You are on page 1of 14

Elsevier Editorial System(tm) for Chemical Physics Letters

Manuscript Draft
Manuscript Number: CPLETT-15-396
Title: Length and Width Dependent Thermal Conductivity of Graphene Nanoribbons
Article Type: Regular Article
Section/Category: Nanostructures and Materials
Corresponding Author: Dr. Sanjeev Gupta, PhD
Corresponding Author's Institution: St. Xavier's College
First Author: Yogesh Sonavane, Ph.D.
Order of Authors: Yogesh Sonavane, Ph.D.; Sanjeev Gupta, PhD; Pooja Raval, M.Sc.; Igor Lukačević,
Ph.D.; P. B Thakor, Ph.D.
Abstract: This letter reports on the investigation of thermal conductivity along a long free-standing
graphene nanoribbon (GNR) by a one-dimensional direction-dependent phonon-boundary scattering
into the linearized phonon Boltzmann transport method. It is observed from the theoretical
investigation that GNR has highly length and width dependent thermal conductivities due to difference
in the in/out-plane phonon mode in the phonon dispersion curve (PDC). The calculations show that
thermal conduction of graphene nanoribbon increases with increasing edge roughness of graphene
nanoribbons. Our findings are helpful for understanding and engineering of the thermal conductivity of
graphene nanoribbons in potential electronic devices.
Suggested Reviewers: R Khenata PhD
Senior Scientist, Physics, LPQ3M-Laboratory Faculty of Sciences & Technology Mascara University
khenata_rabah@yahoo.fr
He is expert in this area and working in the same field.
Sandeep Nigam PhD
Senior Scientist, Chemistry , BARC, Trombay, Mumbai
snigam.jpr@gmail.com
expert
Satyaprakash Sahoo PhD
Senior Scientist, Physics, University of Puerto Rico San Juan, PR USA
satya504@gmail.com
expert in experimental techniques

doc India. Yours sincerely (Sanjeev Gupta) . March 24th. we present the results of the study on the thermal conductivity properties of the Graphene Nanoribbons.*Cover Letter Click here to download Cover Letter: cover_letter. In our opinion paper is suitable for the any good journals like Chemical Physics Letters in the field of Applied Physics. not under consideration for publication elsewhere. In this letter. Thanking You. The submission is original. Kindly find our paper entitled “Length and Width Dependent Thermal Conductivity of Graphene Nanoribbons” for the consideration for the publication in Chemical Physics Letters. 2015 Dear Editor.

3 4 Department of Physics. India This letter reports on the investigation of thermal conductivity along a long free-standing graphene nanoribbon (GNR) by a one-dimensional direction-dependent phonon-boundary scattering into the linearized phonon Boltzmann transport method. Xavier‟s College. Veer Narmad South Gujarat Univeristy. India. Igor Lukačević3.*. Sanjeev K. V.J. Strossmayer. S.India. Osijek.gupta@sxca. Croatia Department of Physics. 2 Department of Physics.in) 1 . University J. The calculations show that thermal conduction of graphene nanoribbon increases with increasing edge roughness of graphene nanoribbons. National Institute of Technology. Gupta2. Our findings are helpful for understanding and engineering of the thermal conductivity of graphene nanoribbons in potential electronic devices. Keywords: Graphene nanoribbon. Thermal transport * Corresponding author: Dr. Pooja Raval1. Surat. B. Thakor4 1 Department of Applied Physics. and P.*The Manuscript Click here to view linked References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Length and Width Dependent Thermal Conductivity of Graphene Nanoribbons Yogesh Sonvane1. Surat 395 007. Gupta (sanjeev. Ahmedabad. St. 395007. Sanjeev K. 380009. It is observed from the theoretical investigation that GNR has highly length and width dependent thermal conductivities due to difference in the in/out-plane phonon mode in the phonon dispersion curve (PDC). Phonon.edu.

GNRs have emerged as a potential few-atom-thick material in width and length considered as important parameters to tailoring the carbon based nanoelectronics and have emerged as important candidates for thermoelectric applications with high figure-of-merit due to this external arrangement of atoms which affects its thermal transport. Recently.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2D layered honeycomb network based on sp2 bonded carbon structures. Graphene nanoribbons (GNR). They have used three-phonon Umklapp scattering directly considering all phonon relaxation channels allowed by the energy and momentum conservation in graphene 2D Brillouin zone (BZ). and carried out systematic studies of the scaling of the transport gap in GNRs of various dimensions. Furthermore. The unique properties like electronic. They have found evidence of a transport mechanism in disordered GNRs based on hopping through localized states whose size is close to the GNR width. by unzipping nanotubes. thermal. which is well known in literature as „Graphene‟. Nika et al [1-2] performed a detailed study of the lattice thermal conductivity of graphene using the phonon dispersion obtained from the valence-force field (VFF) method. A large growing number of experimental and theoretical investigations [1-27] have been performed to study the thermal properties. Han and Kim have performed experiments [28]. by lithographic routes and chemical vapour deposition. or chemical routes. is the building smallest unit of most carbon family materials. transport and mechanical have been given hope for improvement in future nanodevices. Jiang et al [3] calculated the thermal conductance of graphene in the pure ballistic limit 2 . the precise knowledge of effect of width and edge roughness in GNRs can give rise to unusual physics. have attracted much attention nowadays and they are realized by various chemical. specially thermal transport assisted by phonon vibrations. where charge carriers are confined in two dimensions and free to move along the ribbon axis.

we have also calculated thermal conductivities of phonon longitudinal acoustic (LA) and transverse acoustic (TA) branches. 3 . In addition. Ghosh et al. which translates to the thermal conductivity in excess of ~6600 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 W/mK. [4] have experimentally found that graphene has thermal conductivities in the range of 3000–5000 W/mK depending on the specific sizes.obtaining a high value. Separate analysis of longitudinal and transversal branch contributions allows us to gain better understanding of phonon dispersions and its characteristics as the main activator of heat conduction in GNR. Thermal conductivity of graphene nanoribbons have potential applications in nanoscale thermal management such as on-chip cooling and energy conversion by controlling the phonon heat transport which is also fundamental in several recently proposed novel schemes of “thermal circuits” or information processing [26]. which vary from 1 to 5 µm. Our investigations of the thermal conductivity and of controlling mechanisms by longitudinal acoustic (LA) and transverse acoustic (TA) phonon branches in graphene nanoribbon can be important for the development of energy-efficient nanoelectronics based on graphene. provides an ample opportunity for theories to test their models to understand the quantum phenomena that dictates phonon modulation in GNRs. which carry heat. Numerical and ab initio calculations have shown that there is a strong need for a simple analytical model. Rapid development on the experimental side. The general expression for the thermal conductivity of graphene can be written as [1-2]. which can explain the differences in the phonon transport in graphene nanoribbon at different temperatures due to different edge types and widths.

    U1. The proper treatment of phonon-boundary scattering rate is critical in nanoscale system.  1  q.  U .  is given as [7].y is the y component of the group-velocity vector in branch λ.  q   M2  q  max.  B.   q     q  = LA. 4 .  U .  q.  2 kBT 2  q  (3) M. It controls the strength of the phonon-phonon scattering process for each branch. T and q. especially when the characteristic size of the system is close to or less than the phonon mean free path.  .   q. kB .  q  is given as [1. reduced Planck constant. S.35 nm is the effective layer thickness. are the area of the sample. TA (2) In this work.  q    B. ћ . qdqd       2 S     q     exp    1  k BT    2 = S qmax 2 2  2  0 0 (1) where λ = LA. absolute temperature and wave vector respectively. Here νλ.λ are mass of a graphene unit cell. TA. δ = 0. Cph . Grüneisen parameter and maximum cut-off frequency for each branch λ respectively. γλ and ωmax. Boltzmann constant.6].  q. therefore 1 according to the Matthiessen‟s rule. we do not consider other scattering mechanisms such as defects. θ (ranging from 0 to 2π ) is the angle between the wave vector and y axis. and qmax is the cut-off wave vector and ωλ [1-2. volumetric specific heat of each mode.  are relaxation times of Umklapp phonon-phonon scattering and phonon-boundary scattering respectively.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65   S c  2   2 ph  2   d q    q      q      exp  k B  k BT   k BT   2 q cos 2   q.6]. where  U .  q  and  B.

while the group velocity for LA and TA branches is 15045 m/s and 10640 m/s.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65  ll   /   q   le   /   q   B .0 and 0. respectively [8]. In Figure 1. 5 . [3]. the atoms denoted with (yellow) circles at the ends are fixed to avoid the spurious global rotation of the GNRs in the our simulation [27].    if ll  le otherwise (4) where le is the averaged distance traveled ballistically by a phonon before hitting the end boundary. ll and le are given as follows.  q. respectively. We have also performed simulations regarding width and length conditions quantitatively. Out hypothesis is that thermal conductivity is limited and dependent on the finite length and width of GNRs. The parameters for dispersion originate from ref. The Grüneisen parameter for LA and TA branches is 2. and ll is the averaged distance traveled ballistically by a phonon before hitting the lateral boundary.75. le    ll    L 1 L cos  1 W sin  L  ydy  2 cos  (5) 0 W W  ydy  2 sin (6) 0 Where L and W are length and width of GNRs.

as acoustic phonons with longer wavelengths become available for heat transfer. It is noticeable that the length dependence of thermal conductivity in GNRs is controlled by the phonon-boundary scattering mechanism. we consider the width of GNRs is 50 nm and the length ranges from 100 nm to 1050 nm. while phonon mean free path (MFP) is controlled by the length of GNRs. Results suggest that our calculated thermal conductivity is limited by the finite length of GNRs and corresponds to the value for graphene of macroscopic size.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Figure 1: The schematic diagram for Graphen Nanoribbon (GNR). This is consistent with the phonon mean free path (MFP) in graphene extracted from the experiment (775 nm) [4] being approximately the same as the length (up to ∼1000 nm) of the GNRs simulated in this study. 6 . In the situation. It is observed in Figure 2 that thermal conductivity rapidly increases with length. Figure 2 shows the length dependence of thermal conductivity in GNRs with different edge roughness along with temperature effect.

Temperature dependence deviates from 1/T law. resulting in MFP being dependent on both end and lateral boundary scattering. due to the boundary restrictions on phonon MFP from graphene edges. which is believed to be the phonon MFP in graphene at different temperatures thermal conductivity increases with increasing temperature as the number of scattered phonons increases. In the case of large edge roughness (δ = 2. almost all phonons undergo purely diffuse reflection at lateral boundaries. so the phonons diffuse normally and the thermal conductivity converges to a constant when the length of GNRs is longer than 700 nm.0 nm). An increase in the GNRs length increases the number of phonons that experience lateral boundary scattering.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Figure 2: The length dependence of thermal conductivity of GNRs with different edge roughness along with temperature variations. particularly for small nanoribbons. The finite value of thermal conductivity of GNRs results from the existence of diffuse phonon-boundary scattering. This has correlations in heat transport of polycrystalline materials with similar restrictions on phonon MFP due to the boundaries. 7 .

thus. Thermal conductivity of LA and TA branches increases as the length of GNRs increases. The 8 . Edge roughness in this case is 0. indicating that longitudinal and transversal modes both positively contribute to the length dependence of thermal conductivity. greater distribution of phonon velocities. Higher velocities provide them larger phonon MFPs and larger MFPs of longitudinal modes give larger contribution to thermal conductivity. On the other hand. Figure 3 gives the width and length dependence of thermal conductivity of individual branch in GNRs at different temperatures from 300K to 800K. the thermal conductivity of both branches almost converges with increasing temperature. Longitudinal modes have greater energy dispersion and.35 nm. The difference in length dependence of thermal conductivity between the in-plane branches (LA and TA) stems from the different phonon dispersion.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Figure 3: The width and length dependence of thermal conductivity of LA and TA branches in GNRs.

e. the thermal conductivity is governed by both factors i. 3. We also observed similar trend as discussed by Han and Kim in their recent experiment [28] that at the charge neutrality point. Therefore. according to eqns (4) -(6). as shown in Fig. the width dependence becomes more evident when the edge roughness is larger. The width-dependence of thermal conductivity of GNRs is also determined by the edge roughness along with temperature variation. In the case of purely specular reflection at smooth lateral boundaries (δ =0).relative influence to the thermal conductivity of individual branch is also length-dependent. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 which is presented in Fig. a length-independent transport gap forms 9 . length and width. Figure 4: The width dependence of thermal conductivity of GNRs with different edge roughness along with temperature variations. 4. However. the averaged relaxation time of phonon-boundary scattering τB is only determined by the length of GNRs.

electrons are localized. while in GNRs whose width is much less than phonon MFP in graphene (about 800 nm). Although the simulated GNRs have much smaller width. flexural phonons have a dominant contribution to the thermal conductivity of graphene whose length and width are larger than micron [29]. Length and width of grapheme nanoribbons have significant influence on thermal 10 . [4] and [5]. The suppression of thermal conductivity by the diffusive phonon-boundary scattering in TA branch is more remarkable than that in LA branch. When the edge roughness increases from 0 (purely specular reflection) to 2. From this we can see that due to the small phonon group velocity. To more insight. In particular. The diffuse phonon-boundary scattering suppresses the thermal conductivity of GNRs. It explains the main features of pure 2D phonon transport in graphene nanoribbon. indicating that the suppression of in-plane phonons is significantly stronger than that of out-of-plane phonons. the thermal conductivity of phonon modes decreases. and charge transport exhibits a transition between thermally activated behavior at higher temperatures and variable range hopping at lower temperatures.0 nm (almost purely diffusive reflection for all phonons). which is clearly shown. which is also confirmed in the previous study of graphene phonon dispersion [28]. only 2 nm with respect to that of graphene inexperiment. which distinguishes it from that of graphite. we have investigated a simple model for calculating the thermal conductivity of graphene nanoribbon of different widths and lengths. effect of edge rougness on graphene nanoribbon gives that the contribution of flexural phonons increases as the width decreases.whose size is inversely proportional to the GNR width. the number of phonons that experience lateral boundary scattering is larger. we found that in this 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 gap. which enhances the contribution of flexual phonons to heat conduction in GNRs. the mean free path of in-plane phonons is suppressed by diffuse phonon-boundary scattering at lateral boundaries of GNRs with rough edge. we observed the strong size dependence in agreement with Refs. In conclusion. The mechanism is that when the width is smaller.

J. K. Science 324 (2009) 1530. M. Balandin. E.. S. A. D. 6. 95 (2009) 163103 10. Rev. Zhang. Hermet. and P. Wakabayashi. Balandin. H Li. K. B. Phys. Miao and C. 20. W. 17. A. Y.. W. L. 9. 458 (2009) 877. L. Ghosh. J. C. Ghosh. Nardelli and K. 78 (2008) 195401. Miao and C. D. 3 (2007) 172. Zhang. Teweldebrhan. Phys. 98 (2011) 141919. 77 (2008) 054302. 92 (2008) 151911. Price. Y. 9 (2010) 555. Dai. Balandin. Ohno. App. Phys. Geim and K. Phys. Phys. B. F. Nature (London) 438 (2005) 201. 80 (2009) 033406. Phys. Z. A. D. Wang. N.W. F. Nika. X. L. L. D. The thermal conductivity calculated with our model gives results consistent with the ab initio theory of heat conduction in graphene and in excellent agreement with the experiments. Bao. Vandescuren. Phys. A. L. B. Lee. X. J. J. Lambin. W Zhang.conductivity. Balandin. P. and H. I. Ghosh. P. S. Geim. Lau. Balandin. Kim. M. and P. Y. 79 (2009) 205418. W. Lett. Sigrist. M. Kong. 79 (2009) 155413 3. Askerov and A. Bao. K. 8 (2008) 902. I. Chem. L. Li. H Tang. Nature Mater. Wang. D. Pokatilov.. I. Kim. F. Henrard.and thermoelectric device applications. Pokatilov. Appl. Nika. Subrina. 8. L. Calizo. Lett. L. Calizo.. Lett. D. Nika. B. Li. A. Ghosh. Phys. N. Lett. S. Jiao.S. Tour. P. W. App. and J. Kosynkin. S. Zhang. Pokatilov. Bao. 11 . Wang and Jian-Sheng Wang. 19. Stormer. L. Dai. P. J. Phys. Lett. H. Nature Mater. D. A. Nano Lett. Nature (London) 458 (2009) 872. S. Reference: 1. N. Miao. Nat. Lau. Zhang. Zhang and Xin-Gao Gong. and M. G. Lett. Tworzydło. 319 (2008) 1229. A. I. A. A. width and 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 edge roughness can imbue them with properties which will benefit their nano. Balandin. A. J. S. Jiang. Diankov. A. K. Lau. Nature (London). and K. Sinitskii. Novoselov. D. Guo. Appl. 4. 6 (2007) 183. 14. Meunier. Paul. A. S. 22. 18. D. D. P. Nika. W. B. Yamada. Beenakker. and G. B. The model utilizes parameters which are responsible for the heat conduction LA and TA phonon branches. L. Knezevic. 94 (2009) 203103 2. B. Ghosh. 21. D. E. D. Y. 13. Yamakita. V. S. Phys. E. Rev. B. 99 (2007) 036601. Phys. Phys. M. Chem. Phys. Rev. 88 (2006) 111909. Y. E. Lett. P. L. Tan. Xie. Teweldebrhan. and H. K. B. Wang. 16. S. Teweldebrhan. N. F. V. and A. Higginbotham. Bao. M. Lau. L. E. W. Rev. Guo. Balandin. Dimiev. 7. S. A. Our model advocates that appropriate combinations of GNRs‟ length. X. 8 (2008) 902. Science. Takane. and C. Aksamija and I. and B. 92 (2008) 151911 5. Appl. A. S. G. W. Z. 12. Calizo. Miao. Ghosh. and C. Teweldebrhan. Lau and A. A. R. Pokatilov. W Yang. Phys. 11.. and C. Rycerz. Bao. App. 15 (2013) 14647. W. Lomeda. N.. 23. Rev. Phys. Pokatilov. Calizo. Nika. F. Rev. 15. Nano Lett.

. J. 25. A. K. Lett. Yang. and Natalio Mingo. Appl. Gupta. Broido. Hunenberger. Lindsay. H. 9 (2009) 2730. Soni and P. G. Sci. Wei. Polym.. 28.L. Chen. Ruan and Y. Rep..1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 24.Y. 26.. Adv. Phys. Xie. Jha. Shen. 27.H. P. D. R. Nano Lett. Dong. 4 (2014) 5085. Sci.. Zhang. P. 12 . K. L.X. 173 (2005) 105. S. 3 (2013) 032117. 82 (2010) 115427. 29. X. 91 (2007) 173108. Hu. J. Rev. X. Phys. H. B. Zhou and J. AIP Adv.