Nanoscale

COMMUNICATION

Cite this: Nanoscale, 2017, 9, 12867

Received 29th June 2017,Accepted 8th August 2017

DOI: 10.1039/c7nr04686c

rsc.li/nanoscale

High thermal conductivity through simultaneouslyaligned polyethylene lamellae and graphenenanoplatelets†

Mortaza Saeidijavash,a Jivtesh Garg, *a Brian Grady,b Benjamin Smith, c

Zheling Li,d Robert J. Young,d Fatema Tarannuma and Nour Bel Bekria

The effect of simultaneous alignment of polyethylene (PE) lamellae

and graphene nanoplatelets (GnP) on the thermal conductivity (k)

of PE–GnP composites is investigated. Measurements reveal a

large increase of 1100% in k of the aligned PE–GnP composite

using 10 wt% GnPs relative to unoriented pure PE. The rate of

increase of k with applied strain for the pure PE–GnP composite

with 10 wt% GnP is found to be almost a factor of two higher than

the pure PE sample, pointing to the beneficial effect of GnP align-

ment on k enhancement. Aligned GnPs are further found to be

3 times as effective in enhancing k as in the randomly oriented

configuration. Enhancement in k is correlated with the alignment

of PE lamellae and GnPs through wide-angle X-ray scattering and

polarized Raman spectroscopy. At the maximum applied strain of

400% and using 10 wt% GnPs, a composite k of 5.9 W mK−1 is

achieved. These results demonstrate the great potential of simul-

taneous alignment effects in achieving high k polymer composites.

Polymer heat exchangers1,2 based on polymers such as poly-ethylene and polypropylene are widely used in applicationsincluding water desalination,3,4 solar energy harvesting,5 auto-motive control units6 and micro-electronics cooling.7–10

Polymers offer several advantages such as lower cost andweight which make them more economically competitive com-pared with metallic heat exchangers.1 Polymeric materials,however, have much lower intrinsic thermal conductivity,11

k (<0.5 W mK−1) compared to metals (>20 W mK−1) which limitstheir more widespread applicability in thermal managementtechnologies. As means to achieve high k polymers, the align-ment of polymer chains has emerged as a new and growingfield of research. Recently, k of a single PE fiber with ultra-

aligned PE chains was measured to be 104 W mK−1 almost 200times12 larger than k of bulk PE (∼0.5 W mK−1). In anotherreport, the effect of the alignment of polymer chains was usedto achieve a thermal conductivity of ∼16 W mK−1 in polyethyl-ene films at draw ratios approaching ∼100.13 Similarly, thealignment of filler material has been explored in the past toenhance k. In recent work, magnetic GnP–Fe3O4 hybrids atvolume fractions of less than 1% were aligned through a mag-netic field in epoxy composites.14 Aligned GnP–Fe3O4 hybridswere found to enhance epoxy k by almost 40% relative to epoxywith randomly oriented GnPs. In this work, only the GnPswere aligned. Other studies have also explored the alignmentof GnPs through an electric field,15,16 flow-assisted align-ment17 as well as self-alignment.18–20 Large enhancements in kthrough the alignment of either the PE chains or filler materialsuggest that by combining the k-enhancements through thealignment of both the polymer chains and the GnPs, evenlarger k values can be achieved.

Graphitic nanoplatelets (GnPs) with the thickness in therange of ∼10–100 nm are excellent heat conductors with in-plane k ∼ 1500 W mK−1 at room temperature.21–23 GnPs alsoshow lower thermal boundary resistance (TBR)24 with poly-mers relative to carbon nanotubes (CNTs). TBR between CNTsand the polymer matrix is estimated to be in the range of10−8–10−7 m2 K W−1;25 the corresponding value for GnPs is inthe range of 10−9 m2 K W−1.24 This lower TBR has led to asignificant growth in the use of GnPs as the filler material toincrease the thermal conductivity of a polymer. While in-planek of GnPs is very high (∼1500 W mK−1), it should be noted thatthe through thickness value is much lower of the order of ∼10–20W mK−1.26 Alignment achieves full advantage of the high in-plane k of GnPs. Similarly, by aligning polymer chains, highlyefficient heat transfer along their chain axis can be achieved.

In this work, the effectiveness of such simultaneous align-ment in enhancing the thermal conductivity is explored. Thealignment of both the polymer chains and embedded GnPs isachieved through the application of mechanical strain to thePE–GnP composites. Mechanical strain is known to align thepolymer chains in crystallites along the direction of applied

†Electronic supplementary information (ESI) available. See DOI: 10.1039/c7nr04686c

aSchool of Aerospace and Mechanical Engineering, University of Oklahoma, Norman,

OK 73019, USA. E-mail: garg@ou.edubChemical, Biological and Materials Engineering, University of Oklahoma, Norman,

OK 73019, USAcVision Sciences, University of California, Berkeley, CA 94720, USAdNational Graphene Institute and School of Materials, University of Manchester,

Manchester M13 9PL, UK

This journal is © The Royal Society of Chemistry 2017 Nanoscale, 2017, 9, 12867–12873 | 12867

strain27 enhancing k along it (Fig. 1a and b). Similarly, strainhas also been shown to align the embedded filler material(Fig. 1c) such as carbon nanotubes.28 The alignment of GnPsin this work is characterized using laser scanning confocalmicroscopy and polarized Raman spectroscopy. The alignmentof PE crystallites is measured using wide-angle X-ray scattering(WAXS) analysis.

While GnPs are chosen both because of their high intrinsick and better thermal interaction with the polymer matrix, PE ischosen as the base polymer because it exhibits a large changein k upon alignment.12,13 The GnPs used in this work areabout ∼5 µm (Fig. 2a) in lateral size and ∼60 nm in averagethickness. They were produced through mechanical cleavage of

raw graphite and have relatively low defect density. This is seenin the Raman analysis as a low intensity of the D peak (Fig. 2b).The oxygen content of the used GnPs was measured to be lessthan 2.9% through X-ray photoelectron spectroscopy (Fig. 2c).The polymer matrix used in this work was high density poly-ethylene with a melt index of 2.2 g per 10 min and with amolecular weight of Mn ∼ 10 600. PE–GnP composites wereprepared by melt-compounding29–31 PE pellets and GnP nano-powder together in a DSM-5 micro-compounder. A mixingtime of 40 min and a mixing temperature of 200 °C were usedwhile preparing the composite through melt-compounding.The resulting nanocomposite was then compression-moldedto achieve the composite samples with a thickness of ∼1 mm.GnPs used in this work are not functionalized; the interactionbetween GnPs and the polymer matrix is therefore through vander Waals forces.

Mechanical strain was applied to the composite samplesusing a motorized slide to induce the alignment of the gra-phene nanoplatelets and polymer chains. The use of low drawrates (∼40 µm min−1) coupled with heating the sample to60–70 °C during the drawing process prevented brittle failureand allowed total strains of up to 400% to be achieved. Fig. 2dand e show an unstrained and strained PE–GnP composite,respectively.

The thermal conductivity of both the unstrained andstrained composites was measured through the measurementof the thermal diffusivity (α), specific heat capacity (cp) anddensity (ρ), using the relationship k = ρcpα. The specific heatcapacity of composites was measured using a TA instrumentsQ-2000 differential scanning calorimeter. The density ofsamples was measured using Archimedes’ principle. Thermaldiffusivity was measured using the Angstrom method32 byapplying a sinusoidal heat signal at one end of the sample andmeasuring the temperature (T ) response at two differentlocations along the sample. The accuracy of the measurementset-up was established through the good agreement betweenthe measured k of the pristine PE (0.5 W mK−1, Fig. 3) andthat reported in the literature.11

The measured thermal conductivities of pure PE and PE–GnP composites as a function of applied strain (ε = Δl/lo,where Δl is the change in sample length and lo is the originallength) are shown in Fig. 3. The thermal conductivities of boththe pure PE sample and the composite samples increase withapplied strain due to the alignment of PE chains andembedded GnPs. An extraordinary increase in k for alignedPE–GnP composites can be observed. At the highest strain ofε = 4, a k value of 5.9 W mK−1 is achieved for the alignedPE–GnP composite with 10 wt% GnP. This represents ak-enhancement of almost 1100% relative to the unorientedpure PE matrix (k ∼ 0.5 W mK−1). In addition, k of the 7 wt%composite at a strain of ε = 4 is measured to be 5.1 W mK−1,almost 5 times higher compared to the value for theunstrained composite (ε = 0) k = 1.1 W mK−1. These resultsshow significant enhancements in k of the aligned versus theunoriented composites and point to the great potential ofalignment effects in yielding high k polymeric materials. The

Fig. 1 (a) Randomly oriented polymer lamellae in a semi-crystallinepolymer, (b) alignment induced by strain and (c) aligned polymer-GnPcomposite.

Fig. 2 (a) SEM image of a typical graphene nanoplatelet used in thestudy. (b) Raman and (c) XPS analysis of GnPs. (d) Unstrained and (e)strained PE–GnP composite.

Communication Nanoscale

12868 | Nanoscale, 2017, 9, 12867–12873 This journal is © The Royal Society of Chemistry 2017

results provide the first evidence that simultaneous alignmenteffects can provide large enhancements in composite k.

While the above values provide an estimate of the absolutek-enhancement in aligned systems, direct evidence of thebeneficial effect of aligned GnPs in enhancing k is gained bycomparing the slopes of the best-fit lines for the threesamples. For the pure PE sample, a slope of k-increase withrespect to applied strain is found to be (dk/dε)pure-PE ≈ 0.67. Inthe pure PE sample, this k enhancement with respect to theapplied strain is solely due to the alignment of PE chains andhas been well studied before.33 As 7 wt% GnPs are added tothe sample, the slope increases to (dk/dε)composite ≈ 1.03. Inthe composite sample, k-enhancement with applied strain isthrough the alignment of both PE chains and GnPs. The largerslope is a direct manifestation of the beneficial effect ofaligned GnPs in enhancing composite k (if the aligned GnPsdid not cause any additional increase in k relative to randomlyoriented GnPs, the only effect responsible for k-increase withrespect to applied strain in the composite would be the align-

ment of PE chains, this would have resulted in the same slopeas that of the pure PE sample assuming that the alignment ofthe PE was the same in the pure and composite samples).As the weight fraction of GnPs in the composite is increasedto 10 wt%, the slope increases to (dk/dε)composite ≈ 1.24.This points to the enhanced effect of alignment of larger quan-tities of the GnPs within the composite. This almost a factor oftwo higher rate of k-increase with applied strain in the PE–GnP(10 wt%) composite relative to the pure PE sample is dueto the cumulative contributions of the alignment of PEchains and GnPs to k enhancement and clearly points to thepromise of such simultaneous alignment in enhancingcomposite k.

Quantitative estimate of the effectiveness of aligned versusrandomly oriented GnPs can also be achieved by comparingthe difference between the thermal conductivity values of thepure PE sample and composite at the same strain. At zerostrain, the difference between the two is due to randomlyoriented GnPs. At higher strains, the difference between thetwo k values is through GnPs which are now aligned providinga direct comparison of the k enhancement between alignedGnPs and randomly oriented GnPs. For the 7 wt% sample, theenhancement at zero strain is kε¼0

composite − kε¼0PE = 0.5 W mK−1.

As the strain is increased to ε = 3, the difference becomeskε¼3composite − kε¼3

PE = 1.7 W mK−1. Comparing these two enhance-ment values, aligned GnPs are found to be more than 3 timesas effective as randomly oriented GnPs in enhancingthe thermal conductivity. Similar results are foundcomparing enhancements for the 10 wt% sample. Thisanalysis clearly demonstrates the increased effectiveness ofGnPs in enhancing k in the aligned form versus randomlyoriented configuration.

We further compare the effectiveness of mechanical strainmediated alignment in enhancing k with that through othermeans such as electric field. In Table 1, we compare theresults of k-enhancement through GnPs aligned by differenttechniques.14–19,34,35 In the first study in Table 1, GnPs withvolume fractions of up to ∼1.08% were aligned through anelectric field in epoxy composites.15 At low volume fractions∼0.6%, k-enhancement through aligned GnPs was almosttwice that due to the randomly oriented case. At highervolume fractions (∼1%), however, the effectiveness of the elec-tric field in aligning GnPs diminished due to the presence of

Fig. 3 Measured thermal conductivities of pure PE and PE–GnP com-posites as a function of applied strain. A highest k of 5.9 W mK−1 isachieved representing a 12-fold improvement over pristine PE.Increasing slopes of linear-fits point to the role of aligned GnPs inenhancing k.

Table 1 Thermal conductivity enhancement through alignment effects

Sample composition k (W mK−1) Fraction Alignment method Ref.

Aligned GnP/epoxy 0.45 1.05 vol% Electric field 15Aligned-interconnected GnP/PDMSa 2.13 0.92 vol% Freeze casting of self-aligned GOc liquid crystal 18Aligned GnP/epoxy 0.6 1 vol% Magnetic field 14Aligned GnP/epoxy 0.37 0.52 vol% Magnetic field 34Aligned GOc/cellulose 6.17 (3.56) 30 wt% (10 wt%) Self-alignment 19Aligned GnP/PDMSa 6.05 (1.04) 20 vol% (7.5 vol%) Electric field 16Aligned GnP/PVDFb 10 (1.06) 25 vol% (5 vol%) Flow induced 17Aligned GnP/aligned PE 5.9 10 wt% (6.2 vol%) Mechanical strain This work

a PDMS: polydimethylsiloxane. b PVDF: polyvinylidene fluoride. cGO: graphene oxide.

Nanoscale Communication

This journal is © The Royal Society of Chemistry 2017 Nanoscale, 2017, 9, 12867–12873 | 12869

GnP agglomerates, leading to smaller k-enhancements. Thealignment of only low volume fractions (<1 vol%) limited thehighest k value achieved to 0.45 W mK−1. In this work, we findthat k-enhancements for 7 wt% and 10 wt% GnP content areproportional to the volume fractions, indicating no decrease inthe ability of mechanical strain to align greater quantities ofGnPs. This points to the advantages of mechanical strainrelative to an electric field in achieving higher thermal con-ductivity values. The results suggest that even higher k valuescan be achieved through alignment of larger quantities of GnPusing mechanical strain induced alignment.

Aligned GnPs are also found to be more effective in enhan-cing k relative to aligned CNTs through comparison with pre-viously reported results on aligned PVA (polyvinyl alcohol)–CNT composites.36 For the 10 wt% CNT composite and at astrain of 4.5, aligned CNTs were found to enhance thermaldiffusivity by αε¼4:5

PVA–CNT − αε¼4:5PVA = 0.55 mm2 s−1 relative to pure

PVA (at the same strain of 4.5) in the PVA–CNT composites. Inthis work, the same comparison for 10 wt% GnP compositeyields an enhancement of αε¼4:0

PE–GnP − αε¼4:0PE = 1.13 mm2 s−1

through the use of aligned GnPs. Aligned GnPs are thus foundto be twice as effective in enhancing k as aligned CNTs.

The above comparisons shed light on the advantages ofmechanical strain and the use of GnPs for alignment mediatedthermal conductivity enhancement. We next show that simul-taneous alignment of polymer lamellae and GnPs provideshigher k enhancement than achievable through alignment ofGnPs alone. In Table 1 we further compare k-enhancementthrough the alignment of GnPs induced by techniques such asmagnetic field, flow processes and self-alignment with thepresent work. For each study, we present the highest kachieved, and also the k values at volume fractions close to the6.2 vol% used in this study (volume fraction for the 10 wt%composition is ∼6.2 vol%, determined through densitymeasurements). Comparison of these studies clearly showsthat the k value achieved in this work (5.9 W mK−1) throughsimultaneous alignment of PE chains and GnPs (10 wt%) issignificantly higher than achieved through the alignment ofGnPs alone at similar compositions. These comparisons shedlight on the great potential of simultaneous alignment effectsin developing high k polymeric composites.

To analyze the influence of GnP alignment on the compo-site thermal properties, the degree of its alignment must bedetermined. The alignment of GnPs within the nanocomposite

was analyzed in this work through the use of laser scanningconfocal microscopy37 (LSCM) and polarized Raman spec-troscopy. While LSCM provides a visual understanding of GnPalignment, polarized Raman allows a quantitative estimate ofthe graphene orientation. LSCM generates optical sections asthin as 300 nm axially by using reflected light to create imageswhile blocking out any out-of-focus light. By collecting a seriesof these optical sections along the optical axis, one can gene-rate a 3D reconstruction of a volume within an intact speci-men. In this work, we use a Leica SP8 laser scanning confocalmicroscope with a 561 nm diode pumped solid state (DPSS)laser. The samples are imaged with a 63 × 1.4 oil immersionobjective with a pinhole aperture at 0.2 airy units (AU) andvoxel dimensions of 120 nm × 120 nm × 120 nm. The imagevolume for each sample was 246 μm × 246 μm × 10 μm withthree images per sample. Fig. 4 shows the LSCM maximumintensity projections of GnPs in the unstrained nanocompositesample (7 wt% GnP) and the same composition sample withstrains (ε) of 1, 2, 3 and 4. While GnPs are seen to be randomlyoriented in Fig. 4a for the unstrained sample, increasing align-ment of GnPs with increasing applied strain is clearly visiblein Fig. 4b, c, d and e. These LSCM images visually demonstratethe alignment of GnPs upon application of strain.

The absolute alignment of GnPs in terms of their angle oforientation is also measured using polarized Raman spec-troscopy. Raman spectra were obtained using a RenishawinVia Reflex Spectrometer System, equipped with a laser withwavelength λ = 532 nm. To measure orientation using polar-ized Raman spectroscopy, we used the 7 wt% graphene/PEsample with a strain of ε = 4. The sample was set up with thelaser illuminating along the X direction, as shown in Fig. 5(a).The orientation of GnPs is described by the angle ϕi betweenthe surface normal of the composite sample and the surfacenormal of an arbitrary GnP flake (Fig. 5(a)). The angle betweenthe incident laser polarization and the Y axis when viewedalong the X axis is defined as Φ (Fig. 5(b)). The scattered radi-ation was not polarized in this work. During the measurement,the incident laser polarization was rotated, while the samplewas kept fixed. The intensity of the Raman 2D band (I2D) as afunction of Φ measured at five areas is shown in Fig. 5(c). Foreach area, the intensity values at different Φ were normalizedto the maximum. To quantify the orientation of GnPs, themodel in ref. 38 is modified to account for the absence ofpolarization of scattered radiation in this work. The modified

Fig. 4 Confocal microscopy images of GnPs in the 7 wt% PE–GnP composite sample for different applied strains (ε) of (a) 0, (b) 1, (c) 2, (d) 3 and(e) 4. Increasing alignment of GnPs with increasing strain is clearly visible in the above maximum intensity projections. Scale bar = 20 μm.

Communication Nanoscale

12870 | Nanoscale, 2017, 9, 12867–12873 This journal is © The Royal Society of Chemistry 2017

model is used to express the Raman intensity as a function ofangle Φ and the parameter ⟨P2(cos ϕi)⟩ as below,

I2DðΦÞ ¼ I023þ hP2ðcosϕiÞi � 2

3þ cos2 Φ

� �� �ð1Þ

where Io is the amplitude. ⟨P2(cosϕi)⟩ describes the average orien-tation of GnPs through the relation ⟨P2(cosϕi)⟩ = {3⟨cos

2ϕi⟩ − 1}/2,where ⟨cos2ϕi⟩ is the average value of cos2ϕi. By fitting themeasured result to eqn (1), ⟨P2(cosϕi)⟩ can be obtained, as an indi-cation of the spatial orientation of GnPs as shown in Fig. 5c. Thevalues of ⟨P2(cosϕi)⟩ are 0 and 1 for random and completelyaligned orientations, respectively (corresponding values of⟨cos2ϕi⟩ are 0.33 and 1.0). The value of ⟨cos2ϕi⟩ obtained byfitting eqn (1) to the measured intensity data is achieved to be 0.47(Fig. 5c). This large value (⟨cos2ϕi⟩ ∼ 0.47) indicates a high level ofGnP alignment in the strained composite sample. We find thatthis orientation level of GnPs approaches the alignment level of PEcrystallites discussed next.

While the above measurements shed light on the alignmentof GnPs, the alignment of crystalline PE chains is measuredthrough wide-angle X-ray diffraction measurements (WAXS).Through WAXS, the average relative orientation of the PEchains in crystallites with respect to the stretch direction ismeasured in terms of ⟨cos2 γ⟩ (the average value of cos2 γ)where γ is the angle between the PE chain axis and draw direc-tion. The WAXS patterns in the pure PE sample obtained for

different strains are shown in Fig. 6a–c. The 2D patternsclearly show alignment effects; upon stretching, the ringsbecome arcs in WAXS spectra.39 The obtained intensities fromthese plots are integrated to compute the average orientation⟨cos2 γ⟩, giving the alignment of PE chains in crystallites withrespect to the stretch direction. The value of ⟨cos2 γ⟩ is pre-sented in Fig. 6(d) where the increasing value of ⟨cos2 γ⟩ withincreasing strain indicates increasing alignment. A highdegree of PE alignment at the largest strain (ε = 4) is indicatedby a large value of ⟨cos2 γ⟩ = 0.87.

WAXS measurements were also performed on the 7 wt%composite sample for strain ε = 4 to compare PE alignmentlevels with the pure PE samples. The value of ⟨cos2 γ⟩ was iden-tical to the one obtained for the pure PE sample, indicatingthat PE chain orientation is identical in the composite and thepure PE sample for the same applied strain. This also impliesthat the small GnP content of 4.4 vol% (for the 7 wt% sample)does not impact the alignment of polymer lamellae.

We further use Raman and WAXS alignment measurementsto extract the relative contributions of aligned PE chains andGnPs to total k-enhancement. For 10 wt% filler content,increasing the strain from ε = 0 to ε = 2 leads to a totalk-enhancement of 2.9 W mK−1 (Fig. 3). The contribution ofalignment of PE chains to this is considered to be the same asin the pure PE sample, 1.1 W mK−1 from the differencebetween the k values of the pure PE sample for ε = 2 and ε = 0.This is due to the same PE alignment in both the compositeand the pure PE sample (as shown by the above WAXSmeasurements). The further enhancement induced in compo-site k due to the alignment of GnPs is thus 1.8 W mK−1

(obtained by subtracting the PE contribution from totalenhancement). For a strain of ε = 3, the total k-enhancementreaches 3.9 W mK−1. The contributions of the aligned PEchains and aligned GnPs to this are estimated to be 1.9 and2.0 W mK−1, respectively. These near equal contributions ofthe alignment of PE and GnPs to total k-enhancement lead toan almost doubling of the slope of k-increase with the strainfor the 10 wt% composite relative to the pure PE sample and

Fig. 5 Schematic illustration of (a) set-up of the sample with threecoordinates defining the spatial position of a graphene flake, and alsowith the laser propagation direction. ϕi is the angle between the surfacenormal of the sample and the surface normal of an arbitrary flake; (b)side view along the X axis, Φ is the angle between the laser polarizationdirection and the Y axis, (c) I2D as a function of Φ of five areas of the7 wt% graphene/PE sample with a strain of 4. The red line correspondsto the curve fitting of all the points from the 5 different areas.

Fig. 6 (a, b and c) WAXS patterns of strains of 0, 2 and 4, in the pure PEsample, (d) average orientation of PE chains in crystallites with respectto the draw direction.

Nanoscale Communication

This journal is © The Royal Society of Chemistry 2017 Nanoscale, 2017, 9, 12867–12873 | 12871

yield overall k-value of 5.9 W mK−1 These results provide thefirst proof that strain induced higher k-enhancement via mech-anical stretching can occur through combined contributionsof the alignment of the polymer matrix and the embedded gra-phitic nanoplatelets and may provide new pathways for achiev-ing high k polymeric materials.

It should be noted that alignment levels presented in thiswork can be realized on a bulk scale through widely usedmanufacturing processes such as polymer extrusion. In thepast, extrusion has been found to align both polyethylene40

and polypropylene41,42 chains. The alignment of filler materialsuch as CNTs43,44 and GnPs45 upon extrusion has also beenreported. The simultaneous alignment of polymer chains andhigh k filler material such as GnPs through composite extru-sion may lead to inexpensive and scalable methods for proces-sing of high thermal conductivity polymer composites.

In summary, we have demonstrated the effect of the straininduced simultaneous alignment of PE chains and GnPs onthe k-enhancement of PE–GnP composites. These compositeswere prepared by melt-compounding PE pellets and GnPs.GnP weight fractions of up to 10% were included in the study.Both laser scanning confocal microscopy and polarizedRaman spectroscopy demonstrated a high level of GnP align-ment in strained samples. Similarly, wide-angle X-ray scatter-ing measurements revealed the alignment of PE chains. At thehighest strain of 400% and for 10 wt% GnP composition, anoverall thermal conductivity of 5.9 W mK−1 was achieved repre-senting an 1100% enhancement over the k of unoriented pris-tine PE (0.5 W mK−1). Aligned GnPs were found to be 3 timesas effective in enhancing k as randomly oriented GnPs. Asimilar comparison with aligned CNTs showed the alignedGnPs to be twice as effective as aligned CNTs in enhancingcomposite k. These results shed light on the great potential ofsimultaneous alignment effects in enhancing k and providenew avenues for developing high k polymeric composites.

Conflicts of interest

There are no conflicts of interest.

References

1 X. Chen, Y. Su, D. Reay and S. Riffat, Recent Research devel-opments in polymer heat exchangers-A review, RenewableSustainable Energy Rev., 2016, 60, 1367.

2 D. A. Reay, The use of polymers in heat exchangers, HeatRecovery Syst. CHP, 1989, 9, 209.

3 C. Dreiser, L. J. Kratz and H.-J. Bart, in InternationalConference on Heat exchanger fouling and cleaning, ed.H. Muller-Steinhagen, A. P. Watkinson and M. R. Malayeri,2013, p. 296.

4 H. T. El-Dessouky and H. M. Ettouney, Plastic/compact heatexchangers for single-effect desalination systems,Desalination, 1999, 122, 271.

5 W. Liu, J. Davidson and S. Mantell, Thermal analysis ofpolymer heat exchangers for solar water heating: A casestudy, J. Sol. Energy Eng., 2000, 122, 84.

6 S. Mallik, N. Ekere, C. Best and R. Bhatti, Investigation ofthermal management materials for automotive electroniccontrol units, Appl. Therm. Eng., 2011, 31, 355.

7 P. Procter and J. Solc, Improved Thermal Conductivity inMicroelectronic Encapsulants, IEEE Trans. Compon.,Hybrids, Manuf. Technol., 1991, 14, 708.

8 X. Lu and G. Xu, Thermally Conductive PolymerComposites for Electronic Packaging, J. Appl. Polym. Sci.,1997, 65, 2733.

9 Z. Lin, Y. Liu, K. Moon and C. Wong, in Electronic com-ponents and Technology Conference, 2013, p. 1692.

10 Z. Lin, Y. Yao, A. Mcnamara, K. Moon and C. P. Wong, inElectronic components and Technology Conference, 2012,p. 1437.

11 J. Yu, B. Sundqvist, B. Tonpheng and O. Andersson,Thermal conductivity of highly crystallized polyethylene,Polymer, 2014, 55, 195.

12 S. Shen, A. Henry, J. Tong, R. Zheng and G. Chen,Polyethylene nanofibres with very high thermal conduc-tivities, Nat. Nanotechnol., 2010, 5, 251–255.

13 H. Ghasemi, N. Thoppey, X. Huang, J. Loomis, X. Li,J. Tong, J. Wang and G. Chen, in 14th IEEE ITHERMConference, Orlando, FL, 2014.

14 H. Yan, Y. Tang, W. Long and Y. Li, Enhanced thermal con-ductivity in polymer composites with aligned graphenenanosheets, J. Mater. Sci., 2014, 49, 5256.

15 S. Wu, R. B. Ladani, J. Zhang, E. Bafekrpour, K. Ghorbani,A. P. Mouritz, A. J. Kinloch and C. H. Wang, Aligning multi-layer graphene flakes with an external electric field toimprove multifunctional properties of epoxy nano-composites, Carbon, 2015, 94, 607.

16 Y. Guo, Y. Chen, E. Wang and M. Cakmak, Roll-to-Roll con-tinuous manufacturing multifunctional nanocompositesby electric-field-assisted “Z” direction alignment of graph-ite flakes in poly(dimethysiloxane), ACS Appl. Mater.Interfaces, 2017, 9, 919.

17 H. Jung, S. Yu, N. Bae, S. Cho, R. H. Kim, S. H. Cho,I. Hwang, B. Jeong, J. S. Ryu, J. Hwang, S. M. Hong,C. M. Koo and C. Park, High through-plane thermal con-duction of graphene nanoflake filled polymer compositesmelt-processed in an L-shape kinked tube, ACS Appl. Mater.Interfaces, 2015, 7, 15256.

18 G. Lian, C.-C. Tuan, L. Li, S. Jiao, Q. Wang, K. S. Moon,D. Cui and C.-P. Wong, Vertically aligned and inter-connected graphene networks for high thermal conductivityof epoxy composites with ultralow loading, Chem. Mater.,2016, 28, 6096.

19 N. Song, D. Jiao, P. Ding, S. Cui, S. Tang and L. Shi,Anisotropic thermally conductive flexible films based onnanofibrillated cellulose and aligned graphene nanosheets,J. Mater. Chem. C, 2016, 4, 305.

20 W. Zhao, J. Kong, H. Liu, Q. Zhuang, J. Gu and Z. Guo,Ultra-high thermally conductive and rapid heat responsive

Communication Nanoscale

12872 | Nanoscale, 2017, 9, 12867–12873 This journal is © The Royal Society of Chemistry 2017

poly (benzobisoxazole) nanocomposites with self-alignedgraphene, Nanoscale, 2016, 8, 19984.

21 A. A. Balandin, S. Ghosh, W. Bao, I. Calizo,D. Teweldebrhan, F. Miao and C. N. Lau, Superior thermalconductivity of single-layer graphene, Nano Lett., 2008, 8,902.

22 S. Ghosh, W. Bao, D. L. Nika, S. Subrina, E. P. Pokatilov,C. N. Lau and A. A. Balandin, Dimensional crossover ofthermal transport in few-layer graphene, Nat. Mater., 2010,9, 555.

23 G. Fugallo, A. Cepellotti, L. Paulatto, M. Lazzeri, N. Marzariand F. Mauri, Thermal conductivity of graphene and graph-ite: Collective excitations and mean free paths, Nano Lett.,2014, 14, 6109.

24 K. M. F. Shahil and A. A. Balandin, Graphene-Multilayergraphene nanocomposites as highly efficient thermal inter-face materials, Nano Lett., 2012, 12, 861.

25 S. T. Huxtable, D. G. Cahill, S. Shenogin, L. Xue, R. Ozisik,P. Barone, M. Ursey, M. S. Strano, G. Siddons, M. Shim andP. Keblinski, Interfacial heat flow in carbon nanotube sus-pensions, Nat. Mater., 2003, 2, 731.

26 A. A. Balandin, Thermal properties of graphene and nano-structured carbon materials, Nat. Mater., 2011, 10, 569.

27 S. L. Aggarwal, G. P. Tilley and O. J. Sweeting, Changes inorientation of crystallites during stretching and relaxationof polyethylene films, J. Polym. Sci., 1961, 51, 551.

28 S. H. Yao, J. K. Yuan, T. Zhou, Z. M. Dang and J. Bai,Stretch-Modulated Carbon Nanotube Alignment inFerroelectric PolymerComposites: Charctaerization of theOrientation State and its influence on the dielectric pro-perties, J. Phys. Chem. C, 2011, 115, 20011.

29 J. R. Potts, D. R. Dreyer, C. W. Bielawski and R. S. Rwoff,Graphene based polymer nanocomposites, Polymer, 2011,52, 5–25.

30 H. Kim, S. Kobayashi, M. A. AbdurRahim, M. J. Zhang,A. Khsainova, M. A. Hillmyer, A. A. Abdala andC. W. Macoska, Graphene/polyethylene nanocomposites:Effect of polyethylene functionalization and blendingmethods, Polymer, 2011, 52, 1837–1846.

31 G. Ozkoc, G. Bayram and J. Tiesnitsch, Microcompundingof organoclay-ABS/PA6 blend based nanocomposites,Polym. Compos., 2008, 29, 345–356.

32 Y. Zhu, Heat-loss modified Angstrom method for simul-taneous measurements of thermal diffusivity and conduc-tivity of graphite sheets:The origins ofheat loss inAngstrom method, Int. J. Heat Mass Transfer, 2016, 92, 784.

33 D. B. Mergenthaler and M. Pietralla, Thermal conductivityin ultra-oriented polyethylene, Macromolecules, 1992, 25,3500.

34 H. Yan, R. Wang, Y. Li and W. Long, Thermal conductivityof magnetically aligned graphene-polymer composites withFe3O4-decorated graphene nanosheets, J. Electron. Mater.,2015, 44, 658.

35 S. N. Leung, M. O. Khan, H. Naguib and F. Dawson,Multifunctional polymer nanocomposites with uniaxiallyaligned liquid crystal polymer fibrils and graphene nano-platelets, Appl. Phys. Lett., 2014, 104, 081904.

36 W.-L. Song, W. Wang, L. M. Veca, C. Y. Kong, M.-S. Cao,P. Wang, M. J. Meziam, H. Qian, G. E. LeCroy, L. Cao andY. P. Sun, Polymer/carbon nanocomposites for enhancedthermal transport properties - carbon nanotubes versusgraphene sheets as nanoscale fillers, J. Mater. Chem., 2012,22, 17133.

37 H. Yazdani, B. E. Smith and K. Hatami, Multi-walledcarbon nanotube-filled polyvinyl chloride composites:Influence of processing method on dispersion quality, elec-trical conductivity and mechanical properties, Composites,Part A, 2016, 82, 65–77.

38 Z. Li, R. J. Young, N. R. Wilson, I. A. Kinloch, C. Vallés andZ. Li, Effect of the orientation of graphene-based nano-platelets upon the Young’s modulus of nanocomposites,Compos. Sci. Technol., 2016, 123, 125–133, DOI: 10.1016/j.compscitech.2015.12.005.

39 W. B. Genetti, R. J. Lamirand and B. P. Grady, Wide-AngleX-ray scattering study of crystalline orientation in reticula-tep-doped polymer composites, J. Appl. Polym. Sci., 1998,70, 1785.

40 K. Nakayama and H. Kanetsuna, Hydrostatic extrusion ofsolid polymers, J. Mater. Sci., 1975, 10, 1105.

41 C. H. Wang and D. B. Cavannaugh, Effect of chain align-ment on the Brillouin scattering spectra of hydrostaticallyextruded polypropylene, Macromolecules, 1981, 14, 1061.

42 X. Chen, M. A. Leugers, T. Kirch and J. Stanley, Orientationmapping of extruded polymeric composites by polarizedmicro-raman spectroscopy, J. Spectrosc., 2015, 2015,518054.

43 A. B. Sulong and J. Park, Alignment of multi-walled carbonnanotubes in a polyethylene matrix by extrusion shearflow: mechanical properties enhancement, J. Compos.Mater., 2010, 45, 931.

44 V. K. Rangari, M. Yousuf, S. Jeelani, M. X. Pulikkathara andV. N. Khabashesku, Alignment of carbon nanotubes andreinforcing effects in nylon-6 polymer composite fibers,Nanotechnology, 2008, 19, 245703.

45 K. Gaska, X. Xu, S. Gubanski and R. Kaddar, Electrical,mechanical and thermal properties of LDPE graphenenanoplatelets composites produced by means of meltextrusion process, Polymers, 2017, 9, 11.

Nanoscale Communication

This journal is © The Royal Society of Chemistry 2017 Nanoscale, 2017, 9, 12867–12873 | 12873