Published: April 11, 2011

r 2011 American Chemical Society 8753 dx.doi.org/10.1021/jp200838s | J. Phys. Chem. C 2011, 115, 8753–8758

ARTICLE

pubs.acs.org/JPCC

Enhanced Thermal Conductivity in a NanostructuredPhase Change Composite due to Low ConcentrationGraphene AdditivesFazel Yavari,||,† Hafez Raeisi Fard,||,† Kamyar Pashayi,† Mohammad A. Rafiee,† Amir Zamiri,†,*Zhongzhen Yu,§ Rahmi Ozisik,‡ Theodorian Borca-Tasciuc,†,* and Nikhil Koratkar†,*†Department of Mechanical, Aerospace, and Nuclear Engineering, ‡Department of Materials Science and Engineering, RensselaerPolytechnic Institute, 110 Eighth Street, Troy, New York 12180-3590, United States§State Key Laboratory of Organic-Inorganic Composites, College of Materials Science and Engineering, Beijing University of ChemicalTechnology, Beijing 100029, People's Republic of China

’ INTRODUCTION

One of the main sources of energy in nature is thermal energy,which could be obtained from sources such as the sun, geother-mal fields, and oceans. This type of energy is also released aswaste heat from many man-made devices such as power plants,reactors, and engines. There is a large interest to develop efficientsystems or materials to scavenge thermal energy and reuse iteither directly or convert it to another type of useful energy suchas electricity.1�3 Solid�liquid organic phase change materials(PCMs) have drawn a lot of attention4,5 because they possess aconsiderable heat of fusion of the order of 100s J/g. PCMs can beused as latent heat storage and release units for thermal manage-ment of computers, electrical engines, solar power plants, and forthermal protection of electronic devices.4 However, a majordrawback of organic PCMs is their low thermal conductivity(κ) which leads to large temperature gradients during heattransfer in or out of the material, reduced heat transfer rates,and large time constants. Therefore, increasing the thermalconductivity of the solidified material is one of the main issuesin the application of organic PCMs.3 However, although theconductivity needs to be improved, it has to be achieved whilemaintaining the phase change enthalpy. To achieve this goal,

several methods have been developed including improvingencapsulation techniques,6 mixing the PCMs with highconductive fillers such as carbon nanotubes7 and silver nano-wires,8 and dispersing PCMs in highly conductive cellularstructures.9,10

Graphene, a recently discovered form of carbon that consistsof a sheet of carbon atoms arranged in a honeycomb lattice,11

exhibits a number of fascinating properties including very highthermal conductivity. The value of thermal conductivity (κ) forsingle layer graphene is reported to be in the range of 4840�5300W/mK,12 which is more than an order of magnitude higher thanthat of copper. Graphene is a promising thermally conductivefiller because of its ultrahigh thermal conductivity and low density.Nanosheets of graphene have previously been used to improvethermal conductivity of different organic materials13 such asepoxy,14�18 polypropylene,19 polystyrene,20 polyethylene, andpolyamide.21

Received: January 26, 2011Revised: March 23, 2011

ABSTRACT: The liquid�solid phase change enthalpy, crystallization,and thermal conductivity of graphene/1-octadecanol (stearyl alcohol)composite, a nanostructured phase change material, was investigated as afunction of graphene content. The thermal conductivity (κ) of thenanocomposite increased by nearly 2.5-fold (∼140% increase) upon∼4% (by weight) graphene addition while the drop in the heat of fusion(i.e., storage capacity) was only ∼15.4%. The enhancement in thermalproperties of 1-octadecanol obtained with the addition of graphene ismarkedly superior to the effect of other nanofillers such as silvernanowires and carbon nanotubes reported previously in the literature.Boosting the thermal conductivity of organic phase change materialswithout incurring a significant loss in the heat of fusion is one of the keyissues in enabling their practical application as latent heat storage/releaseunits for thermal management and thermal protection.

8754 dx.doi.org/10.1021/jp200838s |J. Phys. Chem. C 2011, 115, 8753–8758

The Journal of Physical Chemistry C ARTICLE

In the current work, we are proposing to use nanosheetsof graphene to improve the thermal conductivity of PCMs.Although the conductivity of the PCM/graphene nanocompositewould be higher than that of the PCM alone, which will help theheat transfer rate, it is expected that the phase change enthalpywould be lower than that of PCM because some of the PCMvolume would be replaced by the graphene sheets that do notundergo phase change in the operating temperature range.Therefore, it is important that the enhancement in the thermalconductivity of the PCM should be done with as small as possiblea sacrifice in the phase change enthalpy of the PCM. We show inthis work that the large increase in thermal conductivity coupledwith a small reduction in the heat of fusion for graphene nano-composites is far superior when compared with the publishedliterature for other nanofillers such as carbon nanotubes andsilver nanowires.

’MATERIALS AND EXPERIMENTAL SECTION

The graphene fillers used in this work were obtained fromgraphite using the method developed in ref 11. In this method,graphite oxide is prepared by oxidizing graphite in a solution ofsulfuric acid, nitric acid, and potassium chlorate for 96 h. Partiallyoxidized graphene sheets are then generated by the rapid thermalexpansion (>2000 �C/min) of the graphite oxide. A transmissionelectron microscope (TEM) image of a graphene plateletproduced by this method is shown in Figure 1a. The grapheneflakes have relatively wrinkled surface texture, which could play abeneficial role in enhancing the interlocking of the flakes witheach other and enable strong interaction with the matrix.22

Figure 1b is a high-resolution transmission electron microscopy(HRTEM) image of the graphene platelet demonstrating thelayered structure of the platelet (inset shows the electrondiffraction image).

The PCM used in the current work is 1-octadecanol (stearylalcohol), with a melting temperature of∼66 �C which is close toroom temperature and has an outstanding solid�liquid phasechange enthalpy (∼250 J/g). 1-octadecanol is nontoxic, has arelatively low density (0.812 g/cm3), and boils at ∼210 �C. Tofabricate 1-octadecanol/graphene composites, as seen inFigure 1c, graphene sheets were first dispersed in a 25 mL/mgacetone solution and were ultrasonicated for 15 min (10 s on, 5 soff, 50% power). Then the mixture was heated on a hot plate to120 �C while being sonicated for another 5 min. The 1-octade-canol was then mixed with the graphene dispersion and wassonicated for another ∼15 min on a hot plate at ∼120 �C. Themixture of acetone, graphene, and PCM were stirred and heatedto ∼150 �C to evaporate the remaining acetone. The nanocom-posite in liquid phase was then poured into preheated Si rubbermolds of cylindrical shape, ∼6.35 mm thick and ∼12.70 mm indiameter, and was left at room temperature to solidify for ∼20min. The resulting composite was affixed to a sample holder, cutto different thicknesses using a hot blade, and polished on a sandpaper pad.

’RESULTS AND DISCUSSION

The SEM images of the freeze-fractured surfaces of pristine1-octadecanol and 4% (by weight) graphene/1-octadecanol nano-composites are shown in Figure 2a,b, respectively. High magnifica-tion images of 4% nanocomposite are also shown in Figure 2c,d. Asseen in these figures, graphene flakes were dispersed uniformlythroughout the matrix providing a three-dimensional network of

high thermal conductivity graphene films. Good dispersion andnetwork formation facilitates heat transfer and allows phonons totravel efficiently through the graphene fillers and between the flakes.It is well-known that the type of the polymer matrix, degree ofexfoliation of the graphene flakes, orientation of the fillers, andinterfacial interaction influences the thermal transport in graphenenanocomposites.21 The SEM image of the fracture surface of anepoxy/graphene composite, fabricated using the same grapheneflakes, is shown in Figure 2e, which is comparable to Figure 2d interms of magnification. However, there is a drastic difference interms of the interaction of graphene with the surrounding polymer.In the case of graphene/1-octadecanol, the graphene sheets are stillcovered with a thick layer of polymer, whereas in the case of

Figure 1. (a) Transmission electron microscopy (TEM) image of agraphene flake obtained by thermal exfoliation of graphite oxidedeposited on a standard TEM grid. (b) HRTEM image of the edgesof the flake showing ∼3�4 individual graphene layers within theplatelet. The inset shows the measured electron diffraction pattern.(c) The fabrication process of PCM/graphene composites: (i) disper-sion of graphene in acetone using ultrasonic agitation, (ii) dispersion ofgraphene and PCM in acetone solution using heat and sonication, (iii)solvent evaporation using hot plate and stir magnet, (iv) insertion of themolten PCM in the mold, and (v) polishing the sample on sand paper.

8755 dx.doi.org/10.1021/jp200838s |J. Phys. Chem. C 2011, 115, 8753–8758

The Journal of Physical Chemistry C ARTICLE

graphene/epoxy, the graphene sheets seem to be completelyseparated from the epoxymatrix. This suggests that there is a stronginterface between graphene and 1-octadecanol molecules.

To measure the heat conductivity of the composites, a steady-state one-dimensional heat conduction method was used. Theexperimental setup consists of an electrical heater, a heat sinkand two thermocouples to measure the temperature gradient(Figure 3a). To minimize the interface thermal resistance, fine-diameter electrically insulated thermocouples were embedded intotwo soft indium layers to measure the temperature at both sides of athin cylindrical sample. Pressure is applied using a screwmechanismthat is thermally insulated from the sample by a thick Teflon block.The heat losses in the experimental setup were calibrated using glasssamples of known conductivity. The thermal conductivity of thepure 1-octadecanol measured with this setup (∼0.38 W/mK)matches the value reported in the literature.23 To measure κ, theexperimental thermal resistance is first obtained from the slope of

the temperature difference across the sample as a function of heaterpower (Figure 3b). Next, the calibrated heat loss (Rhl) contributionis accounted for by using a parallel thermal resistance networkmodel. To find the intrinsic thermal conductivity (κ), the interfacethermal resistance (Rint) between the composite sample and theindium layer must be subtracted from the overall conductionresistance (Rt). This interface thermal resistance was determinedby testing samples with different thicknesses, then extrapolatingthe plot of Rs þ Rint vs thickness to zero thickness using linearregression (Figure 3c). The thermal resistance and conductivity ofthe sample are calculated using the following equations:

Rs ¼ � Rint þ RtRhl

Rhl � Rtð1Þ

k ¼ 1RsA

� �� t ð2Þ

Figure 2. Scanning electron microscopy (SEM) image of (a) fracture surface of pristine 1-octadecanol, (b�d) fracture surface of ∼4% byweight graphene/1-octadecanol composite with different magnifications, and (e) fracture surface of ∼4% by weight graphene/epoxynanocomposite.

8756 dx.doi.org/10.1021/jp200838s |J. Phys. Chem. C 2011, 115, 8753–8758

The Journal of Physical Chemistry C ARTICLE

where t and A are the thickness and cross sectional area of thesample, respectively. The value of κ was obtained using theaforementioned method for the samples with different gra-phene contents. The results are shown in Figure 3d. Asexpected, thermal conductivity is considerably enhanced bythe presence of graphene and reaches ∼0.91 W/mK at ∼4%graphene content. This value of κ is about 2.5-times higherthan the measured κ value (∼0.38 W/mK) for pure 1-octa-decanol. This increase can be attributed to the high thermalconductivity of the network of graphene fillers that provide apath of lower resistance for phonons to travel. Also, the highaspect ratio and large interfacial contact area of graphene aswell as strong interface between graphene and the polymermay help to increase the thermal transport capacity ofgraphene/PCM nanocomposites.

Considering the very high theoretical conductivity ofgraphene (4840�5300 W/mK), one might expect a moredramatic improvement in thermal conductivity of the com-posite with the addition of graphene fillers. However, it hasbeen reported that since the dominant heat transfer mechan-ism is due to the lattice vibrations or phonons; poor phononcoupling in the vibrational modes at the polymer�filler andfiller�filler interfaces cause thermal resistance, also calledthe Kapitza resistance, which decreases the overall thermalconductivity of the material.24�27 Introduction of defectsassociated with the oxidation and thermal exfoliation ofgraphite can also reduce the thermal conductivity belowthe ideal value for defect-free graphene. Previous studieshave also shown that the thermal conductivity enhancementdue to the addition of conductive nanofillers differs depend-ing on the type of nanofiller and the polymer matrix. Forexample, an up to 4-fold increase in thermal conductivity canbe attained by adding 5% graphene platelets (by weight) into

epoxy28 while only ∼20%8 and ∼26%7 increase in thermalconductivity were observed by adding the same amount ofsilver nanowires and multi wall carbon nanotubes intoPCMs, respectively. Besides these factors, a minor mechan-ism affecting the heat transfer could be related to the changein the volume fraction of the crystalline phase of the polymer.The crystallinity of the organic PCM could be lowered by theaddition of graphene fillers.29 In the current work, thechanges in crystallinity and crystallization mechanism wereobserved via differential scanning calorimetry (Figure 4).The crystallization and melting of 1-octadecanol has beenwell studied.30 It was shown that 1-octadecanol has a stable γphase (monoclinic) crystal structure at room temperature,however melting transition does not proceed directly fromthe γ phase to liquid but an intermediate rotator phase exists.Therefore, during melting the γ phase transforms to therotator phase and upon further heating, the rotator phasemelts into the liquid state. During melting the solid�solid(γ-rotator) and solid�liquid (rotator-liquid) phase transi-tions occur very close to each other (within 1�2�), therefore,show up as a single melting peak. However, during crystal-lization, the two phase transitions appear as distinct peaks(8�10� apart from each other). In our results, these char-acteristics are clearly seen in Figure 4a,b; the nanocompositeshows two crystallization temperatures, both of these tem-peratures decrease with increasing filler content suggestingthat crystallization became more difficult in the presence ofgraphene. These results were highly reproducible and thedata from the first heating cycle were identical to thesubsequent ones.

The phase change enthalpy is a critical factor in PCMs.The phase change enthalpy could be used as a measure toevaluate the thermal energy storage capacity of the PCM.Figure 4c shows that with the addition of graphene platelets,the melting enthalpy of 1-octadecanol decreases. The de-crease is ∼15% in the case of ∼4% graphene filler content.This is expected given that some of the PCM volume is nowreplaced by the graphene sheets that do not undergo phasechange. We did not increase the graphene weight fractionbeyond 4% to avoid further reduction in the melting en-thalpy. In the case of silver nanowire/PCM nanocom-posites,8 in order to achieve the same (∼2�3-fold) increasein the thermal conductivity of graphene/PCM, about ∼45%silver nanowires (by weight) needs to be used, which alsoleads to∼50% decrease in phase change enthalpy, and hence,the heat storage capacity. Therefore, compared to silvernanowires, the graphene loading is an order of magnitudelower and the reduction in phase change enthalpy is also ∼3times lower than silver nanowires. Similarly, adding ∼2% byweight of multiwalled carbon nanotubes leads to ∼9%increase in thermal conductivity and ∼9% decrease in thephase change enthalpy of the PCM,7 whereas adding∼2% byweight of graphene platelets leads to ∼63% increase in thethermal conductivity and only ∼8.7% decrease in the phasechange enthalpy of the PCM used in the current work(Figure 3d and 4c). These comparisons with silver nanowiresand multiwalled carbon nanotubes indicate the superiorityof graphene as a conductive nanofiller for organic phasechange materials. We also measured the melting tempera-ture of 1-octadecanol as a function of graphene loading(Figure 4d); no significant change in the melting tempera-tures was observed.

Figure 3. (a) Experimental setup for thermal conductivity measure-ment; (b) temperature difference vs power for ∼4% graphene/PCMnanocomposite; (c) Rsþ Rint (sample resistanceþ interface resistance)as a function of thickness to determine indium-sample thermal interfaceresistance; and (d) measured thermal conductivity of the sample vs fillercontent for graphene/PCM composites.

8757 dx.doi.org/10.1021/jp200838s |J. Phys. Chem. C 2011, 115, 8753–8758

The Journal of Physical Chemistry C ARTICLE

’CONCLUSIONS

It is shown that adding graphene platelets to organic phasechange materials significantly boosts their thermal conductivitywithout incurring a large reduction in their liquid�solid phasechange enthalpy. Addition of ∼4% graphene (by weight) to1-octadecanol led to a ∼140% increase in thermal conductivitywith only ∼15% decrease in the phase change enthalpy. Theseimprovements were markedly superior to other nanofillers suchas multiwalled carbon nanotubes and silver nanowires. Theseresults indicate that graphene is a promising candidate for theenhancement of thermal conductivity of organic phase changematerials.

’AUTHOR INFORMATION

Corresponding Author*E-mail: zamira@rpi.edu, borcat@rpi.edu, koratn@rpi.edu.

Author Contributions

)These authors contributed equally.

’ACKNOWLEDGMENT

N.K. acknowledges funding support from the U.S. Office ofNaval Research (Award: N000140910928) and the U.S. NationalScience Foundation (Award: 0900188). T.B. acknowledgesfunding from the U.S. National Science Foundation (Award:0348613)

’REFERENCES

(1) DiSalvo, F. J. Thermoelectric Cooling and Power Generation.Science 1999, 285, 703–706.

(2) Hasnain, S. M. Review on Sustainable Thermal Energy StorageTechnologies, Part I: Heatt Storage Materials and Techniques. EnergyConvers. Manage. 1998, 39, 1127–1138.

Figure 4. (a) Differential scanning calorimetry (DSC) plots of graphene/1-octadecanol composites for various graphene concentrations showing themelting and crystallization events. (b) The change in the first and second crystallization temperature for graphene/1-octadecanol composites for variousgraphene concentrations. (c) The change in heat of fusion for graphene/1-octadecanol composites for various graphene concentrations. (d) Meltingtemperature of 1-octadecanol shown as a function of the graphene weight fraction in the composite.

8758 dx.doi.org/10.1021/jp200838s |J. Phys. Chem. C 2011, 115, 8753–8758

The Journal of Physical Chemistry C ARTICLE

(3) Velraj, R.; Seeniraj, R. V.; Hafner, B.; Faber, C.; Schwarzer, K.Heat Transfer Enhancement in a Latent Heat Storage System. Sol.Energy 1999, 65, 171–180.(4) Zalba, B.; Marín, J. M.; Cabeza, L. F.; Mehling, H. Review on

Thermal Energy Storage with Phase Change: Materials, Heat TransferAnalysis and Applications. Appl. Therm. Eng. 2003, 23, 251–283.(5) Abhat, A. Low-Temperature Latent-Heat Thermal-Energy

Storage—Heat Storage Materials. Sol. Energy 1983, 30, 313–332.(6) Wirtz, R.; Zhao, T.; Jiang, Y. Thermal and Mechanical Char-

acteristics of a Multi-Functional Thermal Energy Storage Structure. TheNinth Intersociety Conference on Thermal and Thermomechanical Phenom-ena in Electronic Systems IEEE Cat No04CH37543 2009, 32, 549–556.(7) Zeng, J. L.; Cao, Z.; Yang, D. W.; Xu, F.; Sun, L. X.; Zhang, X. F.;

Zhang, L. Effect of MWNTs on Phase Change Enthalpy and ThermalConductivity of a Solid-Liquid Organic PCM. J. Therm. Anal. Calorim.2009, 95, 507–512.(8) Zeng, J. L.; Cao, Z.; Yang, D. W.; Sun, L. X.; Zhang, L. Thermal

Conductivity Enhancement of Ag Nanowires on an Organic PhaseChange Material. J. Therm. Anal. Calorim. 2010, 101, 385–389.(9) Xiao, M.; Feng, B.; Gong, K. Preparation and Performance of

Shape Stabilized Phase Change Thermal Storage Materials with HighThermal Conductivity. Energy Convers. Manage. 2002, 43, 103–108.(10) Py, X; Olives, R; Mauran, S. Paraffin/Porous-Graphite-Matrix

Composite as a High and Constant Power Thermal Storage Material.Int. J. Heat Mass Transfer 2001, 44, 2727–2737.(11) Rafiee, J.; Rafiee, M. A.; Yu, Z.-Z.; Koratkar, N. Super-hydro-

phobic to Super-hydrophilic Wetting Control in Graphene Films. Adv.Mater. 2010, 22, 2151–2154.(12) Balandin, A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.;

Miao, F.; Lau, C. N. Superior Thermal Conductivity of Single-LayerGraphene. Nano Lett. 2008, 8, 902–907.(13) Cai, D.; Song, M. Recent Advance in Functionalized Graphene/

Polymer Nanocomposites. J. Mater. Chem. 2010, 20, 7906–7915.(14) Veca, L. M.; Meziani, M. J.; Wang, W.; Wang, X.; Lu, F.; Zhang,

P.; Lin, Y.; Fee, R.; Connell, J. W.; Sun, Y. P. Carbon Nanosheets forPolymeric Nanocomposites with High Thermal Conductivity. Adv.Mater. 2009, 21, 2088–2092.(15) Yu, A.; Ramesh, P.; Sun, X.; Bekyarova, E.; Itkis, M. E.; Haddon,

R. C. Enhanced Thermal Conductivity in a Hybrid Graphite Nanoplate-let—Carbon Nanotube Filler for Epoxy Composites. Adv. Mater. 2008,20, 4740–4744.(16) Gangulia, S.; Roya, A. K.; Anderson, D. P. Improved Thermal

Conductivity for Chemically Functionalized Exfoliated Graphite/EpoxyComposites. Carbon 2008, 46, 806–817.(17) Yu, A.; Ramesh, P.; Itkis, M. E.; Bekyarova, E.; Haddon, R. C.

Graphite Nanoplatelet-Epoxy Composite Thermal Interface Materials.J. Phys. Chem. C 2007, 111, 7565–7569.(18) Wang, S.; Tambraparni, M.; Jingjing, Q.; Tipton, J.; Dean, D.

Thermal Expansion of Graphene Composites. Macromolecules 2009,42, 5251–5255.(19) Kalaitzidou, K.; Fukushima, H.; Drzal, L. T. Multifunctional

Polypropylene Composites Produced by Incorporation of ExfoliatedGraphite Nanoplatelets. Carbon 2007, 45, 1446–1452.(20) Fang, M.; Wang, K.; Lu, H.; Yang, Y.; Nuttb, S. Single-Layer

Graphene Nanosheets with Controlled Grafting of Polymer Chains.J. Mater. Chem. 2010, 20, 1982–1992.(21) Kim, H.; Abdala, A. A.; Macosko, C. W. Graphene/Polymer

Nanocomposites. Macromolecules 2010, 43, 6515–6530.(22) Rafiee, M. A.; Rafiee, J.; Srivastava, I.; Wang, Z.; Song, H.; Yu,

Z.-Z.; Koratkar, N. Fracture and Fatigue in Graphene Nanocomposites.Small 2010, 6, 179–183.(23) Nan, C. W.; Liu, G.; Lin, Y.; Li, M. Interface Effect on Thermal

Conductivity of Carbon Nanotube Composites. Appl. Phys. Lett. 2004,85, 3549–3551.(24) Zhong, H.; Lukes, J. R. Interfacial Thermal Resistance between

Carbon Nanotubes: Molecular Dynamics Simulations and AnalyticalThermal Modeling. Phys. Rev. B: Condens. Matter 2006, 74, 125403/1–125403/10.

(25) Pollack, G. L. Kapitza Resistance. Rev. Mod. Phys. 1969,41, 48–81.

(26) Huxtable, S.; Cahill, D.; Shenogin, S.; Xue, L.; Ozisik, R.;Barone, P.; Usrey, M.; Strano, M.; Siddons, G.; Shim, M.; Keblinski,P. Interfacial Heat Flow in Carbon Nanotube Suspensions. Nat. Mater.2003, 2, 731–734.

(27) Shenogin, S.; Liping, X.; Ozisik, R.; Keblinski, P. Role of ThermalBoundary Resistance on the Heat Flow in Carbon-Nanotube Composites.J. Appl. Phys. 2004, 95, 8136–8144.

(28) Dervishi, E.; Li, Z. R.; Watanabe, F.; Biswas, A.; Xu, Y.; Biris,A. R.; Saini, V.; Biris, A. S. Large-Scale Graphene Production by RF-cCVD Method. Chem. Commun. 2009, 27, 4061–4063.

(29) Raghu, A. V.; Lee, Y. R.; Jeong, H. M.; Shin, C. M. Preparationand Physical Properties of Waterborne Polyurethane/FunctionalizedGraphene Sheet Nanocomposites. Macromol. Chem. Phys. 2008, 209,2487–2493.

(30) Ventol�a, L.; Ramírez, M.; Calvet, T.; Solans, X.; Cuevas-Diarte,M. A. Polymorphism of N-Alkanols: 1-Heptadecanol, 1-Octadecanol,1-Nonadecanol, and 1-Eicosanol. Chem. Mater. 2002, 14, 508–517.