Thickness-dependent in-plane thermal conductivity of suspended MoS₂ grown by chemical vapor deposition

Abstract

The in-plane thermal conductivities of suspended monolayer, bilayer, and multilayer MoS films were measured in vacuum by using non-invasive Raman spectroscopy. The samples were prepared by chemical vapor deposition (CVD) and transferred onto preformed cavities on a Au-coated SiO/Si substrate. The measured thermal conductivity (13.3 ± 1.4 W m K) of the suspended monolayer MoS was below the previously reported value of 34.5 ± 4 W m K. We demonstrate that this discrepancy arises from the experimental conditions that differ from vacuum conditions and small absorbance. The measured in-plane thermal conductivity of the suspended MoS films increased in proportion to the number of layers, reaching 43.4 ± 9.1 W m K for the multilayer MoS, which explicitly follows the Fuchs-Sondheimer suppression function. The increase in the thermal conductivity with the number of MoS layers is explained by the reduced phonon boundary scattering. We also observe that the Fuchs-Sondheimer model works for the thickness-dependent thermal conductivity of MoS down to 10 nm in thickness at room temperature, yielding a phonon mean free path of 17 nm for bulk.

Transient thermal characterization of suspended monolayer MoS2

“…We thus observe a considerable spread between devices. Moreover, most of the values of k found here are smaller compared to previous observations in literature that used exfoliated MoS 2 devices [17,18], but are larger than CVD MoS 2 values [19].…”

“…By exploiting the temperature-dependent phonon frequency shifts in Raman spectroscopy [16], several experimental works have measured the thermal conductivity of singlelayer MoS 2 . Experimental values of k = 34.5 and 84 W/(m·K) of exfoliated single layer MoS 2 have been reported [17,18], while single-layer MoS 2 grown by chemical vapor deposition was found to show a significantly lower thermal conductivity of 13.3 W/(m·K) [19].…”

Thermal transport in MoS2 from molecular dynamics using different empirical potentials

“…This part was not included in the SW potentials. From here on, we only use the REBO-LJ potential and the efficient HNEMD method, focusing on comparisons with experiments [8][9][10][11][12][13][14][15][16][17][18] and results from BTE approach combined with DFT calculations [19,20].…”

“…Variations in the experimental results could be due to differences in the quality of each sample, measurement calibration, and the presence of thermal contact resistance. Indeed, some experimental samples[12,16,17] were grown by chemical vapor deposition (CVD), which are usually polycrystalline, consisting of grains separated by grain boundaries. It has been recently shown that dense grain boundaries can heavily reduce[69] the thermal conductivity of MoS 2 .…”

Spatiotemporal Mapping of a Photocurrent Vortex in Monolayer MoS2 Using Diamond Quantum Sensors

“…The high-quality MOCVD-grown monolayer 15 readily covers our entire 2x2x0.5 mm diamond sample, but we deliberately expose a portion of the diamond to facilitate control measurements (see Supplementary Section 2). The slight absorption 27 Additionally, we band-pass filter the detected PL between 690 nm and 830 nm to predominantly isolate NV center emission, as shown in the room-temperature PL spectra of monolayer MoS 2 and a typical ensemble NV sample ( Fig. 1f).…”