Carbon Nanotube Polymer Nanocomposites the Role of Interfaces

Composite Interfaces, Vol. 11, No. 8-9, pp. 567586 (2005) VSP 2005.Also available online - www.vsppub.com

Review

Carbon nanotube-polymer nanocomposites:The role of interfaces

C. VELASCO-SANTOS 1,2,3, A. L. MARTINEZ-HERNANDEZ1,2,3and V. M. CASTANO 1,1 Centro de Fsica Aplicada y Tecnologa Avanzada, Universidad Nacional Autonoma de Mexico,

A. P. 1-1010 Quertaro, Quertaro 76000, Mexico2 Department of Materials Science, University of North Texas, Denton, TX 76203-5310, USA3 Departamento de Materiales, Departamento de Mecatronica, Instituto Tecnolgico de Quertaro,

Av. Tecnolgico, Col. Centro, Santiago de Quertaro, Quertaro 76000, Mexico

Received 25 June 2004; accepted 14 September 2004

AbstractResearch aimed at producing new nanocomposites with improved properties has dramat-ically increased in the last decade, especially on materials tailored at a nanometric level, such asfullerenes and carbon nanotubes. The use of nanoforms as reinforcement of organic polymers hasopened the possibility of developing novel ultra-strong and conductive nanocomposites. Neverthe-less, the challenge of manufacturing multifunctional composite materials based on nanostructures isstill open, in particular in the details of the corresponding interfacial properties, which are particularlyrelevant in these systems. This paper reviews the main technical activities in this field, focusing on themost important parameters that influence the behavior of their interface, discussing recent advances,as well as current and future trends in research.

Keywords: Carbon nanotubes; polymer composites; interface; chemical functionalization.

1. INTRODUCTION

Carbon nanotubes (CNs) are among the most interesting materials discoveredduring the last twenty years; these materials have diverse arrangements at ananometric level that lead to different properties depending on the specific kindof nanotubes. In fact, a number of different types of nanotubes, from single-walled nanotubes (SWNTs), double-walled nanotubes (DWNTs), and multiwallednanotubes (MWNTs) to their variants with helix and bamboo shapes are alreadyknown to this date.

To whom correspondence should be addressed. E-mail: [email protected]

568 C. Velasco-Santos et al.

Also, a number of methods have been reported so far for producing carbonnanotubes: arc discharge [1, 2], pyrolysis of hydrocarbons over catalysts [3], laservaporization of graphite [4], electrolysis of metal salts with graphite electrodes[5], and hydrothermal methods [6]; also, nanotubes have been found recently tobe produced under relatively mild natural environments [7].

CNs display different properties that suggest potential for use in various advancedtechnological applications, such as tips for Atomic Force Microscopy (AFM) [8],cells for hydrogen storage [911], nanotransistors [12, 13], electrodes for electro-chemical applications [14, 15], sensors of biological molecules [16], catalysts [17],etc. In particular, one possible application that has attracted a great deal of attentionin the materials engineering community is the incorporation of CNs as reinforce-ment of composite materials, mainly within an polymeric matrix. These materialsare expected to exhibit outstanding mechanical properties, such as extremely highYoungs modulus, stiffness and flexibility, as has been demonstrated by both ex-perimental data and theoretical models [1821]. In addition, the unique electronicproperties of nanotubes suggest possibililities for use as either semiconductor ormetallic conductive nanomaterials [22]. Also, these structures possess high thermalstability [23], which could be advantageous for aerospace applications. Addition-ally, the nearly perfect structure of CNs, their small diameter, and their high surfacearea and high aspect ratio, provide an amazing inorganic structure with unique prop-erties extremely attractive to reinforcing organic polymers.

However, in spite of the fact that CNs were originally discovered in 1991 andthat the first report on their amazing mechanical and electronic properties appearedin 1996 [18, 22], the idea of the using CNs to produce new composite materialshas only just begun to be explored in the past five years, when different papershave reported the incorporation of CNs in various polymer matrices, aiming tosynergetically increase diverse properties of the material. Results so far have beencontrasting, since several types of CNs have been used: some composites were madewith SWNTs, others by using MWNTs, and the difference in the number of wallscauses important variations on the diameter of the nanotubes, which in turn producesdifferent characteristics in the composites. It has also been found that the specificroute used to synthesize CNs yields different graphitization degrees and, therefore,diverse properties of the final material. Various polymers have been used along withdifferent processing methods and, recently, some groups have used functionalizedcarbon nanotubes (f-CNs) to improve interfacial interaction between the inorganicfiller and the organic matrix. In general, the field of developing strong, conductiveand smart materials based on CNs is just beginning and promises to become a veryimportant area of R&D.

Accordingly, in this review we have compiled and discussed the leading resultsavailable in this emerging field of materials science; with special emphasis on therole of interfaces, aiming not only to provide an insight into the field, but also to tryto foresee the main areas of activity in the coming years.

Carbon nanotube-polymer nanocomposites: The role of interfaces 569

2. PROPERTIES OF CARBON NANOTUBES

Carbon nanotubes possess many features that make them unique materials: forexample, although standard carbon fibers have long been utilized in industrialcomposites as reinforcements for polymers that produce excellent properties, suchas low density with high specific strength and specific modulus, CNs have superiorproperties as compared to any carbon fiber, in spite of being chemically equivalent.

The very first paper reporting the Youngs modulus of CNs-based composites waspublished by Treacy et al. [18], where a figure of 1.8 TPa was found as an average,and these authors mentioned there that Youngs modulus in SWNTs can reachvalues as high as 5 TPa. Recently, Yu et al. [21] found smaller Youngs modulus forMWNTs. However, the results varied depending on the specific approach used toperform the mechanical probes and also on the particular type of NTs. For instance,Wong et al. [24] calculated the Youngs modulus through atomic force microscopy(AFM) and found a value of 1.28 TPa for MWNTs. The method was applied tosilicon carbide nanorods which showed a Youngs modulus of around half of that ofMWNTs. There, the modulus was calculated from bending force vs. displacementcurves.

Other studies have evaluated the mechanical properties of carbon nanotubesproduced by arc-discharge and catalytic methods, since these two kinds of CNspresent different advantages over each other. For example, catalytic CNs canbe produced in aligned patterns, which produces useful results like well-definedarrangements and uniform electrical properties in the composites. In addition,the principle used promises to become an important tool to produce CNs in largequantities. However, the degree of graphitization in these structures is poor and thisis detrimental to the mechanical and electronic properties.

On the other hand, CNs produced by arc-discharge methods show a high degreeof graphitization and excellent mechanical and electrical properties, which areideal when these nanomaterials are used in polymeric matrix composites. Thedisadvantage is that they have a high content of impurities, which limits theirpossible uses. The arc-discharge approach to produce SWNTs requires catalyzedreactions, which induce defects in some regions where the bonds between the carbonmaterial and the metallic element of the catalytic particle are formed, in the sameway as those produced by catalytic methods. However, SWNTs produced by thismethod are the strongest fibers known to date, though the SWNT bundles formed bythis process could cause some problems in the dispersion of nanotubes in polymercomposites.

In this context, Salvetat et al. [25] evaluated the mechanical properties ofarc-discharge MWNTs (arc-MWNTs) and grown catalytic MWNTs (cat-MWNTs)using AFM. The results show an important range in the elastic modulus betweentwo forms of CNs: the average elastic modulus (E) for arc-MWNTs has a valueof 0.87 TPa while E for cat-MWNTs is around 0.027 TPa [25]. However, anotherrecent report indicates that the tensile strengths of these CNs might be relativelyhigh, inasmuch as the defects could assist in stress transfer between layers [26].


Another team has measured the elastic modulus by nanoindentation in vertically-aligned multiwalled carbon nanotubes produced by catalytic method with PlasmaEnhanced Chemical Vapor Deposition (PECVD) [27]. The parameters evaluated inthese CNs are: effective bending modulus (Eb), axial modulus (Ea) and nanotubewall modulus (Ew). The mechanical results for these catalytic CNs are morepromising than others with the aim of producing polymer composites, since Ebvaries from 0.9 TPa to 1.24 TPa, Ea is in the range of 0.9 to 1.23 TPa and Ew hasa value between 4.14 and 5.61 TPa. The latter figure is in agreement with thoseobtained by theoretical calculations with the local density approximation model[28] and with the finite element approach [29], which report effective moduli of4.7 TPa and 4.84 TPa, respectively. Other important reports have obtained differentresults in simulations with respect to the dependence of the elastic properties ofnanotubes; Lu found that these properties in CNs are insensible to helicity, numberof walls and size, with an elastic modulus around 1 TPa for all nanotubes with aradius larger than 1 nm [30]. However, other research with similar calculations hasfound that elastic modulus is dependent on the structure and tube diameter with avalue for E of 1.24 TPa [31]. As mentioned before, the vast majority of theoreticaland experimental results agree that CNs produce the strongest fibers in the world.

Table 1.Comparison of diameter, mechanical and electrical conductive properties in different carbon materials

Carbon fibersa Youngs modulus Electrical resistivity Diameter(precursor) (GPa) (104 cm) (nm)Rayon 390 10 70009000Polyacrylonitrileb 230 18 50007000L.T.c pitch mesophase 41 100 11 000H.T.d pitch mesophase 41 50 11 000Fibras tipo carburo 340 9 5000resina L.T.cFibras tipo carburo 690 1.8 5000resina H.T.dGraphitee 1000 0.4 hCarbon nanotubes De 800 a 5000f 0.05 a 2g 0.490

0.05 a 1.2ga Reference [32].b In the reference [33] are cited different carbon fibers that were developed using polyacrylonitrile

as precursor. The elastic modulus differs from 220 to 390 GPa in different commercial fibers such as:AVCO, Celanese, Courtaulds, Great Lakes, Carbon, Hercules, Union Carbide, USA, Toray.

c Low temperature.d High temperature.e The values were calculated in the sheet plane.f The majority of the elastic modulus values in different researches fluctuate between these ranges.g Values of different researches Ref. [22] and cited in the reference [34]. Note: The values fluctuate

depending on the arrangement in CNs.h Does not apply because the graphite is a sheet.


Table 1 shows a comparison of the mechanical properties and other features betweencarbon fibers, graphite and carbon nanotubes of different sources.

There is no doubt that another important reason to incorporate CNs in nanocom-posites is the electronic arrangement of these materials, inasmuch as different workshave shown that CNs can be metallic or semiconducting depending on their diam-eter and structure [3537] and another report has shown that the properties changein each tube depending on its structure [22]. Thus, the electrical conduction range israther broad, which diversifies potential uses of CNs. Table 1 also shows a compar-ison of electrical properties of CNs with other carbon materials. These propertiescould be very useful in order to use CNs in multifunctional composites. However,other features need to be taken into account if CNs are to be effectively incorpo-rated in polymer composites, since CNs must take part in the load transfer througha good interfacial interaction, to truly take advantage of their excellent mechanicalproperties.

It is expected that a deformation in the CNs could modify their electrical prop-erties in some parts of the nanotube and change the behavior of the composite.Accordingly, the effect of the deformation in SWNTs has been calculated, wherethe electronic structure, and therefore electrical conduction, is affected when thenanotubes are deformed [38]. The authors found that the conducting SWNTs maybecome semiconducting once they were mechanically deformed, but semiconduct-ing SWNTs will never become conducting when they are mechanically deformed.This could be useful to build novel smart polymeric composite materials, since adifferent electrical response could be expected depending on the mechanical defor-mation.

Other studies of SWNTs, produced by approaches such as laser ablation andarc-discharge, were aimed to form ropes and bundles [39], and some theoreticalresearches have studied the mechanical properties of these bundles in differentconditions, with the object of predicting mechanical and electrical properties ofthese aggregates of CN groups. These models could be useful in predictingproperties and employing CNs in composites and other applications [40, 41].

In addition to the relevant CN properties discussed above, there exist otherimportant features that position CNs well above any other existing reinforcementfibers for polymer composites. Indeed, in nanotubes, the nanometric scale diameteris a key factor in improving the interface, since their very small size producesclose interaction at the molecular scale, useful in minimizing the zones withoutcontact at the interface, enhancing the interactions and thus providing compatibilitybetween two quite different materials: inorganic CNs and organic polymer matrix.CN diameter plays an important role in providing a near-molecular interaction, andthe corresponding high aspect ratio is key to reinforce composites.

Another important characteristic of CNs becomes evident as they are first de-formed and then the stress released: they recover their original shape. This flexibil-ity gives nanotubes unique opportunities as reinforcements, for neither carbon fibersnor other known reinforcements exhibit this property [26]. It is known that the syn-


thetic fibers used nowadays are more sensitive to fracture when they are deformed.In addition, the mechanical deflections in CNs can be electrically induced [42].

All the properties described above make CNs exciting materials to developcompletely new polymer composite materials. However, the development ofthose novel composites still has important barriers to overcome, mainly relatedto the understanding of the phenomena at the interface between CNs and organicpolymers. The path to develop these new materials began only a few years ago,and the results available at this time are scattered and partial and sometimesmisleading, for some studies promise that these materials are ready for technologicalapplications, while other reports reveal still a poor adhesion and poorly performingcomposites. Thus, it is extremely important to analyze and discuss the resultsreported so far in the light of the underlying fundamental scientific problem: theinterfaces.

3. CARBON NANOTUBE POLYMER NANOCOMPOSITE

Different approaches have been used to produce carbon nanotube polymer nanocom-posites (CNPN) from melt blend [43], dissolving and casting [44], in situ polymer-ization [45] and extrusion [46] among others. Nevertheless, the first studies thatincorporated nanotubes in polymers were developed to align and to observe thepossible interaction of CNs with the polymer. The first report where CNs wereincorporated in polymer dates back ten years. In this study Ajayan et al. [47]aligned CNs in an epoxy matrix, then cut small slides to obtain a partially alignedsegment. Aligned carbon nanotubes in polymer matrices can be considered as adesigned arrangement, inasmuch as this would produce possibilities of control ofthe properties along the direction of the carbon nanotubes in the nanocomposites.This first evidence of aligned CNs in a polymer matrix applies only to small areasof these nanocomposites. Jin et al. [48] report alignment of CNs in polymer ma-trix by mechanical stretching. They used MWNTs synthesized by the arc-dischargemethod and the thermoplastic polymer polyhydroxyaminoether that was dissolvedin chloroform and later molded in Teflon. The results show a partial alignment thatdepends on the stretching ratios and on the carbon nanotube fraction.

In another report, SWNTs are aligned by a combination of solvent casting andmelt processing methods: there the CNs were dispersed in polymethylmethacrylate(PMMA) using dimethylformamide as solvent. Alignment in composites provideshigher conductivity along the flow direction. Elastic modulus (E) and yieldstrength (Y ) were measured in melt-spun fibers: both E and Y increase withnanotube content and draw ratio. Although these increases are not uniform, in bothparameters higher values are reached with 5 wt% and 8 wt% of SWNTs [49].

Wagner et al. developed an interesting study where stress transfer ability wasshown in carbon nanotube polymer nanocomposite. In this report, the fragmentationof MWNTs under tensile stresses in thin films was observed. The results showthat the efficiency to transfer stress in this composite is at least one order of


magnitude larger than in conventional fiber composites [50]. A later report byLourie and Wagner indicates that the polymer nanotube interface is not inert andload is transferred to the nanotubes via the surrounding matrix. In this paper, itwas suggested that, in spite of the huge difference in scale between traditional fibercomposites and carbon nanotube nanocomposites, the fundamental concepts of themechanics of conventional composites could hold at the nanometric level [51].

Deformation studies by transmission electron microscopy (TEM) of polymerCN composites were developed with a nanocomposite containing MWNTs in apolystyrene matrix. The studies were realized using a TEM strain holder and a filmwith a notch. The results show that nanotubes tend to align and bridge the crackwhen a tensile stress was applied to the composite, showing load transfer across thenanotubepolystyrene interface [52].

Other PMMA-based studies have explored other parameters, such as CN distrib-ution and the possible modification of CNs as they are incorporated in a polymermatrix. The studies were realized with SWNTs obtained from arc-discharge methodand not purified. The investigation shows that low concentrations of CNs in polymergive a uniform distribution with this kind of CN [53, 54].

Although some papers claim a good distribution and load transfer at the interface,this should be reflected in the properties of nanocomposites. However, contradictoryresults have been obtained in this kind of composite, mainly with respect tothe mechanical properties, showing that many factors affect the production, theinterface and, therefore, the successful development of carbon nanotube polymernanocomposites.

Schadler et al. developed MWNTs-epoxy composite cured with triethylene tetra-amine. They evaluated both tensile and compression strength. The results show thatcompression modulus is higher than tensile modulus in this composite. The loadtransfer was evaluated through Raman spectroscopy, and it was found that strain incarbon bonds only shifts significantly under compression. The authors speculatethat all the walls in MWNTs only participate in compression, unlike the case whenthe composite is subjected to tensile stress, where the outer walls participate [55].

Shaffer and Windle formed CN-polyvinyl-alcohol nanocomposite with differentloads of catalytically grown nanotubes: the content of CN was varied from 10 wt%to 60 wt% and dynamical mechanical parameters were measured. The values ofstorage modulus (E) increased over two-fold, but only by using a huge quantityof CN. At the same time, the onset of thermal degradation was retarded in thiscomposite. Therefore, these authors suggest the application of CNs as modifiers.Also, the conductive properties in composites were measured, and a percolationthreshold was found at a concentration between 5 and 10% of CN [56]. In otherstudy, MWNT-PMMA composites were developed and tested mechanically. Twotypes of MWNTs were incorporated in the polymer matrix: as-synthesized andtreated MWNTs (t-MWNTs: CNs ground in a high rate ball mill for 20 min andthen boiled for 0.5 h in nitric acid); the method used to form the compositeswas polymerization using azoiso-butyronitrile (AIBN) as initiator. Satisfactory


increases in the tensile strength and hardness were reached with 3 wt% and 5 wt%of t-MWNTs; however, with higher loads the tensile strength decreased. The resultsare attributed by the authors to the good interaction between CNs and polymermatrix, due to the fact that the initiator could break the bonds of CNs and, inthis way, it is possible to form a chemically bonded interface. However, manyother factors could also be responsible for this success: for example, the fact thatt-MWNTs were treated with nitric acid solution could cause a slight oxidationin some zones of the nanotubes, where the defects might be increased by theinitiator action, producing functionalized nanotubes that would interact with thepolymer chains. More research is needed to confirm this and to understand the roleplayed by the polymeric initiator in CN bonds and on the development of nano-composites [57].

PMMA is one of the most studied polymers for these composites. Cooper et al.evaluated mechanical properties of nanocomposites formed with MWNTs, SWNTsand nanofibrils. The results of these analyses present a tensile modulus almostinsensitive to the presence of CNs; however, the impact properties are significantlyimproved by both MWNTs and SNWTs [58]. Tatro et al., in another recentstudy with PMMA and MWNTs, have evaluated thermal and electrical properties,Vickers microhardness and dynamical mechanical analysis in nanocompositesirradiated with a Cesium-137 source, with the aim of determining the effect ofcarbon nanotubes when the sample is irradiated. The results show that MWNTsmay improve radiation hardness of mechanical properties and glass transitiontemperature when a composite is formed. The effect is attributed to conjugatedbonds in CNs, which absorb some part of the radiation energy, limiting the damageof PMMA molecules. More aging studies with radiation and evaluation of theinterface are proposed and currently being developed by these authors [59].

Another recent report evaluates the dynamical mechanical properties of purePMMA and nanocomposites formed with SWNTs and PMMA matrix at twodifferent frequencies. The study was developed with the aim of understanding theeffect produced by the incorporation of the nanotubes on the ability of PMMAto store and dissipate the mechanical energy. SWNTs used in this researchwere produced by laser ablation method and the nanocomposites were preparedby in situ polymerization using a solution of dimethylformamide (DMF) and2,2-azobisisobutyronitrile (AIBN) as initiator. -Relaxation, related with theonset of the glass transition temperature (Tg) and -transition associated withhindered rotation of the side chain, were evaluated in PMMA and SWNT-PMMAnanocomposites. Transitions were detected in both tan curves and elastic moduluscurves (E) at low and high temperatures. Results at two different frequenciesshow that PMMA and nanocomposites formed with slight quantity of SWNTsexhibit similar -transition temperature and identical Tgs. However, E at lowtemperature (150C) in the nanocomposite with a tiny quantity (0.014 wt%) ofSWNTs exhibits an important increase in comparison with E of the pure PMMA,from 5.1 GPa in PMMA to 7.2 GPa for the nanocomposite, both evaluated with a


frequency of 100 rad/s. This represents a significant increase, considering the smallquantity of SWNTs used.

Thermogravimetric analysis (TGA) and nuclear magnetic resonance (NMR)studies have been used to evaluate possible changes in the structure of the polymernanocomposite caused by the addition of SWNTs. The authors cite that neithermicrostructure nor thermal degradation changes were localized in nanocomposites,as compared to pure PMMA. This indicates that these parameters could not beliable for the change in the mechanical properties of the nanocomposites. However,the interface in these nanocomposites plays an important role and the propertiesof the polymer changed in the vicinity of the SWNTs, probably by cohesiveinteractions produced by the large surface area of CNs [60]. This researchalso involved discussion and possible adaptation of the traditional theory of CNnanocomposites; the authors found that nanotubes have a greater effect on theproperties of composites than those expected from traditional composite theory [60].Therefore, new models will be needed to adjust and predict CN nanocompositeproperties because the interface produced with polymers and nanomaterials isgenerating unexpected properties.

Dufresne et al. processed and characterized nanocomposites that were developedwith catalyzed MWNTs and poly(styrene-co-butyl acrylate) [61]. The compositeswere prepared by casting, after stirring with different loads of CNs from 0 to15 wt%. Mechanical properties of these composites were evaluated by dynamicalmechanical analysis (DMA) and tensile test: increases in the tensile modulus withthe content of CN were found in almost all samples. However, an unexpecteddecrease in the modulus was observed in the sample with 7 wt% of CNs. Althoughgood distribution was achieved in most samples, some areas with nanotubesappeared like clusters causing unpredictable behavior in those samples. This hasbeen observed in other composites prepared with CNs when the content is higherthan 5 or 7 wt% and explains the unexpected behavior in this particular study.

Wong et al. have analyzed two kinds of nanocomposite that show agglomeratesthat cause similar effects to those just mentioned. Although it was shown by fieldemission scanning electron microscopy (FESEM) and TEM that CNs have goodinteraction and adhesion with the polymers, other regions show agglomerates ratherthan even distribution and cause a deterioration in the nanocomposite properties. Inthis research, CNs reinforcing polystyrene rod and epoxy thin film were analyzed.In the case of CNpolystyrene composites, tensile strength increases from 22.1 MPa(for polystyrene) to 24.4 MPa (for the sample with 0.1 wt% of CN); however, whenthe content of CNs was increased at 5 wt% the strength was diminished to 17.9 MPa.Furthermore, the samples with 1 wt% and 2 wt% CN showed lower values in tensilestrength than the sample of pure polystyrene [62].

Also in this study, epoxy thin films were evaluated mechanically. Microscopyresults of these samples revealed the two mentioned regions. One corresponds tothe failure of the matrix but not at the CN-epoxy interface. Nevertheless, in otherregions, CN agglomerates were observed showing poor dispersion locally.


Finally, the paper published by Wong et al. gives an interesting analysis involvingmolecular simulations of CN pullout. Molecular models of SWNTs in polystyrenematrix and SWNTs in epoxy matrix were generated, and then total work requiredfor fiber pullout was calculated. The results obtained for CN-epoxy are very similarto those obtained in experimental pullout for the same system. The authors concludethat the results of this simulation and the previous experimental studies suggest thatthe interfacial shear stress of CNpolymer systems is at least one order of magnitudehigher than micro fiber reinforced composites and it is assumed that the surface ofthe CN is important to achieve intimate contact at nanometric level between CNsand polymer matrix [62]. Thus, these results show that a strong interface can beachieved between inorganic CNs and organic matrix when the polymer interactswith nanotubes. However, once more, large quantities cause problems by formingagglomerates.

An opposite effect to those discussed above has been reported in CN nanocom-posites when the hardness has been measured; here, the increase in the quantityof nanotubes incorporated in the polymer has a favorable effect. The addition of2 wt% of CNs improves the hardness of an epoxy matrix by some 20%; however,when quantities higher than 1.5 wt% were used, the hardness diminished as com-pared to the hardness of pure epoxy. The authors have speculated that this effectis caused by a weak interface between CNs and the polymer. On the other hand,a favorable effect on the hardness with 2 wt% of CNs is attributed to the mesh-like structures formed when the concentration of CNs is increased. The authorsmentioned that the formation of these structures is caused by the high aspect ratioof the nanotubes, which produces a tangle arrangement that enhances the scratchresistance. In this report, different temperatures were used to treat CN nanocom-posites and to test their failure after flexural strength in diverse conditions; sinceCNs have been proposed for new composites for the space industry these variable-temperature states are needed to simulate this environment. Results indicate thatthe mesh-like structures mentioned do not improve the flexure strength under thedifferent conditions studied. The authors conclude that, in this case, the cause ofthe reduction could be due to structural non-homogeneity or the presence of a weakbonding interface [63].

Epoxy matrix is another polymer used in many studies in CN nanocomposites,showing excellent electrical properties with small amounts of carbon nanotubes.Allaoui et al. report a percolation threshold between 0.5 wt% and 1 wt% of carbonnanotubes whereas the tensile strength improved significantly with 1 wt% of CNs inyield point and in 10% in strain [64]. However, only a small increase was observedwhen 4 wt% of CNs was added. As we mentioned, this effect has been observedby other researchers, with the addition of unpurified CNs. In fact, the majority ofthe results demonstrate that the spatial distribution is an open question and that it ispossible that the inhomogeneity of the samples has a strong influence, as importantas the type of polymer used [44].

Carbon nanotube-polymer nanocomposites: The role of interfaces 577Table 2.Comparative results (E, % increase in E) of Carbon nanotubes composites obtained in differentpolymers matrices at 40C 1 Hz of frequency [44]a

Polymer E (40C, 1 Hz) E (40C) CNs wt%d Increase % Ee Reference(matrix) MPab CN composite

MPacPMMA =800 =1600 26 100 [43]PS =2400 =3500 5 44 [65]PSBA =0.681 =1.584 7 132 [61]PVA =5000 =11 200 60 124 [56]MEMA 708 2340 1 230 [44]

a Modified table of the reference [44].b Storage modulus at 40C of the polymer matrix.c Storage modulus at 40C of the carbon nanotube composite.d Weight percent of carbon nanotubes in the composite.e Increase percent in storage modulus of composite with respect to polymer matrix.

The alignment of carbon nanotubes in nanocomposites has an important influenceon the electrical properties. Mechanical properties also change when CNs arealigned. Thostenson and Chou found relevant improvements in tensile stressproperties and toughness when 5 wt% of CNs were incorporated into a polystyrenematrix, up to five times with respect to the randomly oriented composite, when thesame quantities of CNs were aligned by extruder [65].

Other important topic that helps to enhance the compatibility at the interfacelevel between CNs and an organic polymer matrix is the incorporation of additivesand other polymers that could assist in the dispersion and compatibility. In thiscontext, different behaviors have been observed when an additive is incorporatedinto nanocomposites. Gong et al. [66] incorporated nonionic surfactants inthermosetting polymer to improve the interfacial interaction between nanotubes andthe polymer. The results show higher increases in storage modulus (E) and in glasstransition temperature when both surfactant and CNs were used, with respect tothe composite where only carbon nanotubes were incorporated. In both cases only1 wt% of CNs was employed; however, E with surfactant is increased by more than30% and glass transition increases by 25C. In contrast, the addition of CNs withoutsurfactant produced only a moderate increase in the temperature and mechanicalproperties. In other reports, better results have been reached without additives,providing great care is taken during mixing [44]. A non-ionic surfactant was alsoused but with thermoplastic copolymer, methylethylmethacrylate (MEMA). Thisstudy is one of the first that compares their own results with other researchers whereboth only carbon nanotubes and additives are used. An important increase in E bymore than 200% at 40C is reached with only 1 wt% of CNs, without additives. Thecomparative results presented involving different groups are shown in Table 2, forCN nanocomposites where additives were used, and in Table 3 for studies whereonly CNs were used. The authors suppose that, although the non-ionic surfactantutilized is a good dispersant for carbon, the wetting could occur with the impurities


Table 3.Comparative results of DMA on carbon nanotubes composites using a surfactant in two differentpolymersa

Sample Epoxy (66) MEMA (44)E40C Tg C (tan )c E 40C Tg C (tan )c(MPa)b (MPa)b

Polymer 1390 63 708 92Polymer + Surf 1150 62 774 89Polymer 1 wt% CN 1550 72 2340 99Polymer 1 wt% CN, 1750 88 1983 1021 wt% surf

a Modified table of the reference [44].b Storage modulus at 40C.c Glass transition temperature (tan ).

in the CNs surroundings and an interface is created with only with some part ofthe CNs. This would produce interfaces that are not so effective, since they form alittle barrier that produces different conductivity behavior [67]. The results of theconductive behavior in these films are currently in preparation.

Reports where other substances take part in the interfacial and dispersion be-havior of CN nanocomposites were published by Jin et al. [68]. In this study,MWNTs were sonicated in a dimethylformamide solution of poly(vinylidene flu-oride) (PVDF). The mix was melt blended with PMMA to form a nanocompos-ite. The results show that there is a threshold in the addition of PVDF, since slightamounts of the PVDF improve the storage modulus, whereas higher PVDF contentslower the storage modulus.

4. CHEMICAL FUNCTIONALIZATION OF CARBON NANOTUBES

The results analyzed in the previous section show that CNs obtained from methodsthat include impurities together with the CNs lead to unpredictable properties, eventhough the composites in some cases show an important increase in mechanicalproperties below a content of 5% of CNs. Other papers report a poor interface dueto poor wettability [69], which could be influenced by the impurities. Accordingly,different methods have been proposed to eliminate impurities in CNs. The mostimportant impact has been produced by oxidation methods which, in addition toreducing impurities, cause chemical modifications of CNs. This has proven usefulfor composites and other applications of these materials. This section describessome important studies in the field of chemical modification.

The oxidation technique has opened a new gate in this field, by producingcarboxyl and carboxylate groups on the tips and surfaces of CNs. These moietieshave been used to attach different chains in CNs that allow different possible usesfor these materials [70]. Oxidation approaches were developed few years afterCNs were discovered. The theoretical basis is the different resistance to oxidation


(a)

(b)Figure 1. A schematic representation (a) carbon nanotube (b) oxidized and functionalized carbonnanotube.

between CNs and other nanoparticles formed during the production. Thus, it wasobserved that the nanotubes when oxidized are consumed from the tips inward atthe same time [71] and the surface wall is modified if the oxidation is severe.

Although the first oxidations were developed with the aim of opening andfilling CNs, the carboxyl, hydroxyl and carboxylate groups generated through theoxidation on the surfaces of these materials are useful moieties in order to bondnew reactive chains that improve solubility, processability and compatibility withother materials and, therefore, improve the interfacial interactions of CNs with othersubstances. This route has become one of the most important: it is the focus of manycurrent researches because of the outstanding results it has produced [72].

The defects caused by oxidation have been studied and it has been foundthat nanotubes tolerate a limited quantity of them, before a macroscopic sampleloses its special mechanical and electronic properties [7376]. This kind offunctionalization is currently being researched in order to understand and to improvethe utilization of these chemical modifications. The organic carboxyl groupsformed in the nanotube surface are localized on the defects on functionalizedsingle-walled and multi-walled nanotubes (f-SWNTs and f-MWNTs, respectively);suitable reactive organic groups for other chemical chains tend to react in this zone.An idealize scheme of unfunctionalized and functionalized CNs with these organicmoieties is shown in Fig. 1.

Next, different reports have focused on the chemical functionalization afteroxidation. In these papers, the COOH groups generated in the oxidation processare used to attach different molecules useful to improve surface compatibility ofCNs with other materials. The relevance of these groups to join CNs with organicpolymer matrices is also mentioned.


One of the routes used as complementary procedure to insert long organic chain isthe use of SOCl2 (thionyl chloride) in dimethyl formamide. In this way, a long chainof alkylamine (octadecylamine) has been coupled to activated f-SWNTs [77]. Thismethod is very useful to insert different chains. The approach also has been usedto enhance the solubility properties of SWNTs with the mentioned chains and othermoieties as alkyl-aryl amine and 4-dodecyl-aniline [78]. The same method with asubsequent reaction has been used to add other organic moieties. For instance, ithas been followed by a reaction with monoamine terminated poly(ethylene oxide)(PEO), which was useful to facilitate the linkage of this chain in a grafted PEO-SWNTs system [79]. Also, the system (f-SWNTS + SOCl2) has been used inMWNTs to functionalize these structures via addition of an aminopolymer, such aspoly(propionylethylemine-co-ethylenemine) (PPEI-EI) with the purpose of forminghydrophilic MWNTs.

Direct reactions have been carried out with molten octadecylamine in carboxyl-terminated SWNTs, in the range of 120130C, forming a carboxylate group inthe SWNT, which interact with an octadecylamine functionalized free radical [73].This produces greater solubility in SWNT (s-SWNTs) than those formed usingSOCl2 [77]. Other researchers have developed a wide variety of aliphatic aminescoupled to carboxyl groups in the SWNT surface [80], showing that amines ingeneral can be attached to CNs successfully. The insertion of these groups in theCN surface could be useful in improving compatibility between these nanomaterialsand aminopolymer matrices.

A comprehensive study of SWNTs and MWNTs using carboxyl organic groupstreated with SOCl2, with further functionalization through esterification reactionswith lipophilic and hydrophilic dendra, has been presented recently [81]. Defunc-tionalization reactions of these nanostructures under alkaline and acid catalyzedhydrolysis have been carried out to provide evidence of the ester linkages betweenthe carboxyl-terminated carbon nanotubes and the mentioned organic moieties [82].

Dodecylamine groups have been bonded to CNs through carboxyl moieties via theSOCl2 reaction. The functionalized carbon nanotubes have shown an improvementin the solubility with the polymer matrix and organic solvents and have beenused to develop carbon nanotube-oxotitanium phthalocyanine composite. Thematerial showed an important improvement in photosensitivity with that content offunctionalized CNs and an important five-fold increase over that with oxotitaniumphthlocyanine alone [83].

The carboxylic terminated organic groups on SWNTs have been used by Quinet al. [84] in order to attach n-butylmethacrylate via atom transfer radicalpolymerization. In this reaction, the polymerization is controlled and mightbe useful to graft copolymer blocks and other polymers, thus improving thecompatibility between nanotubes and synthetic macromolecules.

A recent research study promotes the insertion of many functional groups onthe CN surface. Using the same reaction, different organosilanes were added toMWNT via carboxyl groups on modified CNs. The method allows diversifica-


tion of CN properties depending on the terminal chain of the coupled organosi-lane. The oxidation and silanization achieved in this research yield functionalizedmoieties, which contain terminated organosilane groups designated as R groupsthat are bonded to the nanotubes by silanol groups. This has allowed modifica-tion of organic-terminated CNs with many different chains, chemically-bonded toMWNTs, with the silanization reaction, where the R group (methacryl, glycidoxy,etc.) in the organofunctional silanes can be changed depending on the specific poly-meric matrix employed. The behavior of these MWNTs changes as the R terminalchanges, so it has been found that 3-methacryl-trimethoxysilane (3-MAT) producessoluble MWNTs in organic solvent such as acetone or ethanol [67, 85], since themethacryl chain attached as R group on the MWNTs surface changes the chemi-cal character of this material, increasing their solubility. This does not occur when3-methacryl-mercaptosilane (3-MPT) is attached in MWNT, for the terminal thiolgroup in the chain attached to CNs produces different properties on the MWNTsurface. Figure 2 shows a scheme of the reaction proposed in these studies.

Chemical functionalization has reached an important position in the CN field, asdifferent chemical processes have been developed to diversify CN properties. Themodifications described above represent only an intermediate stage in the route toimprove the interactions between nanotubes and organic polymer matrices. Indeed,other chemical reactions have already been explored to couple these materials toproteins [86] and other substances. Also, the carboxyl groups in the CN surfacehave been shown to improve the compatibility of CNs with biomolecules. This hasbeen employed for nanotechnology applications [87] and might be an important toolto develop good compatibility between organic biopolymers and CNs, which couldfind applications in biomaterials and biomedicine.

Figure 2. Sequence of silanization reactions on nanotubes surfaces [72].


5. FUNCTIONALIZED CARBON NANOTUBE-POLYMER NANOCOMPOSITES

Chemical modification is not a new tool to enhance the compatibility between areinforcement and a matrix, since chemical attack and grafting have long been usedin natural fibers, synthetic fibers and particles [8890]. However, in the field ofthe nanocomposites, chemical functionalization offers an important tool in order toimprove the interface and achieve the amazing properties that carbon nanotubescan provide. Although different groups are very active in this field, interfacialinteractions still need to be fully understood to improve the interface between carbonnanotubes and polymer matrix. The recent results in chemical modification ofcarbon nanotubes are promising and could be the way to form strong and functionalnew advanced materials. Some results obtained in the area of functionalized carbonnanotube nanocomposites will be reviewed in this section.

It was described above how the carboxyl groups found in the tip and surfaceof CNs are useful for the interaction between CNs and other compounds. Thus,the use of these moieties to improve the link in CNpolymer composites havebeen proposed, either by inserting other chemical groups or by using the carboxylgroups produced during the oxidation [85, 91]. As mentioned before, Jia et al.[57] suggested that the initiator opens the bonds found in CNs and, in this way,CNs take part in the polymerization. However, although opening the bonds inCNs would allow them to join to other chemical groups, oxidation provides morepossibilities to bond the nanotubes to the matrix, due to reactive chemical groupssuch as COOH, COO and C O that are found on the tip and on the wall surface.In another report Geng et al. [92] use fluorinated single-walled nanotubes (fl-SWNTs) to enhance the uniformity and the nanotube dispersion using poly(ethyleneoxide) (PEO) as matrix, and dissolving the nanotubes in methanol. Nevertheless, arecent report [45] proposes that the best moment to achieve interactions between thefunctional groups (found in tips and walls in CNs) and polymer chains is definitelywhen the polymer is created by in situ processing, inasmuch as the free radicalformed in monomer molecules by the initiator could either interact or react with theCN moieties more easily than when the polymer is made and after it is dissolvedor melted to produce the composites. In that research, methyl methacrylatemonomer (MMA), 2-2 azobisisobutyronitrile (AIBN) and oxidized f-MWNT wereused, and CN composites were produced by in situ polymerization using AIBNas initiator. The AIBN quantity, reaction time and temperature were controlled toyield uniform molecular weights in all samples. Three samples in this research weremanufactured with only polymer (poly(methyl methacrylate PMMA)): polymerwith 1 wt% of unfunctionalized MWNTs (u-MWNTs), and polymer with 1 wt%of functionalized MWNTs (f-MWNTs), additional composite with polymer and1.5 wt% of f-MWNTs was also produced to observe the behavior when the additionof f-MWNTs was increased slightly. Thee interaction of PMMA with f-MWNTswas analyzed by infrared and Raman spectroscopy. In addition, mechanical resultsshow an important increase in E at 40C by 66% and 88%, with 1 wt% and1.5 wt% of f-MWNTs, respectively, and both samples increase the glass transition


Figure 3. SEM image after tensile test of f-MNWT composite where it is possible to observe thedispersion of the f-MNWTs and the wetting of f-MNWTs with the polymer. Copyright AmericanChemical Society [45].

temperature Tg by some 40C. This is different from the case with unfunctionalizednanotubes, where E at 40C increased by 50% and the Tg by only 6C. Anotherimportant fact is that the performance of f-MWNT composites increases more than11-fold, at relatively high temperature. A comparison of different properties ofthis and other f-CN composites is presented in Ref. [72]. A scanning electronmicrograph (SEM) of functionalized carbon nanotube nanocomposites is presentedin Fig. 3.

Other recent papers have shown that carboxyl-terminated carbon nanotubestreated with amines and incorporated into epoxy matrix have better interactionthan those that were not treated chemically, showing an important improvementin the interface links between these two materials. In this study, TEM pullout testswere carried out, showing a better interaction due to chemical groups on the CNsurface [93]. Other authors have develop in situ functionalized MWNTs in phenoxycomposites by melt mixing with 1-(aminopropyl)imidazole [94]. In this research,the composites with more than 4.8 wt% show higher storage modulus than thepolymer matrix. In this research the modulus at high temperature is diminishedwith the incorporation of in situ functionalized nanotubes.

6. FUTURE PROSPECTS IN CARBON NANOTUBE-POLYMERNANOCOMPOSITES

CNs promise to open up a new age of advanced multifunctional materials; however,the wide variety of results available today demonstrate that the addition of largequantities of CN in polymer-based composites is not favorable in all cases, andthe reason for this behavior is not completely understood yet. Several of the


analyzed reports agree that CNs in large amounts form clusters. This diminishes theinteraction and, therefore, the interface is not optimized. Nevertheless, when smallquantities of nanotubes are incorporated into the polymer, the electrical, optical andmechanical properties improve significantly, reaching important values unseen inother materials.

The remarkable properties obtained when f-CNs are incorporated into polymericcomposites represent a promising route to design ideal materials for aerospace-related structural applications. However, the field requires much deeper fundamen-tal research.

REFERENCES

1. T. W. Ebbesen and P. M. Ajayan, Nature 358, 220 (1992).2. C. Journet, W. Maser, B. Bernier, A. Loiseau and M. de la Chapelle, Nature 388, 756 (1997).3. W. Z. Li, S. S. Xie and G. Wang, Science 274, 1701 (1996).4. A. Thess, R. Lee, P. Nokolaev, H. Dia and P. Petit, Science 273, 483 (1996).5. W. K. Hsu, M. Terrones, H. Terrones, N. Grobert, A. I. Kirkland, J. P. Hare, K. Prassides,

P. D. Townsend, H. W. Kroto and D. R. M. Walton, Chem. Phys. Lett. 284, 177 (1998).6. L. S. Wojciech, J. A. Libera, Y. Gogotsi and Y. Masahiro, J. Solid. State Chem. 160, 184 (2001).7. C. Velasco-Santos, A. L. Martnez-Hernndez, A. Consultchi, R. Rodrguez and V. M. Castao,

Chem. Phys. Lett. 373, 272 (2003).8. J. H. Hafner, C.-L. Cheung, A. T. Woolley and C. M. Lieber, Progr. Biophys. Mol. Biol. 77, 73

(2001).9. F. L. Darkrim, P. Malbrunot and G. P. Tartaglia, Int. J. Hydrogen Energy 27, 193 (2002).

10. H.-M. Cheng, Q.-H. Yang and C. Liu, Carbon 39, 1447 (2001).11. M. Hirscher, M. Becher, M. Haluka, F. Von Zeppelin, X. Chen, U. Detlaff-Weglikowska and

S. Roth, J. Alloy Comp. 356-357, 433 (2003).12. R. H. Baughman, A. A. Zhakidov and W. A. de Heer, Science 297, 787 (2002).13. J. M. Blau and A. J. Fleming, Science 304, 1457 (2004).14. M. L. Pedano and G. A. Rivas, Electrochem. Commun. 6, 10 (2004).15. J. N. Barisci, G. G. Wallace, D. R. MacFarlane and R. H. Baughman, Electrochem. Commun. 6,

22 (2004).16. Y. Lin, F. Lu, Y. Tu and Z. Ren, Nanoletters 4, 191 (2004).17. P. Serp, M. Corrias and P. Kalck, Appl. Cat. A 253, 337 (2003).18. M. M. Treacy, T. W. Ebbesen and J. M. Gibson, Nature 381, 678 (1996).19. O. Lourie and H. D. Wagner, J. Mater. Res. 13, 2418 (1998).20. O. Lourie, D. M. Cox and H. D. Wagner, Phys. Rev. Lett. 81, 1638 (1998).21. M.-F. Yu, O. Lourie, M. J. Dyer, K. Moloni, T. F. Kelly and R. S. Ruoff, Science 287, 637 (2000).22. T. W. Ebbesen, H. J. Lezec, H. Hiura, J. W. Bennett, H. F. Ghaemi and T. Thio, Nature 382, 54

(1996).23. J. Che, T. Cagin and W. A. Gooddard III, Nanotechnology 11, 65 (2002).24. E. W. Wong, P. E. Sheehan and C. M. Lieber, Science 277, 1971 (1997).25. J.-P. Salvetat, A. J. Kulik, J.-M. Bonard, G. A. Briggs, T. Stckli, K. Mtnier, S. Bonnamy,

F. Bguin, N. A. Burnham and L. Forr, Adv. Mater. 11, 161 (1999).26. P. F. J. Harris, Intern. Mater. Rev. 49, 31 (2004).27. H. J. Qi, K. B. K. Teo, K. K. S. Lau, M. C. Boyce, W. I. Milne, J. Robertson and K. K. Gleason,

J. Mech. Phys. Solids 51, 2213 (2003).28. A. Pantano, D. M. Parks and C. M. Boyce, J. Mech. Phys. Solids 52, 789 (2004).


29. Z.-C. Tu and Z.-C. Ou-Yang, Phys. Rev. B 65, 233407 (2002).30. J. P. Lu, J. Phys. Chem. Solids 58, 1649 (1997).31. E. Hernndez, C. Goze, P. Bernier and A. Rubio, Phys. Rev. Lett. 80, 4502 (1998).32. K. K. Chawla, Composite Materials: Science and Engineering. Springer-Verlag, New York

(1998).33. A. Kelly, in: Materials Science and Technology: A Comprehensive Treatment, Structure and

Properties of Composites, T. W. Chou, R. W. Cahn, P. Haasen and E. J. Kramer (Eds), pp. 3-13.VCH Weinheim, Germany (1993).

34. M. Terrones, W. K. Hsu, H. W. Kroto and D. R. M. Walton, Topics in Current Chem. 199, 189(1999).

35. K. Tanaka, M. Okada and Y. Huang, in: The Science and Technology of Carbon Nanotubes,K. Tanaka, T. Yamabe and K. Fukui (Eds). Elsevier, Oxford, UK.

36. J. W. Mintmire, B. I. Dunlap and C. T. White, Phys. Rev. Lett. 68, 631 (2002).37. R. Saito, M. Fujita, G. Dresselhaus and M. S. Dresselhaus, Phys. Rev. B 46, 1804 (1992).38. B. Liu, H. Jiang, H. T. Johnson and Y. Huang, J. Mech. Phys. Solids 52, 1 (2004).39. E. T. Thostenson, Z. Ren and T.-W. Chou, Compos. Sci. Technol. 61, 1899 (2001).40. C. Q. Ru, Phys. Rev. B 62, 10405 (2000).41. N. Popov, V. E. Van Doren and M. Balkanski, Solid State Commun. 114, 395 (2000).42. P. Poncharal, Z. L. Wang, D. Ugarte and W. A. de Heer, Science 283, 1513 (1999).43. Z. Jin, K. P. Pramoda, G. Xu and S. Hong Goh, Chem. Phys. Lett. 337, 43 (2001).44. C. Velasco-Santos, A. L. Martnez-Hernndez, F. Fisher, R. S. Rouff and V. M. Castao, J. Phys.

D. Appl. Phys. 36, 1423 (2003).45. C. Velasco-Santos, A. L. Martnez-Hernndez, F. Fisher, R. S. Ruoff and V. M. Castao, Chem.

Mat. 15, 4470 (2003).46. M. L. Shofner, F. J. Rodriguez-Macias, R. Vaidyanathan and E. V. Barrera, Comp. Part A 34,

120 (2003).47. P. M. Ajayan, O. Stephan, C. Colliex and D. Trauth, Science 265, 1212 (1994).48. L. Jin, C. Bower and O. Zhou Appl. Phys. Lett. 73, 1197 (1998).49. R. Haggenmueller, H. H. Gommans, A. G. Rinzler, J. E. Fischer and K. I. Winey, Chem. Phys.

Lett. 330, 219 (2000).50. H. D. Wagner, O. Lourie, Y. Feldman and R. Tenne, Appl. Phys. Lett. 72, 188 (1998).51. O. Lourie and H. D. Wagner, Compos. Sci. Technol. 59, 975 (1999).52. D. Qian and E. C. Dickey, J. Microsc. 204, 39 (2001).53. C. Stephan, T. P. Nguyen, M. L. De la Chapelle, S. Lefrant, C. Jounet and P. Bernier, Synth.

Methods 108, 139 (2000).54. M. L. De la Chapelle, C. Sthepan, T. P. Nguyen, S. Lefrant, C. Jounet, P. Bernier, E. Munoz,

A. Benito, W. K. Maser, M. T. Martinez, G. F. de la Fuente, T. Guillard, G. Flamant, L. Alvarezand D. Laplaze, Synth. Methods 103, 2510 (1999).

55. L. S. Schadler, S. C. Giannaris and P. M. Ajayan, Appl. Phys. Lett. 73, 3842 (1998).56. M. S. P. Shaffer and A. H. Windle, Adv. Mater. 11, 937 (1999).57. Z. Jia, Z. Wang, C. Xu, J. Liang, B. Wei, D. Wu and S. Zhu, Mater. Sci. Eng. A271, 395 (1999).58. C. A. Cooper, D. Ravich, D. Lips, J. Mayer and H. D. Wagner, Compos. Sci. Technol. 62, 1105

(2002).59. S. R. Tatro, L. M. Clayton, P. A. ORourke Muisener, A. M. Rao and J. P. Harmon, Polymer 45,

1971 (2004).60. K. W. Putz, C. A. Mitchell, R. Krishnamoorti and P. F. Green, J. Polym. Sci. B 42, 2286 (2004).61. A. Dufresne, M. Paillet, J. L. Putaux, R. Canet, F. Carmona, P. Delhaes and S. Cui, J. Mater. Sci.

Lett. 37, 3915 (2002).62. M. Wong, M. Paramsothy, X. J. Xu, Y. Ren, S. Li and K. Liao, Polymer 44, 7757 (2003).63. K.-T. Lau and S.-Q. Shi, Carbon 40, 2961 (2002).64. A. Allaoui, S. Bai, H. M. Cheng and J. B. Bai, Compos. Sci. Technol. 62, 1993 (2002).


65. E. T. Thostenson and T.-W. Chou, J. Phys. D. Appl. Phys. 35, L77 (2002).66. X. Gong, J. Liu, S. Baskaran, R. D. Voise and J. S. Young, Chem. Mater. 12, 1049 (2000).67. C. Velasco-Santos, PhD Thesis, Universidad Autnoma de Quertaro, CFATA UNAM, Mxico

(2003).68. Z. Jin, K. P. Pramoda, S. H. Goh and G. Xu, Mater. Res. Bull. 37, 271 (2002).69. J. N. Coleman, S. Curran, A. B. Dalton, A. P. Davey, B. McCarthy, W. Blau and R. C. Barklie,

Phys. Rev. B 58, R7492 (1998).70. C. Velasco-Santos, A. L. Martnez-Hernndez and V. M. Castao, in: Progress in Nanotechnol-

ogy Research, Vol. II. Nova Science Publishers (2004) (to be published).71. T. W. Ebbesen, Carbon Nanotubes Preparation and Properties. CRC Press, Boca Raton, FL

(1997).72. C. Velasco-Santos, A. L. Martnez-Hernndez and V. M. Castao, in: Progress in Nanotechnol-

ogy Research, Vol. I. Nova Science Publishers (2004) (in press).73. A. Hirsch, Angew. Chem. Int. Ed. 41, 1853 (2002).74. A. Kukovecz, C. H. Kramberger, M. Holznger, H. Kuzmony, J. Schalka, M. Mannsberger and

A. Hirsh, J. Phys. Chem. B 106, 6374 (2002).75. J. Zhang, H. Zhou, Q. Qing, Y. Yang, Q. Li, Z. Liu, X. Guo and Z. Pu, J. Phys. Chem. B 107,

3712 (2003).76. A. Garg and A. B. Sinnott, Chem. Phys. Lett. 295, 273 (1998).77. L. S. Wojciech, J. A. Libera, Y. Gogotsi and Y. Masahiro, J. Solid. State. Chem. 160, 184 (2001).78. W. Huang, Y. Lin, S. Taylor, J. Gaillard, A. M. Rao and Y.-P. Sun, Nanolett. 2, 231 (2002).79. Y.-P. Sun, W. Huang, Y. Y. Lin, K. Fu, A. Kitaygarodskiy, L. A. Riddle, Y. J. Yu and D. L. Carroll,

Chem. Mater. 13, 2864 (2001).80. W. Z. Li, S. S. Xie and G. Wang, Science 274, 1701 (1996).81. E. V. Basiuk, V. A. Basiuk, J.-G. Bauelos, J.-M. Saniger-Blesa, V. A. Pokrovskiy, T. Y. Gro-

movoy, A. V. Mischanchuk and B. G. Mischanchuk, J. Phys. Chem. B 106, 1588 (2002).82. G. N. Fursey, P. V. Novikov, G. A. Dyuzher, A. V. Kotcheryzhenrov and P. O. Vassilier, App.

Surf. Sci. 215, 135 (2003).83. L. Cao, H. Chen, J. Sun, X. Zhang and F. Kong, J. Phys. Chem. B 106, 8971 (2002).84. J. Chen, A. M. Rao, S. Lyuksyutov, M. E. Itkis, M. A. Hamon, H. Hu, R. W. Cohn, P. C. Eklund,

P. T. Colbert, R. E. Smalley and R. C. Haddon, J. Phys. Chem. B 105, 2525 (2001).85. C. Velasco-Santos, A. L. Martnez-Hernndez, M. Lozada-Cassou, A. Alvarez-Castillo and

V. M. Castao, Nanotechnology 13, 495 (2002).86. W. Huang, S. Taylor, K. Fu, Y. Lin, D. Zhang, T. W. Hanks, A. M. Rao and P.-Y. Sun, Nanoletters

2, 311 (2002).87. J. H. Hafner, C. L. Cheung, A. T. Woolley and C. M. Lieber, Prog. Biophys. Mol. Biol. 77, 73

(2001).88. A. L. Martnez-Hernndez, C. Velasco-Santos, M. De Icaza and V. M. Castao, e-Polymers,

no. 016 (2003), http://www.e-polymers.org89. A. K. Bledzki, S. Reihmane and J. Gassan, J. Appl. Polym. Sci. 59, 1329 (1996).90. Y. L. Liu, C. Y. Hsu, M. L. Wong and H. S. Chen, Nanotechnology 14, 813 (2003).91. P. Calvert, Nature 399, 210 (1999).92. H. Z. Geng, R. Rosen, B. Zheng, H. Shimoda, L. Fleming, J. Liu and O. Zhou, Adv. Mater. 14,

1387 (2002).93. F. H. Gojny, J. Nastalczyk, Z. Roslaniec and K. Schulte, Chem. Phys. Lett. 370, 820 (2003).94. H. W. Goh, S. H. Goh, G. Q. Xu, K. P. Pramoda and W. D. Zhang, Chem. Phys. Lett. 373, 277

(2003).

Documents

Carbon Nanotube Polymer Nanocomposites the Role of Interfaces