European Polymer Journal 55 (2014) 210–221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

Contents lists available at ScienceDirect

European Polymer Journal

journal homepage: www.elsevier .com/locate /europol j

Macromolecular Nanotechnology

Synergistic effects of hybrid graphitic nanofillerson simultaneously enhanced wear and mechanical propertiesof polymer nanocomposites

http://dx.doi.org/10.1016/j.eurpolymj.2014.04.0020014-3057/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +1 509 335 7658; fax: +1 509 335 4662.E-mail address: katie_zhong@wsu.edu (W.-H. Zhong).

Tian Liu, Yu Wang, Allen Eyler, Wei-Hong Zhong ⇑School of Mechanical and Materials Engineering, Washington State University, Pullman, WA 99164, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 23 January 2014Received in revised form 31 March 2014Accepted 6 April 2014Available online 13 April 2014

Keywords:Polymer nanocompositeHybrid graphitic nanofillersWear resistanceMechanical properties

In this study, hybrids of organosilane-functionalized graphitic nano-materials, nanofibers(GNFs, 1-D) and nanoplatelets (GNPs, 2-D) were applied to fabricate high-density polyeth-ylene (HDPE) nanocomposites with simultaneously enhanced wear and mechanical prop-erties. Organosilane as a surfactant was shown to effectively assist the nanofiller–polymermatrix interactions, filler dispersion and distribution, GNP interlayer sliding motion, aswell as the establishment of a filler network. The ratio of weight percentage (wt%) of silanecoated onto GNP relative to GNF (defined as factor R) was adjusted by the amount of sur-factant. The results indicated that hybrid nanofillers consisting of the GNF and GNP withthe highest wt% ratio of silane coating on GNP compared to that of GNF suggested a syn-ergistic effect on the improvements in both wear resistance and storage modulus of thenanocomposites. Compared to the pure polymer, the wear resistance of the hybrid GNF–GNP composite with the highest factor R was improved by 89%; the storage moduluswas increased: up to 70% at �70 �C, and 83% at room temperature.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

New functional nanocomposites have been constantlyemerging in industry applications in recent years.Composed of polymers and nano-reinforcements, a varietyof polymeric nanocomposites have been successfullydeveloped with high mechanical, thermal, electrical,dielectric properties, etc. [1–18]. Most of these investiga-tions have shown that a great deal of enhancement inone of the properties can be achieved effectively. However,findings of designed nanocomposites possessing two oreven more desirably improved performances are less fre-quently reported. In reality, the development of novelmaterials with multi-functionalities is a crucially impera-tive in the field of materials science and technology.

Rational combination of various properties can signifi-cantly broaden practical applications of nanocomposite.

As an important multi-functional nano-reinforcement,graphite nanoplatelets (GNPs) are highly popular in elec-trostatic painting, gas storage, thermal interface materials,and other applications [19–21]. Incorporating GNPs withpolymer matrices, the nanocomposites are able to realizemulti-functionalities and show fruitful accomplishmentsin improving electrical/thermal conductivities, thermalstability, and mechanical performances [13,22]. With theaddition of only small amount of GNPs (�0.5 wt%), theelectrical and thermal properties were enhanceddramatically compared to neat polymers. However, to fur-ther heighten electrical/thermal conductivities, a furtherincrease in GNP concentration is necessary due to the highpercolation threshold of GNP, which is in the range of 6–20wt%, and even higher for some polymer matrices [22,23].For mechanical performance enhancement, it is more chal-

T. Liu et al. / European Polymer Journal 55 (2014) 210–221 211

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

lenging to reach an extraordinary improvement [24]. Afundamental cause of all these phenomena is related tothe two-dimensional (2-D) structure of GNPs. Hence, ahigh GNP content in the composites is commonly selectedin order to simultaneously improve the mechanical andelectrical and/or other properties. Recently, it has beendemonstrated that the 2-D graphene multi-layered struc-tures were capable of inducing high lubricating efficiencyduring contact frictional movement, and considerablyenhanced the wear resistance of the resultant polymericnanocomposites at only 3 wt% filler loading [25]. For somespecial practical applications, such as load-bearing jointreplacements and other grinding systems, the wear behav-ior of a material is inseparable from its mechanical perfor-mance [26]. Therefore, combination of this improvementwith mechanical properties enhancement in the samenanocomposite is desired. Thus, an exclusive researchfocusing on both wear and mechanical properties of GNPreinforced polymer composites, while maintaining a lowfiller concentration, is needed. Thus far, no reported workhas addressed on this problem.

Graphite nanofibers (GNFs) are considered as good can-didates to improve the mechanical properties of GNP–polymer composites. The employing GNF and GNP to rein-force polymers for tribology-mechanics joint applicationsis rooted in the following aspects: (1) Both GNFs and GNPsare graphitic nanofillers. They are composed of stackedgraphene layers bonded with van der Waals forces. Thecombination of these two kinds of nanofillers can benefitcompatibility between the hybrid nanofillers in composite;(2) GNFs with 1-D cylindrical structure have a relativelylow percolation threshold compared to GNPs. Accordingly,the addition of GNFs has great potential to promote theformation of a complete filler network in the compositesystem, thereby enhancing the load transfer between thenanofillers and polymer matrix; (3) similar to GNPs, thewide availability, low cost, and high quality and stabilityof GNFs makes them ideal for high-volume applications.Therefore, the incorporation of the two structured gra-phitic nanofillers into the polymer matrix is a promisingapproach to the design of a nanocomposite with improvedmechanical properties and wear resistance.

To produce a high-performance multi-functional com-posite with hybrid nanofillers, there are three essentialprerequisites. There must be well dispersed and distrib-uted nanofillers, a strengthened interface between nanofil-lers and polymer matrix, as well as high compatibility ofthe hybrid nanofillers. The aggregation of nanofillers anda poor filler–matrix interface inevitably compromise theireffectiveness [19]. Thus, the surface modification on bothnano-reinforcements and a suitable fabrication approachcan be effective in achieving the goals. Commercially avail-able GNFs and GNPs have active sites on their edges forfurther functionalization, typically hydroxyl (on bothnanofillers) and carbonyl groups (only on GNPs) [25,27].To date, numerous surface treatments on either GNF orGNP via various chemicals have been identified [28–34].The results demonstrate that the treatments contributeto uniform filler dispersion/distribution and strengthenedinterfacial bonding in the polymeric nanocomposites.However, studies focusing on the surface modifications of

two or more nanofillers in one composite system are rare.A rational combination of hybrid nanofillers with highcompatibility and the application of appropriate fabrica-tion method for the nanocomposites can further improvethe dispersion and distribution of reinforcements, as wellas the load transfer between the matrix and nanofillers.

In this work, we applied both GNFs and GNPs toreinforce high-density polyethylene (HDPE) compositesand aimed to improve wear resistance and mechanicalperformance simultaneously through a synergisticallystrengthening effect of the two graphitic nanofillers. A typeof organosilane coupling agent, octadecyltrimethoxysilane(ODMS), was used as a surfactant to modify hybrid fillers,GNF–GNP, in order to achieve homogenous dispersionand distribution of nanofillers, their strengthened interfa-cial interactions with HDPE, as well as high compatibilitybetween nanofillers. For comparison, nanocomposites withsingle graphitic nanofiller (GNFs and GNPs, respectively)were prepared by the same fabrication method. A rela-tively low filler loading, 3 wt%, was selected for all compos-ite samples (weight ratio of hybrid GNF–GNP is 1:1). Theanalysis of silanized hybrid GNF–GNP was performed toobserve the morphology and determine the weight per-centage (wt%) ratio of silane coating on GNP relative toGNF. It was demonstrated that the GNFs and GNPs wereconnected to each other through silanization and the wt%ratio of silane coating on GNP to GNF could be adjustedby using different dosage of ODMS. The wear and mechan-ical testing results revealed considerably improved wearresistance and storage modulus in the hybrid nanofillercomposite system with the highest wt% ratio of silanecoating on GNP relative to GNF.

2. Experimental procedures

2.1. Materials

High-density polyethylene (HDPE) as the matrix used inthis study was purchased from Bamberger Polymers Inc.with an approximate density of 0.95 g cm�3. The hybrid-ized graphitic nanofillers included graphitic nanoplatelets(GNPs) and graphite nanofibers (GNFs). The GNPs obtainedfrom XG Science Inc. were pre-exfoliated with an averagediameter of �5 lm and �5 nm in thickness. The GNFs sup-plied from Applied Sciences Inc. were pre-oxidized andhave 60–150 nm in diameter and 30–100 lm in length.Octadecyltrimethoxysilane (ODMS) as surfactant was pur-chased from Sigma–Aldrich and has 90% technical gradeaccording to the supplier. 100% ethanol was obtained fromDecon Laboratories Inc. Sodium hydroxide (NaOH) with 5%w/v aq. soln. was received from Alfa Aesar.

2.2. Synthesis of silanized hybrid GNF–GNP

Both GNF and GNP obtained from the suppliers werepre-treated via highly oxidative procedures, resulting inabundant active chemical sites existed on the surface ofthe particles [20,30]. The hydroxyl group on the GNFsand both hydroxyl and carboxylic acid groups on the GNPs

212 T. Liu et al. / European Polymer Journal 55 (2014) 210–221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

are available for further surface modification to improvethe nanofiller–polymer matrix compatibility.

Detailed information on the silanization process for asingle graphitic nanofiller can be found in our previousstudies [25,27]. The sub-sequent surface treatment ofhybrid GNF–GNP was performed with the help of ODMS.The mixture of GNF and GNP was dispersed in 95% ethanolusing room temperature bath-type sonication for 1 h toobtain a homogenous suspension without shortening thenanofiber length or reducing the nanoplatelet diameter.The diluted ODMS with ethanol was then added into thenanofiller suspension under boiling conditions. The con-densation reaction occurred between reactive functionalgroups on the graphitic nanofillers and hydrolyzed silanecoupling agent. Before the reaction started, NaOH as cata-lyst was added to the solution as a catalyst to increase thedegree of hydrolysis of ODMS [27]. Finally, an organosilanelayer was formed to cover GNF and GNP surfaces. Bychanging the dosage of ODMS, we can control the weightpercentages of silane coating on GNP and GNF.

2.3. Manufacturing of HDPE/single and hybrid graphiticnanofiller composites

The dry mixing processing method was applied tofabricate HDPE composites reinforced by either singleor hybrid graphitic nanofillers. 3 wt% silanized-GNF(s-GNF), silanized-GNP (s-GNP), and silanized-GNF&GNP(s-GNF&GNP) were melt-blended with HDPE by a HaakeTorque Rheometer. Half of the HDPE powder was mixedwith the nanofillers at 170 �C with a rotator speed of30 rpm for 2 min. Thereafter, the rest of polymer powderwas added, and the rotation speed was raised to 70 rpmfor 13 min. The polymeric nanocomposites were subse-quently hot-pressed at 180 �C for 15 min via a hydrauliccompression molding machine. The panels were allowedto cool down to room temperature naturally after turningoff the heat. All samples were cut with specific sizes forwear and mechanical testing, respectively. Table 1 showsthe detailed descriptions of each composite sample.

2.4. Characterization

2.4.1. TGA (Thermogravimetric analysis)To determine the weight percentage of silane coating on

the hybrid graphitic nanofillers with different ODMS dos-ages (s-GNF&GNP-075, s-GNF&GNP-150, and s-GNF&GNP-

Table 1Detailed description for each sample.

Sample Surface modification of nanofillers

Filler type

Comp-sGNF-075 Silanized GNFComp-sGNF-150Comp-sGNF-250

Comp-sGNP-075 Silanized GNPComp-sGNP-150Comp-sGNP-250

Comp-s(GNF&GNP)-075 Chemically mixed GNF&GNPComp-s(GNF&GNP)-150Comp-s(GNF&GNP)-250

250), a thermogravimetric analyzer (TA Instruments, SDTQ600, USA) was applied in this work. All the TGA testswere performed at a temperature range from ambient tem-perature to 800 �C, with a heating rate of 10 �C/min in anN2 environment and a flow rate of 100 mL/min. For eachtest, nanofiller samples with similar mass (�10 mg) wereused.

2.4.2. TEM (Transmission electron microscope)TEM (FEI Tecnai, G2 T20, USA) was used to observe the

morphologies of silanized graphitic nanofillers and mea-sure the coating thickness on s-GNF. To prepare the TEMspecimens, nanoparticles were suspended in 100% ethanolfollowed by 2 h bath-type sonication. The middle layer ofthe suspension was then dropped onto TEM copper gridwith 300 meshes. The sample was ready for imaging afterovernight vacuum drying.

2.4.3. FESEM (Field emission scanning electron microscope)For each nanocomposite, the nanofiller dispersion and

filler–matrix interaction were investigated through FESEM(FEI Tecnai, Quanta 200F, USA). The fracture surface of thecomposites was prepared to take SEM images. All sampleswere frozen in liquid nitrogen for a couple of minutes priorto fracturing, and then sputter-coated with gold for electri-cal conductivity.

2.4.4. Wear testingThe wear resistance of the composites was determined

by a custom-built pin-on-disk rig with a vertical 1020 car-bon steel disk (Hitachi, L200, USA). The radius of the weartrack was 70 mm and a 36 N normal force was applied tothe specimen. The detailed schematic diagram of the weartesting apparatus was described in our previous studies[25,35]. The weight loss for each sample was recorded forthe calculation of specific volumetric wear rate by Eq. (1),given below:

W ¼ DmqFd

ð1Þ

where w is specific volumetric wear rate, Dm is weight lossafter each wear period, q is the density of the compositesample, F means the applied normal force, and d refers tolinear sliding distance. The wear disk was polished by finesand paper (silicon carbide 220 grit) and cleaned with100% ethanol prior to each test. For each type of composite,at least 4 specimens were tested.

Filler concentration

Volume of ODMS (ml)

0.75 GNF – 3 wt%1.52.5

0.75 GNP – 3 wt%1.52.5

0.75 GNF&GNP – 3 wt (Weight ratio: 1)1.52.5

T. Liu et al. / European Polymer Journal 55 (2014) 210–221 213

OTE

CHN

OLO

GY

2.4.5. DMA (Dynamic mechanical analysis)DMA testing for the nanocomposites was performed

using a TA Instrument (Tritec 2000) with a liquid nitrogencooling system in a single cantilever bending mode. Allsamples were rectangular in shape with a size of 35 mm(L) � 14 mm (W) � 2 mm (T). The testing was set at 1 Hzfrom �70 �C to 135 �C with a ramp rate of 2 �C/min. Thestorage modulus (E0), loss modulus (E00), and the relaxationpeak were recorded. The testing results shown in thispaper are the averages of at least 5 repeated tests obtainedfrom different specimens for each type of compositesample.

2.4.6. DSC (Differential scanning calorimetry)The melting behaviors of single and hybrid graphitic

nanofiller reinforced HDPE nanocomposites were per-formed in a TA instrument (Mettler-Toledo, Inc., TA DSC822). All samples were sealed in aluminum pans andscanned under a nitrogen atmosphere with a heating rateof 10 �C/min from 25 �C to 160 �C. Samples with similarmass (�5 mg) were used in each test.

2.4.7. AC conductivityAlternating current (AC) conductivity was tested at

ambient conditions using a Novocontrol TechnologiesAlpha-N High Resolution Dielectric Analyzer equippedwith Au parallel plate sensors. Both samples and plate sen-sors were cleaned thoroughly with ethanol and then driedcompletely.

MO

LECU

LAR

NA

N

3. Results and discussion

3.1. Coating analysis on hybrid graphitic nanofillers

The studies on single graphitic nanofiller surface modi-fication by using organosilane coupling agent have beenreported in our previous work [25,27]. The reactive silanolgroups in the hydrolyzed ODMS can create chemical bonds

Fig. 1. Schematic representation of the silanizati

with polar groups (–OH or –COOH) to either GNF or GNPvia condensation reactions. However, when both two gra-phitic nanofillers participate in the same silanization reac-tion, it is possible that the GNF and GNP can be connectedvia ‘‘silane-bridge’’. Also, the chemical reactions of ODMS–GNF and ODMS–GNP should occur not only simultaneouslybut also competitively. Under the same chemical reactionwith a fixed amount of silane coupling agent, the differ-ence in the ratio of functional groups on the two nanofillersand their activities lead to a preferential ‘‘adsorption’’ ofthe silane layer on one type of nanofiller. In this case, thepercentages of organosilane coated on GNP and GNF mustbe different. In particular, it becomes more complex as dif-ferent dosages of ODMS are used in this process since thedegree of hydrolysis of the surfactant can be differentunder fixed water content [27]. The basic mechanism ofhybrid GNF–GNP silanization can be described as Fig. 1.Hence, the determination of the weight percentage ratioof silane coating on GNF to GNP for the hybrid graphiticnanofillers is crucial.

First of all, we used TGA to quantify the total amount ofsilane coating on the hybrid graphitic nanofillers after sil-anization. The influence of ODMS dosage during the silan-ization process of hybrid GNF–GNP was investigatedsimultaneously. According to Fig. 2, three s-GNF&GNPsamples displayed dramatic weight loss at around 500 �C,which can be attributed to the different thermal decompo-sition behaviors between the graphitic nanofillers and theorganosilane layer on them [27]. Clearly, the percentageof weight loss increased with an increase in the dosage ofODMS. The s-GNF&GNP-075 yielded 35 wt% organosilanecoating, the s-GNF&GNP-150 yielded approximately 43wt%, and the sample of s-GNF&GNP-250 yielded up to 52wt%. More surfactant involved in the silanization processcan provide more reactive groups which are available toparticipate in condensation reactions, resulting a thickercoating on both nanofillers.

Secondly, we can evaluate the weight percentage ofsilane coating on GNF in hybridized graphitic nanofillers

on process of hybrid graphitic nanofillers.

MA

CRO

Fig. 2. TGA curves for the comparison of silanized GNF–GNP with three different volumes of ODMS in the surface modification.

214 T. Liu et al. / European Polymer Journal 55 (2014) 210–221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

based on the measurement of GNF coating thickness viaTEM image analysis. In Fig. 3, the morphologies of silanizedhybrid graphitic nanofillers with three different ODMS vol-umes during the surface modification can be observed. Thefirst row of Fig. 3 shows the silane-coated GNPs. The darkerportions on GNP surfaces reflect the locations of silanecoating. The samples of s-GNF&GNP-075 (Fig. 3A) and s-GNF&GNP-250 (Fig. 3C) displayed cleavage features, indi-cating the typically exfoliated structure of the pristineGNPs. The rough surfaces of the two samples demon-strated that the organosilane coating on GNPs was verythin, especially for s-GNF&GNP-075. In contrast, a verysmooth plate surface was seen in s-GNF&GNP-150 as aresult of a thick silane layer covered on the GNP surface(Fig. 3B). It is obvious that most portions of s-GNF&GNP-150 were surrounded by the silane coating (darker sub-stance). Besides, it can be noticed that some GNP andGNF were brought in contact, with the silane layer as abridge. Thus, we can conclude that the connection of thetwo graphitic nanofillers through a ‘‘silane-bridge’’ is apossible mechanism for the silanization of GNF–GNP.

In the second row of Fig. 3, the highly magnified TEMimages clearly show the silanized GNFs within hybridizedgraphitic nanofillers. Note that there exist distinct outerportions which are darker than the other areas, represent-ing the silane coating on GNF surface. The coating thick-ness was measured by an image analysis tool for thefurther calculation of weight percentage of silane coatingon GNF in hybrid graphitic nanofillers. In order to scientif-ically investigate the coating wt% of GNF, 15 images foreach type of the sample were taken for quantitative analy-sis of GNF coating thickness. The average values with cor-responding deviations are shown in Table 2. We observedthat the thickness of silane coating on GNFs visiblyincreased as the ODMS dosage was raised. Three examplesof silanized GNFs with measured coating thicknesses weredisplayed in Fig. 3a–c. The coating thickness was increasedfrom 18 nm in s-GNF&GNP-075 (Fig. 3a), to around 29 nmin s-GNF&GNP-250 (Fig. 3c). According to all obtained TEMimages, it can be verified that the organosilane was indeedcoated onto both GNPs and GNFs through the silanizationof the hybrid graphitic nanofillers. Moreover, the dosageof ODMS can affect the coating thickness on the nanofillers.

Finally, based on the total weight percentage of silanecoating on hybrid GNF–GNP determined by TGA and fromthe coating thickness on GNF measured through TEMimages, the volume fraction of GNF in the composite canbe calculated by the following equation:

pr2 ¼ VGNF � pðr þ s1Þ2 ð2Þ

where r is diameter of GNF, VGNF is the volume fraction ofGNF, and s1 is the average thickness of silane coating onGNF; then, the silane coating thickness on GNP was esti-mated by:

ab ¼ VGNP � aðbþ 2s2Þ ð3Þ

where a and b are the diameter and thickness of GNP,respectively, VGNP is the volume fraction of GNPs, and s2

is the average thickness of silane coating on GNP. Thereaf-ter, the weight percentage ratio of coating on GNP to GNFcan be determined, which defined as factor R,

R ¼Wco-GNP

Wco-GNFð4Þ

where Wco-GNP and Wco-GNF are the coating weight percent-ages on GNP and GNF, respectively. The comparison of allcalculations for three silanized hybrid graphitic nanofillerswas presented in Table 2. It can be noted that the percent-age of silane coating on GNP was much higher than that ofon GNF. This is consistent for all three samples. 2-D platestructured GNPs with abundant functional groups, typi-cally hydroxyl and carboxylic acid, can provide more activesites and attract more hydrolyzed ODMS with reactive sil-anol groups for condensation reactions. Additionally, wefound that the coating weight percentages on either GNFsor GNPs in hybrid graphitic nanofillers increased withincreasing dosage of ODMS during silanization process.However, this regularity is not seen in the factor R. Thesample s-GNF&GNP-150 showed the highest R value, indi-cating the highest wt% ratio of coating on GNP to GNF. Wesupposed that the chemical reaction rate of hydrolyzedsilane coupling agent with GNPs was higher than that withGNFs due to more active sites on GNPs. Thus, the moreODMS provided, the more silane can be coated on GNP sur-face rather than on GNF. As the ODMS dosage is further

Fig. 3. TEM images (scale bar: 500 nm for A and B; 1 lm for C; 100 nm for a and c; 200 nm for b) for silanized GNF–GNP with different ODMS volumes(A–C), and the silane coated GNF with measured coating thickness for each sample (a–c).

Table 2Analysis of the wt% ratios of silane coating on the surfaces of GNP to GNF.

s-(GNF&GNP)-075 s-(GNF&GNP)-150 s-(GNF&GNP)-250

Total coating wt% 35 43 52Average coating thickness on GNF 19.01 ± 1.23 (nm) 20.09 ± 1.04 (nm) 29.55 ± 1.11 (nm)Coating wt% on GNF 6.11 5.69 7.84Coating wt% on GNP 28.89 37.31 44.16R 4.73 6.58 5.63

T. Liu et al. / European Polymer Journal 55 (2014) 210–221 215

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

increased, the chemical reaction rate of ODMS with GNPs isreduced owing to less ‘‘free space’’ on GNPs available forthe reaction. Most of the surface areas of the GNPs werecovered by the silane layer. Although there were still resid-ual functional groups on GNPs under the coating, it is noteasy for the unreacted silane to penetrate the coating layerin order to react with those groups. Instead, the unreactedsilane had a tendency to react with GNFs so as to improvethe relative reaction rate of ODMS and GNFs. At a criticalpoint, the ODMS dosage in hybrid GNP–GNF will result ina maximum value for the wt% ratio of GNP to GNF. Thiscritical point is between 1.5 mL and 2.5 mL in the currentwork. Therefore, the sample s-GNF&GNP-150 showed thehighest R value.

3.2. Morphology study on hybrid GNF–GNP composites

The morphology investigations on HDPE nanocompos-ites reinforced by single graphitic nanofillers (GNFs orGNPs) in our previous work have confirmed that silaniza-tion can assist the dispersion and distribution, as well asthe interfacial interaction of the nano-reinforcement witha PE matrix [25,27]. Generally, no observable big nanofilleraggregations were observed in the composites. Due to the

crystallizable characteristics of ODMS, the cocrystallizationbehavior of organosilane and semi-crystalline PE can occur.Hence, the graphite nanofiller-PE matrix interfacial transi-tion region possessed to enhance strength [36]. For the sil-anized-GNF/HDPE composite, the uniformly dispersed anddistributed GNFs with tubed structure were the primaryimpetus behind the potential of the nanofiber networkstructure formation at a low concentration. For thesilanized-GNP/HDPE composite system, the surface modi-fication helped to improve the dispersion and distributionof GNPs. However, compared to silanized-GNF/HDPEcomposite with the same filler loading, there existed largerun-reinforced polymer regions due to the 2-D multi-layered structure of GNPs.

Similarly, a homogeneous dispersion and distribution ofhybrid GNF–GNP were obtained in the HDPE matrix byintroducing the silane layer between the fillers and poly-mer matrix, as observed from the fracture surface of thecomposites (Fig. 4). No visible nanofiller agglomerateswere observed in the composites. In particular, it shouldbe noted that the GNP un-reinforced areas were filled withGNF (Fig. 4A–C), reflecting the hybrid GNF–GNP can estab-lish a filler network structure compared to the singlegraphite nanofiller composites. Also, the GNFs were

Fig. 4. FESEM micrographs (scale bar: 20 lm for A–C; 4 lm for a–c) of the nanocomposites reinforced by silanized hybrid graphitic nanofillers with threeODMS volumes.

216 T. Liu et al. / European Polymer Journal 55 (2014) 210–221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

attached to the edges of most of GNPs (Fig. 4a–c). This con-firms that the GNFs and GNPs can be connected to eachother via the silane layer, coinciding with the morphologystudy on the hybrid graphitic nanofillers (Fig. 3A–C). In thesample Comp-s(GNF&GNP)-075, there were some individ-ual long GNFs exposed on the fracture surface of the com-posite (Fig. 4A). Tiny gaps existed between the two phasesand a few of the GNFs were pulled out. Besides, obviousGNP plates were found in the polymer. However, theirinterfacial interactions with the matrix were not as goodas those between GNF and PE in the same composite. Obvi-ous GNP–PE gaps were clearly seen, suggesting the coatedsilane was insufficient to hold the multi-layered GNPsunder the external force; numerous GNP plates werepulled out during fracturing. This can be verified by themagnified SEM image with single GNPs, as shown inFig. 4a. The noticeable cleavage features shown on the sur-face of GNPs indicates their exfoliated structure. It demon-strates that the silane layer on GNPs in Comp-s(GNF&GNP)-075 was very thin, which is consistent withthe previous coating analysis results (Table 2). Addition-ally, the oriented GNPs in the composites can be realizedby means of the melt-blending manufacturing approachapplied in this study.

With the increased ODMS volumes, the filler–matrixinteraction was further improved while maintaining uni-form filler dispersion and distribution. In Comp-s(GNF&GNP)-150 and Comp-s(GNF&GNP)-250, only veryshort GNFs (white spots) and fewer GNP plates can beobserved on the fracture surface of the composites(Fig. 4B and C). Also, no gaps were visible between the dif-ferent phases, indicating that the hybrid nanofillers werestably embedded in the polymer matrix. This phenomenonwas attributed to the thick silane layer on the hybrid fillersurfaces that can be sufficient to create robust cocrystalli-zation interface between silane coating and HDPE. The

morphologies of GNPs of the two nanocomposites can beseen clearly in Fig. 4b and c. The GNP plate in Comp-s(GNF&GNP)-250 still displayed cleavage features eventhough it is not so obvious as the GNPs in Comp-s(GNF&GNP)-075 (Fig. 4a). In comparison, from Fig. 4b, arather smooth surface was seen on the GNP plate ofComp-s(GNF&GNP)-150 thanks to the thickest silane cov-ering of the sample (Table 2). Overall, the morphologystudy presented another piece of evidence of that Comp-s(GNF&GNP)-150 had the highest wt% of silane coatingon GNPs out of the three hybrid graphitic nanofiller com-posites. Furthermore, the filler network structure can beperfected in the hybrid composite system.

3.3. Wear resistance

Organosilane modified GNPs have been confirmed tofunction as a solid lubricant; a graphene interlayer slidingmotion occurs during wear process, potentially improvingthe wear performance of polymeric nanocomposites [25].In order to explore whether the treated GNPs were still apromising reinforcement in hybrid graphitic nanofillercomposites for tribological applications, the wear propertyof HDPE/GNF–GNP system was investigated to comparewith that of the single graphite nanofiller composites.The weight loss of nanocomposite prepared in this studywas determined by a pin-on-disk wear testing apparatus.The comparison of specific volumetric wear rate calculatedby Eq. (1) is exhibited in Fig. 5. For each type of silanizedsingle-nanofiller-reinforced HDPE composites, the wearrates of the samples with high ODMS dosages during thesilanization process (�150 and �250 for each type) werevisibly higher than those of the same composite systemwith low silane content (�075). High silane content canprovide a large number of long hydrocarbon silane tailswhich lead to the accumulation of quite a few adjacent

Fig. 5. Specific volumetric wear rates of HDPE nanocomposites reinforced by GNF, GNP, and hybrid GNF–GNP with three ODMS volumes during silanization(dashed line: wear resistance for pure HDPE).

T. Liu et al. / European Polymer Journal 55 (2014) 210–221 217

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

graphite nanofillers through the binding of silane tails [27].In this case, the aggregates were inclined to form, nega-tively impacting filler dispersion. As a result, the wear per-formance of the composites was discounted. This effect ismore obvious in the samples of Comp-sGNP. The specificvolumetric wear rate of Comp-sGNP-250 showed the high-est value, even higher than that of the pure polymer (thedash line in Fig. 5). The unique 2-D structure of GNP makesit difficult to form a network structure in the composite atlow loading, particularly with respect to undispersed fill-ers. The existence of un-reinforced HDPE regions led tothe failure of load transfer from the nanofillers to the poly-mer matrix [25]. Furthermore, the aggregated GNPs wereunfavorable to the storage and dissipation of frictional heatgenerated by wear, decreasing the wear resistance of thecomposites.

Compared to the single graphite nanofiller composites,the hybrid GNF–GNP composite system exhibited signifi-cantly reduced specific volumetric wear rate under highODMS dosages (�150 and �250). From the morphologystudies, we know that a smaller amount of silane couplingagent coated onto the GNP surfaces in Comp-s(GNF/GNP)-075 caused their poor interfacial interactions with polymermatrix (Fig. 4A and a). Consequently, two primary factorsdirectly negatively affect the wear performance of thecomposite: (1) The GNP interlayer sliding motion isrestricted under external force due to weak interfacialbonding between GNP layers (a very thin coating layerunstably holds the multi-layered GNPs); (2) The wear gen-erated heat cannot be stored and dissipated during thewear process as a result of the failure of loading transfer.Hence, its wear resistance was even lower than Comp-sGNF-075 and Comp-sGNP-075. With increasing silanecontent in the hybrid graphitic nanofiller composite, thefiller–matrix interactions were obviously enhanced(Fig. 4B, b, C, and c). We suppose that the surface modifica-tion with sufficient silane content promoted the formationof silane-layer intercalated GNPs, allowing strong GNPlayer sliding motion to occur and contributing to highlubricating efficiency during wear. Besides, the effectivenanofiller network was successfully established throughuniform dispersion and distribution, as well as thestrengthened interface between phases.

For the three Comp-s(GNF/GNP) samples, the wearresistance was not consistently improved with increasing

silane content. The sample with the medium ODMS dosageduring the filler surface treatment showed the lowest wearrate. Compared to pure HDPE, the wear resistance ofComp-s(GNF&GNP)-150 was improved almost 90%. Ourprevious silane coating analysis on the hybrid graphiticnanofillers proposed explanations for this phenomenon(Section 3.1). The sample of s-GNF&GNP-150 demon-strated the highest R value, followed by s-GNF&GNP-250and s-GNF&GNP-075. This pattern is the same as theimprovements on wear resistance of the resultant compos-ites. Accordingly, we can conclude that a high wt% ratio ofsilane coating on GNP to GNF has great benefit in wearresistance enhancement, suggesting GNPs play the role ofsolid lubricant in the hybrid GNF–GNP composite system.In the same composite, a higher wt% silane coating onGNPs compared to GNFs leads to better wear performance.

3.4. Mechanical properties

DMA testing was conducted to quantify the mechanicalproperties of the differently prepared nanocomposites.Fig. 6 and Table 3 show storage modulus and loss moduluswith respect to temperature, as well as the summarizedresults at special points. As can be viewed from theseresults, the storage modulus of composites prepared withhybrid graphitic nanofillers had higher values than thoseof single nanofiller systems at the initial set temperature;thereafter the curves slowly converged with the others(Fig. 6a). This effect is more obvious for Comp-s(GNF–GNP)-150, which started at the highest average E0. To moreintuitively reflect the differences of E0 for all samples in thetesting temperature range, two points were chosen forcomparison, including low (�70 �C) and room tempera-tures (25 �C), as shown in Table 3. Firstly, it is clear thatthe hybrid graphitic nanofiller composites displayedhigher E0 values at both temperature points compared tothe other composite samples. The reason for this comesfrom the hybrid nanofiller network structure formationthrough tuning the distributed GNF–GNP architecture inHDPE. Secondly, Comp-s(GNF&GNP)-150 exhibited thehighest E0 values at the two selected temperature pointscompared to the other hybrid graphitic nanofiller compos-ites. Compared to pure HDPE, the storage modulus ofComp-s(GNF&GNP)-150 increased 70% at -70 �C, and 83%at 25 �C. These increases far surpassed the increase in stor-

Fig. 6. Dynamic Mechanical Analysis showing storage modulus (A–C), loss modulus (a–c) with respect to temperature in HDPE nanocomposites withvarying reinforcements and silane dosages: (A and a) 0.75 ml; (B and b) 1.5 ml and (C and c) 2.5 ml. (Remark: for pure HDPE, storage modulus is 1.66 GPa at�70 �C, and 1.07 GPa at 25 �C.).

Table 3DSC results.

Sample list Tm (�C) Hm (J g�1) Crystallinity (%)

GNF Comp-sGNF-075 133.0 162.7 59.7Comp-sGNF-150 135.2 173.8 62.2Comp-sGNF-250 135.7 180.9 65.8

GNP Comp-sGNP-075 134.5 185.3 62.7Comp-sGNP-150 134.3 189.2 64.8Comp-sGNP-250 136.0 183.6 64.3

Hybrid Comp-s(GNF/GNP)-075 136.3 183.4 64.9Comp-s(GNF/GNP)-150 134.8 197.7 67.1Comp-s(GNF/GNP)-250 134.5 180.2 64.5

Remark: for pure HDPE, Tm is around 130 �C, and the degree of crystallinity is 63.7%.

218 T. Liu et al. / European Polymer Journal 55 (2014) 210–221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

age modulus for Comp-s(GNF&GNP)-075 and Comp-s(GNF&GNP)-250, let alone the single graphite nanofillercomposite systems. This is due to the enhanced interfacialinteraction between GNP and HDPE such that more nanof-illers are available for load transfer compared to a hybridnanofiller–polymer composite system with weak interfa-cial regions. Lastly, similar to the trend found with in wearresistance, the storage modulus of Comp-s(GNF&GNP)-150was higher than that of Comp-s(GNF&GNP)-250, followedby Comp-s(GNF&GNP)-075.

Fig. 6b shows the effect of temperature on the loss mod-ulus (E00) for the different systems. In general, there shouldbe three relaxations in the polyethylene material: (1)alpha-peak, present at around 65 �C, reflects the crystallinephase in semi-crystalline PE; (2) beta-peak, existing ataround �25 �C, reveals the behavior of amorphous regionand glass transition zone; (3) gamma-peak is similar tobeta-peak but appears at around �110 �C [37–39]. In thecurrent work, we examined both alpha and beta relaxationbehaviors in the temperature range of �70 �C to 135 �C. In

T. Liu et al. / European Polymer Journal 55 (2014) 210–221 219

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

Fig. 6b, most of the samples display similar alpha-relaxa-tion peaks, except Comp-sGNF-075, Comp-sGNF-150, andComp-sGNP-075. The low relaxation peaks of these threesamples reflect the limited mobility of large segments inthese composites compared to the rest of the samples.Considering the beta-relaxation peak, it can be clearlyviewed in Table 3 that the peak in the Comp-sGNP compos-ite was markedly larger than that of Comp-sGNF system,corresponding to side group motions in the composites.Because of the large surface area of GNPs, the organosilanecoating cannot cover the whole plate, and thus the nanof-illers become more mobile in response to dynamic loading[25]. Moreover, the larger beta-relaxation peak revealedthat the degree of cocrystallizating compatibility betweensGNP and HDPE might be higher because more active sitesexist on GNP surface. Thus, high energy absorptionbetween GNPs and HDPE can occur. For the system ofComp-s(GNF&GNP), E00 at beta-peak are generally highmainly contributed from sGNP. Among them, the sampleof Comp-s(GNF&GNP)-150 showed slightly higher beta-relaxation. Combining the results of coating analysis, wefound that the R value (coating wt% ratio on GNP to GNF)was relevant to E00 behavior. In the system of Comp-s(GNF&GNP), more silane coated on GNP compared toGNF could lead to higher area of cocrystallized sGNP-HDPEinterface [36]. Also, the coated silane on GNFs could beentangled each other/with GNFs. In comparison, thehydrocarbon tails distributed spreadly on the 2-D GNP sur-face were allowed to be more mobile. Since the relaxationbehaviors of polymeric materials can reflect large seg-ments/side group motions, there must exist a direct rela-tion between relaxation behavior and crystallinity degreeof the composites. This relationship can be examined viadifferential scanning calorimetry (DSC).

3.5. Differential scanning calorimetry

DSC was performed to understand the melting behav-iors and their potential correlations with storage modulusand relaxation peak magnitude of single/hybrid graphiticnanofiller composites. As presented in Table 3, the meltingtemperatures (Tm) obtained in the heating procedure ofvarious graphitic nanofiller composite systems were simi-lar to those of the pure HDPE, suggesting that neither sin-gle nor hybrid graphitic nanofillers considerably affectedthe thermodynamics of their HDPE nanocomposites. Forthe comparison of melting enthalpy (Hm), we found thatinterestingly, the total energy of the Comp-sGNP thermo-dynamic system was generally higher than that ofComp-sGNF. As mentioned previously, the GNPs can pro-vide numerous functional groups as active sites availablefor the silanization. As a result, the coated silane tails andthe residual unreacted polar groups have the potential toincrease the internal energy of the Comp-sGNP system,due to the kinetic energy of abundant active sites, leadingto higher Hm compared to Comp-sGNF [25,27,40,41].Among the three hybrid GNF–GNP composites, Comp-s(GNF/GNP)-150 showed the highest Hm due to the highestwt% ratio of silane coating on GNP relative to GNF (Table2). Besides, though there were slight changes in the degreeof crystallinity between the various composite systems,

Comp-s(GNF/GNP)-150 actually exhibited the highest crys-tallinity, which was up to 67.1%. Combining DMA with DSCresults (Fig. 6 and Table 3), the composite with high stor-age modulus displays a high degree of crystallinity. It hasbeen demonstrated that there exists a direct associationbetween modulus and crystallization [42]. In the currentstudy, the sample Comp-s(GNF/GNP)-150 with hybrid gra-phitic nanofiller network structure was more likely to formregular crystals with ordered arrangement. Thus, it is ben-eficial to the improvement in storage modulus. Addition-ally, we did not find a confirmed relationship betweenthe magnitude of alpha-relaxation and the degree of crys-tallinity. Therefore, it can be concluded that the polymerchain folds and tie molecules during the relaxation processinstead of the crystalline domains [43].

3.6. Electrical performance

In order to confirm the contribution of hybrid GNF–GNPto nanofiller network formation, the AC conductivitybehavior was observed (Fig. 7). In a multiphase nanocom-posite system with an insulating polymer matrix, the elec-trical properties are dependent on the micro-structure ofthe nanofiller phase: its concentration, dispersion and dis-tribution. While the filler concentration is above the criti-cal point, the electrical performances will be dominatedby the conducting nanofillers because the percolationstructure has been formed [36]. At relatively low filler con-tent, the dispersion and distribution of nanofiller are morecrucial. For hybrid s(GNF–GNP) composites, improved ACconductivities compared to single graphitic nanofiller com-posites was observed, demonstrating the formation of con-ductive pathway. It indicates that the combination of GNFand GNP assists the establishment of a filler network in thecomposite at low filler loading, and thus the electricalproperties can be enhanced. In particular, the Comp-s(GNF/GNP)-150 with the largest R value showed almost3 orders of magnitude improvement in AC conductivitycompared to the single graphitic nanofiller compositeswith the same ODMS dosage. It suggests that a high ratioof wt% of silane coated onto GNP relative to GNF facilitatedthe formation of the conductive nanofiller network, pro-ducing a high conductivity.

3.7. Mechanism for the addition of GNF in HDPE/GNPcomposite

Based on all the results obtained across this study, themechanism of the high performance HDPE compositeembedded with hybrid GNF–GNP through silanizationwas summarized in Fig. 8. Both GNF and GNP are graphiticnanofillers, which have the potential to minimize thermaleffect generated through wear and improve mechanicalproperties. A well dispersed and distributed s-GNF&GNPwith the highest R value, wt% ratio of silane coating onGNP to GNF, is benefit to form a nanofiller network struc-ture in the Comp-s(GNF/GNP) system with only 3 wt% fillerconcentration. In this case, the synergistic effects of GNFand GNP can be realized. On one hand, the silanized GNPsas solid lubricant have the ability to allow the interlayersliding motion occurs during the wear process. On the

Fig. 7. AC conductivity of the nanocomposites at 10�2 Hz with single and hybrid graphitic nanofillers.

Fig. 8. Schematic representation of the structure development.

220 T. Liu et al. / European Polymer Journal 55 (2014) 210–221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

other hand, GNFs act as bridges to join one GNP to otheradjacent GNPs with the assistance of silanization andthereby effectively enhance the overall composite system.The analysis results strongly suggest that the hybrid silan-ized GNF–GNP composite with a high R value has greatpotential to improve the mechanical properties whilesimultaneously maintaining superior wear resistance.

4. Conclusions

Hybrid GNF–GNP reinforced HDPE composites werefabricated through the silanization for the two types ofnanofillers. During the nanofiller surface modification,the organosilane coatings on the two graphitic nanofillerswere controlled by the dosage of the silane coupling agent.The highest wt% ratio of silane on GNP to GNF (R = 6.58)was shown to be the most beneficial to the performanceenhancements, i.e. the storage modulus of the compositewas considerably improved, meanwhile, the silane modi-fied GNPs showed efficiently improved the wear resistanceduring the sliding wear process. Therefore, through com-bining the 1-D and 2-D graphitic nanofillers, i.e. GNF and

GNP, a synergistic effect in the prominent mechanicalproperties and wear resistance of the HDPE nanocompos-ites was achieved. The current study strongly suggests thatthe multi-functionalities, such as mechanical and wearresistance can be successfully realized through hybridizingGNF and GNP to reinforce the non-polar polymer, based onthe silanization for the graphitic nanofillers. This studyimplies that the resultant nanocomposites may have greatpotential for tribology-mechanics applications, and thuscontribute to broadening their industrial applications.

Acknowledgments

The authors gratefully acknowledge the financial sup-port from the National Science Foundation (Civil, Mechan-ical and Manufacturing Innovation 0856510).

References

[1] Kumar S, Rath T, Mahaling RN, Reddy CS, Das CK, Pandey KN. Studyon mechanical, morphological and electrical properties of carbonnanofiber/polyetherimide composites. Mater Sci Eng B2007;141:61–70.

T. Liu et al. / European Polymer Journal 55 (2014) 210–221 221

MA

CRO

MO

LECU

LAR

NA

NO

TECH

NO

LOG

Y

[2] Nogales A, Broza G, Roslaniec Z, Shulte K, Ics I, Hsiao BS. Lowpercolation threshold in nanocomposites based on oxidized singlewall carbon nanotubes and poly(butylene terephthalate).Macromolecules 2004;37:7669–72.

[3] Li B, Sui G, Zhong WH. Single negative metamaterials in unstructuredpolymer nanocomposites toward selectable and controllablenegative permittivity. Adv Mater 2009;21:4176–80.

[4] Kumar S, Sun LL, Caceres S, Li B, Wood W, Perugini A. Dynamicsynergy of graphitic nanoplatelets and multi-walled carbonnanotubes in polyetherimide nanocomposites. Nanotechnology2010;21:105702–10.

[5] Ghose S, Working DC, Connell JW, Smith Jr JG, Watson KA, DelozierDM. Thermal conductivity of ultemTM/carbon nanofiller blends.High Perform Polym 2006;18:961–77.

[6] Ci L, Suhr J, Pushparaj V, Zhang X, Ajayan PM. Continuous carbonnanotube reinforced composites. Nano Lett 2008;8:2762–6.

[7] Kumar S, Li B, Caceres S, Zhong WH. Dramatic property enhancementin polyetherimide using low-cost commercially functionalizedmulti-walled carbon nanotubes via a facile solution processingmethod. Nanotechnology 2009;20:465708.

[8] Qi XY, Yan D, Jiang Z, Cao YK, Yu ZZ, Yavari F, et al. Enhancedelectrical conductivity in polystyrene nanocomposites at ultra-lowgraphene content. ACS Appl Mater Interfaces 2011;3:3130–3.

[9] Li B, Zhong WH. Influence of carbon nanofiber network variability onthe AC conductivity of polyetherimide composite films. MacromolMater Eng 2010;295:310–4.

[10] Li B, Wood W, Baker L, Sui G, Leer C, Zhong WH. Effectual dispersionof carbon nanofibers in polyetherimide composites and theirmechanical and tribological properties. Polym Eng Sci2010;50:1914–22.

[11] Gary GT, Max LL, Karla LS, Brian PR. A review of the fabrication andproperties of vapor-grown carbon nanofiber/polymer composites.Compos Sci Technol 2007;67:1709–18.

[12] Li B, Zhong WH. Effective static dissipation of bi-layer structuredthermoplastic nanocomposites at low nanofiber loadings. MacromolMater Eng 2010;295:1136–43.

[13] Li B, Zhong WH. Review on polymer/graphite nanoplateletsnanocomposites. J Mater Sci 2011;46:5595–614.

[14] He Q, Yuan T, Zhu J, Luo Z, Haldolaarachchige N, Sun L, et al.Magnetic high density polyethylene nanocomposites reinforcedwith in situ synthesized Fe@FeO core–shell nanoparticles. Polymer2012;53(16):3642–52.

[15] Zhu J, Sadu R, Wei S, Chen DH, Haldolaarachchige N, Luo Z, et al.Magnetic graphene nanoplatelet composites toward arsenicremoval. ECS J Solid State Sci Technol 2012;1(1):M1–5.

[16] Li Y, Zhu J, Wei S, Ryu J, Sun L, Guo Z. Poly(propylene)/graphenenanoplatelets nanocomposites: melt rheological behaviors andthermal. Elect Elect Prop Macromol Chem Phys2011;212(18):1951–9.

[17] Zhu J, Wei S, Yadav A, Guo Z. Rheological behaviors and electricalconductivity of epoxy resin nanocomposites suspended with in situstabilized carbon nanofibers. Polymer 2010;51(12):2643–51.

[18] Zhu J, Wei S, Ryu J, Budhathoki M, Liang G, Guo Z. In situ stabilizedcarbon nanofibers (CNFs) reinforced epoxy nanocomposites. J MaterChem 2010;20:4937–48.

[19] Li YC, Tjong SC, Li RKY. Electrical conductivity and dielectricresponse of poly(vinylidene fluoride)-graphite nanoplateletscomposites. Synth Met 2010;160:1912–9.

[20] Stankovich S, Dikin DA, Dommett GHB, Kohlhaas KM, Zimney EJ,Stach EA. Graphene-based composite materials. Nature2006;442:282–6.

[21] Ganguli S, Roy AK, Anderson DP. Improved thermal conductivity forchemically functionalized exfoliated graphite/epoxy composites.Carbon 2008;46:806–17.

[22] King JA, Via MD, Morrison FA, Wiese KR, Beach EA, Cieslinski MJ,et al. Characterization of exfoliated graphite nanoplatelets/polycarbonate composites: electrical and thermal conductivity, andtensile, flexural and rheological properties. J Compos Mater2011;46:1029–39.

[23] Jiang X, Drzal LT. Reduction in percolation threshold of injectionmolded high-density polyethylene/exfoliated graphenenanoplatelets composites by solid state ball milling and solid stateshear pulverization. J Appl Polym Sci 2012;124:525–35.

[24] Sengupta R, Bhattacharya M, Bandyopadhyay S, Bhowmick AK. Areview on the mechanical and electrical properties of graphite andmodified graphite reinforced polymer composites. Prog Polym Sci2011;36:639–70.

[25] Liu T, Li B, Lively B, Eyler Allen, Zhong WH. Enhanced wear resistanceof high-density polyethylene composites reinforced byorganosilane-graphitic nanoplatelets. Wear 2014;309:43–51.

[26] Zdravecka E, Tianinen VM, Konttinen YT, Franta L, Vojs M, Marton M,et al. Relationships between the fretting wear behavior andmechanical properties of thin carbon films. Vacuum 2012;86:675–80.

[27] Wood W, Kumar S, Zhong WH. Synthesis of organosilane-carbonnanofibers and influence of silane coating thickness on theperformance of polyethylene nanocomposites. Macromol MaterEng 2010;295:1125–35.

[28] Lakshminarayanan PV, Toghiani JH, Pittman CU. Nitric acid oxidationof vapor grown carbon nanofibers. Carbon 2004;42:2433–42.

[29] Jimenez GA, Jana SC. Oxidized carbon nanofiber/polymer compositesprepared by chaotic mixing. Carbon 2009;45:2079–91.

[30] Planes E, Duchet J, Maazouz A, Gerard JF. Characterization of newformulations for the rotational molding based on ethylene–propylene copolymer/graphite nanocomposites. Polym Eng Sci2008;48:723–31.

[31] Li B, Olson E, Perugini A, Zhong WH. Simultaneous enhancements indamping and static dissipation capability of polyetherimidecomposites with organosilane surface modified graphenenanoplatelets. Polymer 2011;52:5606–14.

[32] Margarita HA, Ahmed AA, Michael JM, Ilhan A, Robert KP.Intercalation and stitching of graphite oxide with diaminoalkanes.Langmuir 2007;23:10644–9.

[33] Si YC, Samulski ET. Synthesis of water soluble graphene. Nano Lett2008;8:1679–82.

[34] Lomeda JR, Doyle CD, Kosynkin DV, Hwang WF, Tour JM. Diazoniumfunctionalization of surfactant-wrapped chemically convertedgraphene sheets. J Am Chem Soc 2008;130:16201–6.

[35] Liu T, Wood W, Zhong WH. Sensitivity of dielectric properties towear process on carbon nanofiber/high-density polyethylenecomposites. Nanoscale Res Lett 2010;6:7. http://dx.doi.org/10.1007/s11671-010-9748-1.

[36] Kim DH, Lee J, Park J, Chung JW, Lee WH, Giri G, et al. Molecularweight-induced structural transition of liquid-crystalline polymersemiconductor for high-stability organic transistor. Adv Funct Mater2011;23:4442–7.

[37] Khanna YP, Turi EA, Taylor TJ, Vickroy VV, Abbott RF. Dynamicmechanical relaxations in polyethylene. Macromolecules1985;18:1302.

[38] Laredo E, Suarez N, Belloa A, de Gascue RB, Gomez MA, Fatou JMG. a,b and c relaxations of functionalized HD polyethylene: a TSDC and amechanical study. Polymer 1999;40:6405.

[39] Matsuo M, Bin Y, Xu C, Ma L, Nakaoki T, Suzuki T. Relaxationmechanism in several kinds of polyethylene estimated by dynamicmechanical measurements, positron annihilation, X-ray and 13Csolid-state NMR. Polymer 2003;44:4325.

[40] Liu T, Wood W, Li B, Lively B, Zhong WH. Effect of reinforcement onwear debris of carbon nanofiber/high density polyethylenecomposites: morphological study and quantitative analysis. Wear2012;294–295:326–35.

[41] Rhodes SM, Higgins B, Xu Y, Brittain WJ. Hyperbranched polyol/carbon nanofiber composites. Polymer 2007;48:1500–9.

[42] Horvath Z, Menyhard A, Doshev P, Gahleitner M, Tranninger C,Kheirandish S, et al. Effect of molecular architecture on thecrystalline structure and stiffness of iPP homopolymers: modelingbased on annealing experiments. J Appl Polym Sci2013;130:3365–73.

[43] Rhodes SM, Higgins B, Xu Y, Brittain WJ. Hyperbranched polyol/carbon nanofiber composites. Polymer 2007;48:1500.