Effect of Growth Parameters on MoS2 Film Quality Deposited by Low-Temperature MOCVD Using I-Pr2DADmo(CO)3 and (t-C4H9)2S2

Abstract

ABSTRACT Molybdenum disulfide (MoS2) is one of the representative two-dimensional layered material that is being actively researched for application in various fields owing to its characteristics such as a moderately wide bandgap (~1.8 eV) and excellent stability. Particularly promising applications include thin film transistors (TFTs) for displays and three-dimensional integrated circuits (3D-ICs). However, applications in these fields require direct deposition of high-quality films at a low temperature in order to meet the conditions such as the softening temperature of the glass for displays (<500-600°C) and the process temperature in the back-end of line of semiconductor device fabrication (<500°C). Low-temperature fabrication of MoS2 thin film using metalorganic chemical vapor deposition (MOCVD) has already been demonstrated in several previous studies [1-3]. However, they include processes such as long deposition times of about 10 hours or more [1,3], the use of highly toxic hydrogen sulfide (H2S) [2,3], and the addition of NaCl as a growth promoter [1,3]. These processes may lead to decrease in productivity and safety problems which would result in cost increase, and decrease in device reliability due to sodium contamination. These problems are considered to be bottlenecks for industrial and practical use. Our previous report examined process strategies that could overcome these challenges [4]. As a molybdenum precursor, we adopted i-Pr2DADMo(CO)3, which shows high adsorption and is expected to improve the throughput and suppress parasitic gas phase reactions. As a sulfur precursor, we used (t-C4H9)2S2, which is non-toxic as opposed to H2S. For the samples in our study, we demonstrated low-temperature (≤500°C) growth of MoS2 thin film without adding NaCl. However, the grain size was very small, about 10 nm, suggesting that it is necessary to improve the film quality for device application. Therefore, the purpose of this study is to investigate the effects of various deposition parameters on film growth and to provide guidelines for process optimization for improvement of film quality.In this study, the growth of MoS2 thin films by MOCVD was carried out in a cold wall type reactor, and the above-mentioned organic compounds were used as Mo and S precursors, and sapphire (0001) was used as the substrate. The samples were evaluated for film morphology by scanning electron microscope (SEM), Atomic Forth Microscopy (AFM) and film composition by X-ray photoelectron spectroscopy (XPS).Figure 1 (a) shows the surface SEM images of each sample with the S precursor supply rate set from 0.2 to 1.5 sccm while keeping the Mo precursor supply rate constant at 4.9 x 10-3 sccm. In addition, Fig. 1 (b) shows the average grain size and nucleation density calculated from 6 points in the sample with respect to the S precursor supply rate. As the S precursor supply rate increased from 0.2 to 0.7 sccm, decrease in nucleation density and increase in grain size up to approximately 110 nm were confirmed. It has been reported that the critical radius of clusters is larger for the clusters with stoichiometric composition than those with higher metal ratio with lacking chalcogen [5]. Therefore, it is considered that nucleation was suppressed by the increase in the critical radius of the cluster induced by the increase in the supply rate of S precursor. On the other hand, when the S precursor supply rate was further increased to 1.5 sccm, the nucleation density increased and the grain size decreased. This suggests that the S precursor supply rate have an optimum value within the range investigated in this study. We also evaluated the effect of temperature, pressure, and so on.This work was partly supported by JST CREST Number JPMJCR16F4, Japan. This work was also partly supported by JSPS KAKENHI Grant Number 18J22879. REFERENCE Kang, S. Xie, L. Huang, Y. Han, P. Y. Huang, K. F. Mak, C.-J. Kim, D. Muller, and J. Park, Nature 520, 656 (2015). Mun, Y. Kim, I.-S. Kang, S. K. Lim, S. J. Lee, J. W. Kim, H. M. Park, T. Kim, and S.-W. Kang, Scientific Reports 6, 21854 (2016). Mun, H. Park, J. Park, D.H. Joung, S.-K. Lee, J. Leem, J.-M. Myoung, J. Park, S.-H. Jeong, W. Chegal, S.W. Nam, and S.-W. Kang, ACS Appl. Electron. Mater. 1 (4), 608 (2019). Yamazaki, Y. Hibino, Y. Oyanagi, Y. Hashimoto, N. Sawamoto, H. Machida, M. Ishikawa, H. Sudo, H. Wakabayashi and A. Ogura, MRS Advances 5, 1643 (2020). Yue, Y. Nie, L. A Walsh, R. Addou, C. Liang, N. Lu, A. T Barton, H. Zhu, Z. Che, D. Barrera, 2D Mater. 4, 045019 (2017). Figure 1