by the morphology of the polymers rather than by the intrinsic properties of the nanotubes. For example the variation from a cylindrical to a lamellar microstructure induced by the nanotubes leads to improvements of mechanical properties superior to those more conventionally observed when the nanotubes act as reinforcement agents. The alteration of matrix morphology has also a direct impact on the electrical properties. It has been shown that even at relatively high fraction of carbon nanotubes a composite can remain insulating. This behavior differs from the behavior of an isotropic medium in which the conductivity is essentially controlled by the fraction of the conductive inclusions. The concepts investigated in this work
can provide some guidance for the future processing of complex nanotube-copolymer composites of technological interest. Acknowledgments We acknowledge financial support from Arkema. We also thank Nicolas Guidolin for NMR and chromatography characterizations. In memory of Maryse Maugey. References [1] Shin HI, Min BG, Jeong WY, Park CM. Amphiphilic block copolymer micelles: new dispersant for single wall carbon nanotubes. Macromolecular Rapid Communications 2005;26(18):1451e7. [2] Nativ-Roth E, Shvartzman-Cohen R, Bounioux C, Florent M, Zhang DS, Szleifer I, et al. Physical adsorption of block copolymers to SWNT and MWNT: a nonwrapping mechanism. Macromolecules 2007;40(10):3676e85. [3] Sluzarenko N, Heurtefeu B, Maugey M, Zakri C, Poulin P, Lecommandoux S. Diblock copolymer stabilization of multi-wall carbon nanotubes in organic solvents and their use in composites. Carbon 2006;44(15):3207e12. [4] Cotiuga I, Picchioni F, Agarwal US, Wouters D, Loos J, Lemstra PJ. Blockcopolymer-assisted solubilization of carbon nanotubes and exfoliation monitoring through viscosity. Macromolecular Rapid Communications 2006;27(13):1073e8. [5] Shvartzman-Cohen R, Levi-Kalisman Y, Nativ-Roth E, Yerushalmi-Rozen R. Generic approach for dispersing single-walled carbon nanotubes: the strength of a weak interaction. Langmuir 2004;20(15):6085e8. [6] Park C, Lee S, Lee JH, Lim J, Lee SC, Park M, et al. Controlled assembly of carbon nanotubes encapsulated with amphiphilic block copolymer. Carbon 2007;45(10):2072e8. [7] Nagarajan R, Bradley RA, Nair BR. Thermodynamically stable, size selective solubilization of carbon nanotubes in aqueous solutions of amphiphilic block copolymers. Journal of Chemical Physics 2009;131(10). 104906e1 to 13. Je ne sais pas si cette notation est très académique. [8] Perié T, Brosse AC, Tencé-Girault S, Leibler L. Nanostructured films and composites from carbon nanotubes dispersed by ABC block terpolymers in selective solvent. Polymer 2011;52(14):3065e73. [9] Perié T, Brosse AC, Tencé-Girault S, Leibler L. Mechanical and electrical properties of multi walled carbon nanotube/ABC block terpolymer composites. Carbon 2012;50(8):2918e28. [10] Wang ZM, Liu QC, Zhu H, Liu HF, Chen YM, Yang MS. Dispersing multi-walled carbon nanotubes with water-soluble block copolymers and their use as supports for metal nanoparticles. Carbon 2007;45(2):285e92. [11] Zou JH, Khondaker SI, Huo Q, Zhai L. A general strategy to disperse and functionalize carbon nanotubes using conjugated block copolymers. Advanced Functional Materials 2009;19(3):479e83. [12] Bates FS, Fredrickson GH. Block copolymer thermodynamics e theory and experiment. Annual Review of Physical Chemistry 1990;41:525e57. [13] Stadler R, Auschra C, Beckmann J, Krappe U, Voigtmartin I, Leibler L. Morphology and thermodynamics of symmetrical poly(a-block-b-bloch-c) triblock copolymers. Macromolecules 1995;28(9):3080e97. [14] Zheng W, Wang ZG. Morphology of ABC triblock copolymers. Macromolecules 1995;28(21):7215e23. [15] Corte L, Yamaguchi K, Court F, Cloitre M, Hashimoto T, Leibler L. Annealing and defect trapping in lamellar phases of triblock terpolymers. Macromolecules 2003;36(20):7695e706. [16] Corte L, Rebizant V, Hochstetter G, Tournilhac F, Leibler L. Toughening with little stiffness loss: polyamide filled with ABC triblock copolymers. Macromolecules 2006;39(26):9365e74. [17] Yamaguchi D, Cloitre M, Panine P, Leibler L. Phase behavior and viscoelastic properties of thermoplastic elastomer gels based on ABC triblock copolymers. Macromolecules 2005;38(18):7798e806. [18] Scalia G. Alignment of carbon nanotubes in thermotropic and lyotropic liquid crystals. Chemphyschem 2010;11(2):333e40. [19] Hegmann T, Qi H, Marx VM. Nanoparticles in liquid crystals: synthesis, selfassembly, defect formation and potential applications. Journal of Inorganic and Organometallic Polymers and Materials 2007;17(3):483e508. [20] De Rosa C, Auriemma F, Di Girolamo R, Pepe GP, Napolitano T, Scaldaferri R. Enabling strategies in organic electronics using ordered block copolymer nanostructures. Advanced Materials 2010;22(47):5414e9. [21] Gilmore KJ, Moulton SE, Wallace GG. Incorporation of carbon nanotubes into the biomedical polymer poly(styrene-beta-isobutylene-beta-styrene). Carbon 2007;45(2):402e10. [22] Jeon SJ, Yang SM, Kim BJ, Petrie JD, Jang SG, Kramer EJ, et al. Hierarchically structured colloids of diblock copolymers and Au nanoparticles. Chemistry of Materials 2009;21(16):3739e41. [23] Meier JG, Crespo C, Pelegay JL, Castell P, Sainz R, Maser WK, et al. Processing dependency of percolation threshold of MWCNTs in a thermoplastic elastomeric block copolymer. Polymer 2011;52(8):1788e96. [24] Mountrichas G, Sandanayaka ASD, Economopoulos SP, Pispas S, Ito O, Hasobe T, et al. Photoinduced electron transfer in aqueous carbon nanotube/
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
[25]
[26]
[27]
[28]
[29]
[30]
block copolymer/CdS hybrids: application in the construction of photoelectrochemical cells. Journal of Materials Chemistry 2009;19(47):8990e8. Park HS, Choi BG, Hong WH, Jang SY. Interfacial interactions of single-walled carbon nanotube/conjugated block copolymer hybrids for flexible transparent conductive films. Journal of Physical Chemistry C 2012;116(14):7962e7. Park I, Lee W, Kim J, Park M, Lee H. Selective sequestering of multi-walled carbon nanotubes in self-assembled block copolymer. Sensors and Actuators B-Chemical 2007;126(1):301e5. Peponi L, Tercjak A, Torre L, Mondragon I, Kenny JM. Nanostructured physical gel of SBS block copolymer and Ag/DT/SBS nanocomposites. Journal of Materials Science 2009;44(5):1287e93. Peponi L, Valentini L, Torre L, Mondragon I, Kenny JM. Surfactant assisted selective confinement of carbon nanotubes functionalized with octadecylamine in a poly(styrene-b-isoprene-b-styrene) block copolymer matrix. Carbon 2009;47(10):2474e80. Brinkmann S, Stadler R, Thomas EL. New structural motif in hexagonally ordered cylindrical ternary (ABC) block copolymer microdomains. Macromolecules 1998;31(19):6566e72. Halpin JC, Kardos JL. Halpin-Tsai equations e review. Polymer Engineering and Science 1976;16(5):344e52.
2291
[31] Balberg I, Anderson CH, Alexander S, Wagner N. Excluded volume and its relation to the onset of percolation. Physical Review B 1984;30(7): 3933e43. [32] Grillard F, Jaillet C, Zakri C, Miaudet P, Derre A, Korzhenko A, et al. Conductivity and percolation of nanotube based polymer composites in extensional deformations. Polymer 2012;53(1):183e7. [33] Di Cola E, Fleury C, Panine P, Cloitre M. Steady shear flow alignment and rheology of lamellae-forming ABC triblock copolymer solutions: orientation, defects, and disorder. Macromolecules 2008;41(10):3627e35. [34] Maugey M, Neri W, Zakri C, Derré A, Nnicaud A, Noe L, et al. Substantial improvement of nanotube processability by freeze-drying. Journal of Nanoscience and Nanotechnology 2007;7(8):2633e9. [35] Zamora-Ledezma C, Blanc C, Puech N, Maugey M, Zakri C, Anglaret E, et al. Conductivity anisotropy of assembled and oriented carbon nanotubes (vol. 84, 062701, 2011). Physical Review E 2012;85(1). [36] Auschra C, Stadler R. New ordered morphologies in ABC triblock copolymers. Macromolecules 1993;26(9):2171e4. [37] Xie XL, Mai YW, Zhou XP. Dispersion and alignment of carbon nanotubes in polymer matrix: a review. Materials Science & Engineering R-Reports 2005;49(4):89e112.
Contents lists available at SciVerse ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
Changes of morphology and properties of block copolymers induced by carbon nanotubes Karell Saint-Aubin, Philippe Poulin, Christèle Jaillet, Maryse Maugey, Cécile Zakri* Université de Bordeaux, CNRS, Centre de Recherche Paul Pascal, 115 Avenue Schweitzer, 33600 Pessac, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 January 2013 Received in revised form 1 March 2013 Accepted 4 March 2013 Available online 13 March 2013
Carbon nanotubes have been extensively used in isotropic polymer media as mechanical reinforcements or as conductive fillers. New phenomena arise when the polymer matrix is made of an ordered block copolymer. Dispersions of nanotubes stabilized by block copolymers in selective solvents can be used to cast composite films in which the nanotubes are segregated in microdomains of the structured polymer. This concept is here investigated for the case of carbon nanotubes in a poly(styrene)-b-poly(butadiene)b-poly(methylmethacrylate) terpolymer (SBM). It is observed that casting films of SBM from different solvents in the presence or absence of nanotubes can lead to different morphologies with distinct mechanical and electrical properties. In particular it was found that neat SBM cast from a mixture of cyclohexane and acetone exhibits a cylindrical microstructure. This metastable form is mechanically weak and brittle. But the polymer adopts a stable lamellar morphology in the presence of nanotubes. This phase exhibits much better mechanical properties. The capability of nanotubes to alter the morphology of a block copolymer and to lead to large improvements of mechanical properties appears therefore as a new mechanism of mechanical reinforcement beyond the already reported mechanisms of direct reinforcement and network formation. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Carbon nanotube Block copolymer Nanocomposite
1. Introduction Block copolymers exhibit a rich phase behavior and the capability to adsorb at interfaces. They have been used over the last years as stabilizers of various colloids, including carbon nanotubes in aqueous or organic media [1e11]. Block copolymers can undergo microphase separations [12e17] and be used as self-organized matrices for nanoparticles; offering thereby the possibility to develop novel nanocomposite structures. Depending on the relative size of the constituents, block copolymers can segregate into microphase lamellae, cylinders and spheres, with several other microstructures between these boundaries. In addition to this rich phase behavior at equilibrium, it is also known that the properties of block copolymers depend on their processing. In particular, metastable microstructures can be quenched via different thermal or casting treatments, resulting in materials with potentially different properties in spite of a similar chemical composition. Embedding colloids in such matrices is reminiscent of the approach which consists in dispersing carbon nanotubes in thermotropic or lyotropic liquid crystals [18,19] to achieve anisotropic functional
* Corresponding author. E-mail address: [email protected] (C. Zakri). 0032-3861/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymer.2013.03.006
materials. Several examples of carbon nanotube-block copolymer composites have been reported in the last years. This class of materials holds a great promise for potential applications in organic electronics, optics, biotechnologies and reinforced polymers [9,20e 28]. Most studies in the field have focussed on diblock copolymers or ABA type triblock copolymers. More recently, Pieré et al. reported pioneering studies of the inclusion of multiwall carbon nanotubes (MWNTs) in ABC triblock copolymers [8,9]. Such polymers offer a greater degree of complexity associated to a large potential of properties and possible functionalities. Pieré et al. investigated poly(styrene)-b-poly-(butadiene)-b-poly(methyl methacrylate) systems (SBM). The properties of such polymers depend on their microstructure, in their neat state [13e17,29] and also on the dispersion of the inclusions [9]. In particular, it was observed that well dispersed MWNTs can induce a direct mechanical reinforcement due to the stress transfer from the polymer to the nanotubes. This type of reinforcement can be described by micromechanical models of particle reinforced composites, such as the HalpineTsaï equations [9,30]. Instead at high temperature the observed increase of Young’s modulus with the particle fraction was ascribed to the formation of a nanotube network in the softened polymer. Regardless the physical mechanisms of reinforcement it was observed that the presence of MWNTs did not affect the SBM polymer morphology in the conditions of investigation. The
2286
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
addition of carbon nanotubes to block copolymers, and to polymers in general, is an effective approach to provide electrical conductivity. Conductive composites can find applications from simple antistatic materials to more sophisticated components of organic electronics. Pieré et al. demonstrate that the addition of MWNTs in SBM terpolymers led to conductive composites [9]. The materials were found to be conductive when the weight fraction of nanotubes was above 1 wt%. Actually the so-called percolation threshold was estimated just below 1 wt%. The same electrical behavior and percolation threshold were observed for different SBM materials [9]. The experimentally measured percolation threshold was consistent with the percolation threshold theoretically expected considering the aspect ratio of the used MWNTs [31,32]. In this work, we study the behavior of single wall carbon nanotubes (SWNTs) or MWNTs stabilized by SBM polymers in two distinct selective solvents. The achieved dispersions are used to produce nanotube-SBM composites by solvent casting. Different morphologies are achieved depending on the solvent. In particular non equilibrium structures can be quenched during solvent evaporation [29]. Surprisingly it is observed in the present work that carbon nanotubes strongly alter the casting behavior. The block copolymers indeed exhibit distinct microstructures in the absence or presence of nanotubes. A related phenomenon was reported for poly(styrene-b-isoprene-b-styrene) triblock copolymer [28]. However, in this case, the nanotubes were functionalized and a surfactant was added to stabilize the system. Changes of morphologies were ascribed to interactions of the nanotubes with the polystyrene (PS) blocks. Here different solvents are used to modify their selectivity towards the blocks of the terpolymer. In particular mixing solvents that exhibit different rate of evaporation allows the selectivity of the medium to vary during film casting. As a result different morphologies can be obtained. A cylindrical morphology is achieved in the absence of nanotubes or any other additive when a mixture of acetone and cyclohexane is used. This metastable morphology results from the segregation of poly-(methyl methacrylate) (PMMA) blocks in the cores of cylindrical micelles. Mechanical characterizations show that the cylindrical morphology is particularly brittle. By contrast the same polymer forms a stiff and thermodynamically lamellar morphology in presence of unmodified nanotubes. The role of the nanotubes is discussed by considering their capability to act as nucleating agent for the most stable phase of the block copolymer. Indeed the nanotubes onto which polybutadiene (PB) blocks are anchored hinder the segregation of the PMMA blocks as observed in the absence of nanotubes. The capability of nanotubes to alter the morphology of a block copolymer and to lead to large improvements of mechanical properties appears therefore as a new mechanism of mechanical reinforcement beyond the already reported mechanisms of direct reinforcement and network formation. The electrical properties were also found to strongly depend on the solvent used to cast the composites. Systems that are expected to be well above the percolation threshold were found to be insulating if the films were cast from the solvent mixture of acetone and cyclohexane (A/C). By contrast systems with exactly the same final chemical composition are conductive if they are cast from solutions in dichloromethane (DCM). This strong difference is qualitatively discussed by considering the variable degree of ordering in systems cast from distinct solvents. 2. Experimental section Raw block copolymer material was provided by ARKEMA. The product contains approximately 70% wt of triblock SBM copolymer, and 30% wt of SB diblock copolymer impurity. The product was purified following a method developed by Di Cola et al. [33]. Raw
material was put in a 60 wt%/40 wt% mixture of cyclohexane/nheptane at 90 C during 2 h. By contrast to SB, SBM molecules are not soluble in this solvent mixture which is a bad solvent for the PMMA block. The non-dissolved fraction which is solely comprised of SBM was collected after centrifugation. The process was repeated twice in order to optimize the purification of the material. Purity was checked using 1H NMR and exclusion chromatography. The purified terpolymer has the following composition: 32% wt of PS (molar fraction 24%), 36% wt of PB (molar fraction 51%) and 32% wt of PMMA (molar fraction 25%). Its average molecular weight is 64,000 g mol1. Chromatography indicates a polydispersity of 1.4. The glass transition temperature of each block is respectively 96 C, 86 C and 144 C for PS, PB and PMMA. Multiwalled carbon nanotubes were provided by ARKEMA (GraphistrengthÒ C100 batch 6068). The nanotubes are synthesized by a catalyzed chemical vapor deposition process. Their external diameter lies in the 10e15 nm range. Their length exceeds several microns before sonication. Raw nanotubes contain catalytic iron nanoparticles supported by alumina particles. The total weight fraction of impurities is about 15 wt%. The nanotubes were purified via the following treatment: 9 wt% of MWNTs were added to a 15 wt% sulfuric acid aqueous solution. After 5 h under reflux at T ¼ 105 C, the materials were washed with deionized water, filtered and kept in water under the form of a wet-cake that contains 10 wt% of carbon nanotubes. Lastly, this cake was freeze-dried. Freeze-drying leads to a fluffy powder that can be easily processed and mixed with liquid and polymer materials [34]. Thermal gravimetric analyzes showed that the weight fraction of impurities in purified and dried nanotubes was decreased down to 3 wt%. Single walled carbon nanotubes (SWNTs) were provided by Thomas Swan (ElicarbÒ batch K3772). These tubes are also made by a catalyzed chemical vapor deposition process. They are directly provided as purified materials by the provider. Nanotube dispersions were prepared as follows for both MWNTs and SWNTs: purified nanotubes were put in a given organic solvent and the solution was sonicated at a power of 20 W during 60 min. A Branson sonifier, model S-250D operating at a 20 kHz and associated to a 13 mm disruptor horn and a 3 mm tapered microtip was used. Two different solvents were used during this study: dichloromethane (DCM) and a mixture of acetone and cyclohexane 50 vol%/50 vol% (A/C). The volume of the sonicated dispersion was 10 mL. Cooling the samples in a water-ice bath prevented the suspension from overheating during sonication. Purified SBM was then added to the dispersion which was again sonicated during 60 min in the same conditions. The amount of added SBM is chosen in order to achieve a desired composite composition after evaporation of the solvent. Table 1 shows the Hildebrand solubility parameters d of the three SBM blocks, compared to that of DCM, acetone and cyclohexane. The Hildebrand value of a solvent mixture can be determined by averaging the Hildebrand values of the individual solvents by volume. The used mixture of acetone and cyclohexane has thus a solubility parameter of about 9.1 (cal cm3)1/2. The components are compatible when their solubility parameters are similar. By contrast dissimilar values yield limited solubility.
Table 1 Hildebrand solubility parameters d, in (cal cm3)1/2 of the PS, PB and PMMA blocks and of DCM, acetone and cyclohexane (values extracted from the Handbook of Polymers). SBM blocks
d
Solvents
PS
PB
PMMA
DCM
Acetone
Cyclohexane
9.1
8.4
9.3
9.8
9.8
8.3
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
Homogeneity of the nanotube dispersions was first assessed by optical microscopy. The nanotube-SBM dispersions were placed in rectangular Teflon@ trough 40 mm in length, 10 mm in width and 0.5 mm deep. Film casting of nanotube-SBM composites was achieved by solvent evaporation. The dilute solutions and composites have been characterized by transmission electronic microscopy (TEM) with a Hitachi 120 kV microscope. Ultrathin slices less than 100 nm in thickness were prepared by microtomy at 100 C. The slices were deposited on Butvar@ or Formar@ membranes and stained by exposure to osmium tetraoxide (OsO4) during 30 min. By contrast to PMMA, the PS and PB phases are stained by OsO4 and respectively appear as light gray and black. The same coloration method was used for the observation of the nanotubes dispersions. In this case, a droplet of solution was deposited on a carbon/nickel grid. TEM observation was realized after the almost instantaneous evaporation of the solvent. The structure of the films was also characterized by small angle X-ray scattering (SAXS) experiments, realized on a Bruker Nanostar diffractometer. This set-up allowed investigations in a scattering wave-vector range from 0.008 to 1 Å1. Mechanical characterizations of neat SBM and nanotube-SBM composite films were realized under tensile load. These characterizations were performed with a Zwick Z2.5/TN1S mechanical testing instrument. The tested samples were cut in the cast films and were 30 mm in length, 2 mm in width and approximately 30 mm thick. The samples were fixed in the jaws of the sample holder in order to have a gage length of 15 mm. Similar samples were used for electrical resistivity measurements. Such measurements were performed via two probes made of silver paste [35]. The resistance was measured with a Keithley 2000 instrument and the resistivity was deduced from normalization of the resistance value by the dimensions of the sample. 3. Results and discussion 1 Solvent effects on nanotube-SBM mixtures in solution e Solutions containing 0.2 wt% of SBM and 0.025 wt% of carbon nanotubes have been prepared following the procedure described in the experimental section. SWNTs and MWNTs were used. Both types of nanotubes could be well dispersed in DCM and A/C mixture. Macroscopic dispersions were black without aggregates visible by naked eye. A control by optical microscopy allowed us to check the absence of aggregates at a smaller scale. As shown in Fig. 1a, there are no visible aggregates in the dispersion, proving that SBM is an efficient dispersant of nanotubes in the investigated organic solvents. Similar observations were done for SWNTs and MWNTs in DCM and A/C mixture. Fig. 1b and c is electron micrographs of dried solutions of MWNTs in DCM. The MWNTs appear well dispersed; reflecting that they remained efficiently stabilized by the SBM polymers during evaporation of the solvent on the grids for electron microscopy experiments. Similar observations were
2287
achieved for nanotubes in the A/C mixture in spite of its different solubility parameters. It is also observed for both systems that the nanotube interfaces are surrounded by a black coating. This coating is more easily seen in Fig. 1c at a greater magnification. Considering the coloration method used, the coating is likely due to the presence of PB blocks adsorbed at the surface of the nanotubes. The fact that PB blocks preferentially adsorb at the surface of nanotubes in both DCM and A/C mixture can be understood by considering its large difference of solubility parameter with the used solvents. The nanotubes are stabilized against aggregation by steric repulsions provided by the PS and PMMA blocks for which DCM and the A/C mixture are good solvents. The present stabilization mechanism in selective organic solvents is similar to that observed in other nanotube-block copolymer dispersions [3,8,9]. 2 Pure SBM films e SBM films were cast from solutions in the two investigated different solvents. The structure of the obtained films was characterized by TEM and SAXS. The solutions, initially containing 5 wt% of SBM, were put into Petri dishes during several hours to let the solvent slowly evaporate. The films were then annealed at high temperature. We tested different speeds of evaporation by opening or closing the Petri dishes and different annealing temperature, from 50 C to 120 C, all of them led to the same structuration of the films. The following results and mechanical characterizations were performed for films dried in closed Petri dishes for 48 h, and then annealed during 2 h at 100 C. Fig. 2a shows an electron micrograph of an SBM film cast from an SBM solution in DCM and Fig. 2b the SAXS spectrum of the material. The X-ray intensity exhibits three broad peaks on top of a large scattered intensity background at low scattering wave vector. The positions of the three peaks can be assigned to the three first orders of a lamellar structure with a smectic distance D of about 60 nm. The large width of the peaks is probably due to a small size of the monodomains and to the presence of a large density of defects. Nevertheless, TEM experiments confirm the lamellar microstructure of the SBM cast from solutions in DCM. This result is in good agreement with previous studies of different but still symmetric SBM [13,29]. Such kind of polymers is indeed expected to form a lamellar microstructure at thermodynamic equilibrium [36]. In addition, it has been shown that the smectic spacing of the lamellar structure should scale as function of the molecular weight of the polymer Mn as D(nm) ¼ 0.063.Mn0:619 [29]. Mn being of about 64,000 g mol1, the D spacing is expected to be 59.5 nm, in very good agreement with the X-ray characterizations. The scenario seems fundamentally different for films cast from A/C solutions. The TEM image of Fig. 2c shows a structure that differs from the equilibrium structure identified with DCM. This structure looks like a hexagonal, yet poorly ordered, phase. The
Fig. 1. Micrographs of a dispersion containing MWNTs and SBM in DCM. (a) optical micrograph (scale bar 200 mm). (b) and (c) TEM micrographs (scale bars: (b)200 nm, (c)100 nm).
2288
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
Fig. 2. (a)-TEM image of an ultrathin film of SBM prepared from a solution in DCM (scale bar 100 nm). (b)-SAXS spectrum of the (a) film (I ¼ scattered intensity and q ¼ scattering wave vector) (c)-TEM image of an ultrathin film of SBM prepared from a solution in A/C solvent (scale bar 100 nm). (d)-SAXS spectrum of the (c) film. The first, second and third order diffraction peaks are labeled by numbers 1, 2 and 3 in (b) and (d).
SAXS spectrum of Fig. 2d confirms this difference. The X-ray data show three large peaks positioned at relative wave vectors which are characteristic of a hexagonal packing, with a lattice distance of about 61 nm. The large widths of the peaks presumably reflect the poor ordering observed by TEM. The possibility to achieve metastable structures with SBM materials cast from different solvents or mixtures of solvent was already reported [29]. Nevertheless, solvents in previously investigated mixtures had approximately similar vapor pressures and rates of evaporation. In the present case, acetone is more volatile and has a greater rate of evaporation than cyclohexane. As a consequence the system becomes temporarily enriched with cyclohexane when the material dries. Cyclohexane is a bad solvent for the PMMA and PS blocks and a good solvent for the PB blocks. This selectivity promotes the segregation of PMMA blocks and the swelling of the PB blocks. This effect can explain the arrangement of white domains of PMMA surrounded by a dark continuous matrix of PB as observed in the TEM image of Fig. 2c. The PMMA blocks arrange in cylindrical domains that form an hexagonal packing. This temporary packing cannot revert towards a lamellar equilibrium structure when the acetone and cyclohexane have been fully evaporated. The PB and PS blocks are maintained at the outer part of the micelles with frozen cores of PMMA; leading to a quenched metastable hexagonal structure. 3 Nanotube-SBM composite films e Composite films have been obtained from casting dispersions of SWNTs or MWNTs and SBM in different solvents. The weight fraction of nanotubes after solvent evaporation is 5 wt%. TEM images and SAXS data of the achieved composites are presented in Fig. 3a and b. The electron micrographs of the polymer structure and X-ray data are similar for SWNTs and MWNTs. Nevertheless, because of their small diameter SWNTs can be hardly seen in the present experimental conditions.
data of Fig. 3b. Nevertheless, the scattering intensity profile differs from that of neat SBM. This difference can be ascribed to the scattering from the nanotubes and from some disorder introduced by the nanotubes. The composite films cast from the A/C mixture exhibits a less ordered structure shown in Fig. 3c. But this structure still clearly contains large domains with lamellar ordering. The morphology is very different from the metastable hexagonal structure of neat SBM. The SAXS experiments confirm this difference and reveal the lamellar ordering of composite films cast from A/C mixture. The same behavior is observed for SWNTs and MWNTs. It can be deduced that the presence of nanotubes promotes the formation of the equilibrium microstructure, i.e. the lamellar phase. This surprising behavior can be explained by considering the evaporation process and the adsorption of polymer at the interface of the nanotubes. Before evaporation the PB blocks are strongly anchored at the nanotube interface, the A/C mixture being a bad solvent for PB and a good solvent for PS and PMMA. The presence of the nanotubes artificially induces the packing of the PB blocks and the segregation of PMMA and PS at the outer part of the micelles. Nevertheless, indicated above, the system becomes temporarily enriched with cyclohexane during drying. In these conditions, the solvent mixture is becoming a bad solvent for PMMA blocks and a better solvent for PB. In an unconstrained material this would lead to the segregation of PMMA blocks in cores of micelles as previously observed. This segregation is here impeded by the trapping and adsorption of the PB blocks at the interface of the nanotubes. Desorption of the polymer from the nanotubes and polymer restructuration require too much energy or too much time to take place before the system fully dries. As a result, the metastable cylindrical morphology achieved in the absence of nanotubes phase cannot form in the present conditions. Instead the nanotubes can be viewed as nucleating agents of the lamellar equilibrium structure.
A lamellar structure is visible in the TEM image of Fig. 3a for films cast from solutions in DCM. This observation is confirmed by the diffraction first and second order peaks observed in the X-ray
4 Mechanical characterizations e Tensile load experiments have been performed on the different films described above. Fig. 4 the stress vs strain curves for neat SBM and composites made
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
2289
Fig. 3. (a)-TEM image of an ultrathin film of MWNT-SBM composite cast from a DCM solution (scale bar 200 nm). (b)-SAXS spectrum of the (a) film (I ¼ scattered intensity and q ¼ scattering wave vector) (c)-TEM image of an ultrathin film of MWNT-SBM prepared from an A/C solution (scale bar 200 nm). (d)-SAXS spectrum of the (c) film. The dashed lines in (b) and (d) represent the X-ray data of the neat SBM films (data shown in Fig. 2).
from DCM and A/C solutions for SWNTs and MWNTs. The Young’s modulus, Y, the strength at break and the failure strain deduced from these experiments are listed in Table 2. As indicated above, the nanotubes do not alter the morphology of the SBM matrix when the materials are cast from DCM solutions. The polymer has always the stable lamellar morphology. In these conditions, the nanotubes are expected to act as direct mechanical reinforcements. This effect of reinforcement is actually observed. The Young’s modulus of composites is increased from 560 MPa to respectively 1350 and 1810 MPa in the presence of MWNTs or SWNTs. The better efficiency of SWNTs, compared to MWNTs, can be understood by their greater aspect ratio and better intrinsic mechanical properties [37]. The strength at break is also clearly improved from 12 MPa for the neat SBM to respectively 32 MPa and 50 MPa in the presence of MWNTs or SWNTs. More surprising effects are observed in the case of films cast from A/C solvents. In this case the neat SBM, which has a cylindrical morphology, is weak and brittle. Its Young’s modulus is about 250 MPa and the strength at break is of only 4 MPa. The Young’s modulus is increased by a factor 3.4 and 11 in the presence of MWNTs or SWNTs. Significant improvements are also observed for the strength at break which is respectively multiplied by 3 and 10 in the presence of MWNTs or SWNTs. These strong variations of mechanical properties are likely due to the changes of morphology induced by the nanotubes. Indeed, as indicated above, the nanotubes promote the formation of
Table 2 Young’s modulus Y, strength and strain to failure of pure SBM, MWNT-SBM and SWNT-SBM composites prepared from DCM or A/C solvents. Solvent
Fig. 4. Stress vs strain curves of neat SBM and nanotube-SBM composites. (a)-films made from DCM solutions, (b)-films made from A/C solutions.
Y (MPa) Strength (MPa) Strain to failure (%)
SBM
MWNT-SBM
SWNT-SBM
DCM
A/C
DCM
A/C
DCM
A/C
560 12 22
250 4 2
1350 32 24
851 12 3
1810 50 14
2720 41 11
2290
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
Table 3 Electrical resistivity MWNT-SBM and SWNT-SBM composites prepared from DCM or A/C solvents.
DCM A/C
MWNT-SBM
SWNT-SBM
450 U cm >108 U cm
190 U cm >108 U cm
the stable lamellar morphology. The latter has better mechanical properties than the cylindrical microstructure. Therefore the effect of the carbon nanotubes is double: they act as direct stiff reinforcements and stabilize the microstructure that exhibits the better mechanical properties. 5 Electrical conductivity of nanotube-SBM composites e Values of the electrical resistivity of nanotube-SBM composites cast from DCM or A/C mixtures are listed in Table 3. All the materials contain the same fraction of carbon nanotubes which is 5 wt%. This weight fraction is above the expected percolation threshold of the composites. This explains why materials cast from DCM solutions are electrically conductive. By contrast, the resistivity of systems cast from the A/C solvent mixture could not be measured via the present experimental procedure. This means that the resistivity for both MWNT-SBM and SWNT-SBM composites is above 108 U cm when they are cast from the A/C mixture. The composites can thus be considered still insulating in spite of their relatively large fraction of carbon nanotubes. The difference can be explained by considering the distinct degree of ordering of composites cast from different solvents. Indeed, materials obtained after evaporation of DCM exhibit a well ordered lamellar morphology with large monodomains. The nanotubes are segregated in the PB domains and can form a continuous conductive network throughout the samples. Composites obtained after evaporation of the A/C mixture exhibit also a lamellar morphology. But, as indicated above, the materials are poorly ordered. The domains are small and the materials contain a large density of defects with even possibly some domains with a cylindrical morphology. In such conditions the nanotubes which are selectively segregated in the PB phase cannot form a continuous conductive network even at relatively high concentration. The present results show how the structure of block copolymer can strongly alter the electrical conductivity of composite in comparison to a conventional isotropic matrix. 4. Conclusion The inclusion of carbon nanotubes in a block copolymer matrix leads to original phenomena and requires a delicate control of processing to achieve desired properties. It was already known that carbon nanotubes could alter the morphology of a block copolymer. It was shown in this work that this alteration of morphology can be associated to large changes of electrical and mechanical properties. These changes are essentially dominated by the morphology of the polymers rather than by the intrinsic properties of the nanotubes. For example the variation from a cylindrical to a lamellar microstructure induced by the nanotubes leads to improvements of mechanical properties superior to those more conventionally observed when the nanotubes act as reinforcement agents. The alteration of matrix morphology has also a direct impact on the electrical properties. It has been shown that even at relatively high fraction of carbon nanotubes a composite can remain insulating. This behavior differs from the behavior of an isotropic medium in which the conductivity is essentially controlled by the fraction of the conductive inclusions. The concepts investigated in this work
can provide some guidance for the future processing of complex nanotube-copolymer composites of technological interest. Acknowledgments We acknowledge financial support from Arkema. We also thank Nicolas Guidolin for NMR and chromatography characterizations. In memory of Maryse Maugey. References [1] Shin HI, Min BG, Jeong WY, Park CM. Amphiphilic block copolymer micelles: new dispersant for single wall carbon nanotubes. Macromolecular Rapid Communications 2005;26(18):1451e7. [2] Nativ-Roth E, Shvartzman-Cohen R, Bounioux C, Florent M, Zhang DS, Szleifer I, et al. Physical adsorption of block copolymers to SWNT and MWNT: a nonwrapping mechanism. Macromolecules 2007;40(10):3676e85. [3] Sluzarenko N, Heurtefeu B, Maugey M, Zakri C, Poulin P, Lecommandoux S. Diblock copolymer stabilization of multi-wall carbon nanotubes in organic solvents and their use in composites. Carbon 2006;44(15):3207e12. [4] Cotiuga I, Picchioni F, Agarwal US, Wouters D, Loos J, Lemstra PJ. Blockcopolymer-assisted solubilization of carbon nanotubes and exfoliation monitoring through viscosity. Macromolecular Rapid Communications 2006;27(13):1073e8. [5] Shvartzman-Cohen R, Levi-Kalisman Y, Nativ-Roth E, Yerushalmi-Rozen R. Generic approach for dispersing single-walled carbon nanotubes: the strength of a weak interaction. Langmuir 2004;20(15):6085e8. [6] Park C, Lee S, Lee JH, Lim J, Lee SC, Park M, et al. Controlled assembly of carbon nanotubes encapsulated with amphiphilic block copolymer. Carbon 2007;45(10):2072e8. [7] Nagarajan R, Bradley RA, Nair BR. Thermodynamically stable, size selective solubilization of carbon nanotubes in aqueous solutions of amphiphilic block copolymers. Journal of Chemical Physics 2009;131(10). 104906e1 to 13. Je ne sais pas si cette notation est très académique. [8] Perié T, Brosse AC, Tencé-Girault S, Leibler L. Nanostructured films and composites from carbon nanotubes dispersed by ABC block terpolymers in selective solvent. Polymer 2011;52(14):3065e73. [9] Perié T, Brosse AC, Tencé-Girault S, Leibler L. Mechanical and electrical properties of multi walled carbon nanotube/ABC block terpolymer composites. Carbon 2012;50(8):2918e28. [10] Wang ZM, Liu QC, Zhu H, Liu HF, Chen YM, Yang MS. Dispersing multi-walled carbon nanotubes with water-soluble block copolymers and their use as supports for metal nanoparticles. Carbon 2007;45(2):285e92. [11] Zou JH, Khondaker SI, Huo Q, Zhai L. A general strategy to disperse and functionalize carbon nanotubes using conjugated block copolymers. Advanced Functional Materials 2009;19(3):479e83. [12] Bates FS, Fredrickson GH. Block copolymer thermodynamics e theory and experiment. Annual Review of Physical Chemistry 1990;41:525e57. [13] Stadler R, Auschra C, Beckmann J, Krappe U, Voigtmartin I, Leibler L. Morphology and thermodynamics of symmetrical poly(a-block-b-bloch-c) triblock copolymers. Macromolecules 1995;28(9):3080e97. [14] Zheng W, Wang ZG. Morphology of ABC triblock copolymers. Macromolecules 1995;28(21):7215e23. [15] Corte L, Yamaguchi K, Court F, Cloitre M, Hashimoto T, Leibler L. Annealing and defect trapping in lamellar phases of triblock terpolymers. Macromolecules 2003;36(20):7695e706. [16] Corte L, Rebizant V, Hochstetter G, Tournilhac F, Leibler L. Toughening with little stiffness loss: polyamide filled with ABC triblock copolymers. Macromolecules 2006;39(26):9365e74. [17] Yamaguchi D, Cloitre M, Panine P, Leibler L. Phase behavior and viscoelastic properties of thermoplastic elastomer gels based on ABC triblock copolymers. Macromolecules 2005;38(18):7798e806. [18] Scalia G. Alignment of carbon nanotubes in thermotropic and lyotropic liquid crystals. Chemphyschem 2010;11(2):333e40. [19] Hegmann T, Qi H, Marx VM. Nanoparticles in liquid crystals: synthesis, selfassembly, defect formation and potential applications. Journal of Inorganic and Organometallic Polymers and Materials 2007;17(3):483e508. [20] De Rosa C, Auriemma F, Di Girolamo R, Pepe GP, Napolitano T, Scaldaferri R. Enabling strategies in organic electronics using ordered block copolymer nanostructures. Advanced Materials 2010;22(47):5414e9. [21] Gilmore KJ, Moulton SE, Wallace GG. Incorporation of carbon nanotubes into the biomedical polymer poly(styrene-beta-isobutylene-beta-styrene). Carbon 2007;45(2):402e10. [22] Jeon SJ, Yang SM, Kim BJ, Petrie JD, Jang SG, Kramer EJ, et al. Hierarchically structured colloids of diblock copolymers and Au nanoparticles. Chemistry of Materials 2009;21(16):3739e41. [23] Meier JG, Crespo C, Pelegay JL, Castell P, Sainz R, Maser WK, et al. Processing dependency of percolation threshold of MWCNTs in a thermoplastic elastomeric block copolymer. Polymer 2011;52(8):1788e96. [24] Mountrichas G, Sandanayaka ASD, Economopoulos SP, Pispas S, Ito O, Hasobe T, et al. Photoinduced electron transfer in aqueous carbon nanotube/
K. Saint-Aubin et al. / Polymer 54 (2013) 2285e2291
[25]
[26]
[27]
[28]
[29]
[30]
block copolymer/CdS hybrids: application in the construction of photoelectrochemical cells. Journal of Materials Chemistry 2009;19(47):8990e8. Park HS, Choi BG, Hong WH, Jang SY. Interfacial interactions of single-walled carbon nanotube/conjugated block copolymer hybrids for flexible transparent conductive films. Journal of Physical Chemistry C 2012;116(14):7962e7. Park I, Lee W, Kim J, Park M, Lee H. Selective sequestering of multi-walled carbon nanotubes in self-assembled block copolymer. Sensors and Actuators B-Chemical 2007;126(1):301e5. Peponi L, Tercjak A, Torre L, Mondragon I, Kenny JM. Nanostructured physical gel of SBS block copolymer and Ag/DT/SBS nanocomposites. Journal of Materials Science 2009;44(5):1287e93. Peponi L, Valentini L, Torre L, Mondragon I, Kenny JM. Surfactant assisted selective confinement of carbon nanotubes functionalized with octadecylamine in a poly(styrene-b-isoprene-b-styrene) block copolymer matrix. Carbon 2009;47(10):2474e80. Brinkmann S, Stadler R, Thomas EL. New structural motif in hexagonally ordered cylindrical ternary (ABC) block copolymer microdomains. Macromolecules 1998;31(19):6566e72. Halpin JC, Kardos JL. Halpin-Tsai equations e review. Polymer Engineering and Science 1976;16(5):344e52.
2291
[31] Balberg I, Anderson CH, Alexander S, Wagner N. Excluded volume and its relation to the onset of percolation. Physical Review B 1984;30(7): 3933e43. [32] Grillard F, Jaillet C, Zakri C, Miaudet P, Derre A, Korzhenko A, et al. Conductivity and percolation of nanotube based polymer composites in extensional deformations. Polymer 2012;53(1):183e7. [33] Di Cola E, Fleury C, Panine P, Cloitre M. Steady shear flow alignment and rheology of lamellae-forming ABC triblock copolymer solutions: orientation, defects, and disorder. Macromolecules 2008;41(10):3627e35. [34] Maugey M, Neri W, Zakri C, Derré A, Nnicaud A, Noe L, et al. Substantial improvement of nanotube processability by freeze-drying. Journal of Nanoscience and Nanotechnology 2007;7(8):2633e9. [35] Zamora-Ledezma C, Blanc C, Puech N, Maugey M, Zakri C, Anglaret E, et al. Conductivity anisotropy of assembled and oriented carbon nanotubes (vol. 84, 062701, 2011). Physical Review E 2012;85(1). [36] Auschra C, Stadler R. New ordered morphologies in ABC triblock copolymers. Macromolecules 1993;26(9):2171e4. [37] Xie XL, Mai YW, Zhou XP. Dispersion and alignment of carbon nanotubes in polymer matrix: a review. Materials Science & Engineering R-Reports 2005;49(4):89e112.