Effect of Substitutional Silicon Dopants on Thermal Properties of Graphene

Abstract

One-atom-thick layer graphene has superlative thermal conductivity k = 2000 ~ 5000 W/mK but it is significantly suppressed by the presence of lattice defects. Among the different types of lattice defects, substitutional defects efficiently reduce the in-plane thermal conductivity but maintain the unique hexagonal structures of graphene, which possibly enables to preserve other novel properties like high electrical mobility and optical transparency. According to the second-order perturbation theory, the phonon scattering rate by the substitutional dopants is known to be proportional to the square of the mass different with the host C atoms (∆M). Previous studies have only demonstrated the thermal conductivity reduction of graphene by the substitutional dopants of small mass difference, including N, 13C, B, and 24C. <br/> In addition, they mostly facilitated the case of small mass perturbation and thermally and atomically isolated suspended graphene. The heavy dopants of larger mass difference will possibly provide more interesting and significant thermal properties alteration in graphene, including by over one order-of-magnitude further reduction in k.<br/> In this point, the selection of our experimental investigation on the thermal properties alteration by substitutional dopants is the silicon atoms, which provides the largest mass difference compared to previous works, in our knowledge. We have successfully synthesized Si-doped graphene (SiG) by using liquid precursor and facilitating the low pressure chemical vapor deposition (LPCVD) synthesis, which suppresses the lattice defects in synthesized graphene samples and thus allows the exclusive understandings on the effects of Si dopants. In addition, the Si doping concentrations indicated by I(D)/I(G) Raman peak ratio are controlled by manipulating the synthesis temperature and pressure. Correlation between the I(D)/I(G) and Si doping concentration enables to measure the in-plane thermal conductivities with the Si doping concentration variation. <br/> The in-plane thermal conductivity (k_SiG) of the supported SiG was probed using optothermal Raman thermometry after being transferred on an ultra-thin 8 nm SiO2 substrate. The measurement results show that k_SiG gradually decreases with the increase of Si dopant mass concentration from 1.41 % to 3.13 %, exhibiting more than one order-of-magnitude reduction from the supported pristine graphene of k = 645.6 ~ 900.2 W/m·K down to k_SiG = 46.7 ~ 79.5 W/m·K for the present supported SiG. When compared to the suspended pristine graphene (PG) case (k = 2662.5 W/m·K), the reduction is accounted for being nearly two order-of-magnitudes. We attribute the observed substantial reduction of k_SiG to the exceptionally large phonon scattering strength of Si dopants, which can be explained by the large mass difference between the doped Si atoms and the C atoms. We have also found that enhancing graphene-substrate conformity through thermal annealing in vacuum further lowers k_SiG from that of the ambient annealing case, due to the increasing phonon scattering rate by underlying substrate. Further aspects in our findings include reduced temperature dependence of k_SiG and negligibly small non-equilibrium effect in Si-doped graphene, both of which are believed to attribute to the enhanced phonon scattering (reduced phonon mean free path) by Si dopants as well as by the presence of the graphene-substrate interface. <br/> Finally, we found that substitutional silicon dopants do not violently destruct the electrical properties of graphene when compared to previously measured thermal conductivities. The measured thermal conductivities show the maximum 8 times further reduction than that of electrical conductivities at about 2.6 % Si doping concentration. This results can be attributable to the resemblance of silicon and carbon atoms in terms of their number of valence electrons. We note that substitutional doping of graphene with heavy Si atoms can be a novel strategy to independently control the thermal properties and thus possibly enhance the thermoelectric efficiency of graphene.<br/>