High thermal conductivity of high-quality monolayer boron nitride and its thermal expansion

Atomically thin boron nitride is one of the best thermal conductors among semiconductors and insulators.


This PDF file includes:
Section S1. Optical and AFM images of atomically thin BN samples Section S2. Raman spectra of the suspended 1-3L and bulk BN Section S3. Temperature coefficients of the 1-3L BN suspended over Au/Si substrate Section S4. Absorbance of 1-3L BN measured on quartz Section S5. Laser beam radius Section S6. Error calculation Section S7. Thermal conductivity of graphene as a control Section S8. Thermal equilibration on MD simulations using LAMMPS Section S9. TEC of SiO 2 /Si substrate simulated by FEM Section S10. Comparison of the TEC of common 2D materials    Table S1. TEC of 2D materials (10 -6 K -1 ). Section S1. Optical and AFM images of atomically thin BN samples Figure S1 shows additional optical and atomic force microscopy (AFM) images of the mechanically exfoliated 1-3L BN suspended over pre-fabricated microwells.   S4d and e). An optical power meter (1916-C, Newport) was then used to measure the absorbance of the BN nanosheets under 514.5 nm laser. We used UV-Vis reflectance spectrum to estimate the reflectance of the bare quartz and quartz covered by few-layer BN produced by CVD, and quartz showed a nuance (i.e. slightly higher) in reflectance with or without the coverage by few-layer BN at 514.5 nm.

Section S5. Laser beam radius
The laser beam radius 0 (objective lens 100×) was measured by performing a micro-Raman scan across the boundary of a partially Au-coated Si wafer. fig. S6a and b display the optical image of the boundary with and without Au coating and the corresponding Raman mapping of the Si frequency at 520.5 cm -1 across the boundary. The measured intensity (I) was proportional to the total laser power incident on the sample. The step size used in the mapping was 0.1 µm. fig. S6c shows the distribution of I as a function of the distance ( ) to the boundary. A Gaussian function (− 2 / 0 2 ) was used to fit the slope / to calculate 0 , which was found to be 0.31±0.01 μm ( fig. S6d).

Section S7. Thermal conductivity of graphene as a control
For validation purpose, we used the same optothermal procedure to measure the thermal conductivity of monolayer graphene suspended over the same substrate with 3.8 µm holes connected by 0.2 µm wide trenches. The graphene sheets were mechanically exfoliated from highly oriented pyrolytic graphite (HOPG) using Scotch tape. The Raman results are shown in fig. S7. The calculated thermal conductivity of 1L graphene was 2102±221 W/mK.

Section S8. Thermal equilibration on MD simulations using LAMMPS
An initial equilibration using NVT ensemble on the systems was followed by the calculation of the thermal conductivity at the NVE ensemble as discussed in the main text. shows the dimension of the model projected on the x-y plane.

Section S10. Comparison of the TEC of common 2D materials
The thermal expansion coefficients of some common 2D materials at 300 K are compared in Table S1, with the values of BN from this study.