Vapor-based synthesis of ultrathin metallic films on weakly interacting substrates (2D materials, dielectric surfaces) is a key step for fabricating heterostructure devices in the areas of optoelectronics, energy saving and energy conversion.
Prior work using real-time growth monitoring of polycrystalline Cu films on SiO2 substrate) has revealed that the thickness at which the film becomes continuous (i.e. full coverage of the substrate) is affected by growth interrupts, but in opposite way depending whether the interruption of the deposition process occurs before or after the percolation threshold (electrical conductivity)[1].
To better understand these experimental findings, we have performed atomistic simulations using Kinetic Monte Carlo (KMC). A dedicated KMC model based on 3D rigid lattice has been developed to study the growth of sputter-deposited Cu thin films. The model takes into account the angular and energy distribution of the incoming particles, as provided by SRIM and SIMTRA calculations. Beside deposition and diffusion events, atomistic mechanisms relevant to growth under energetic deposition conditions, such as re-sputtering, directionally-induced surface diffusion and bulk defect creation, have been explicitly taken into account. The anisotropy of Cu surface diffusion is taken into account as well.
We show here preliminary simulation results on reentrant smooth growth and about the effect of multistep deposition process for a homoepitaxial system (Cu on Cu(001)). To study the reentrant smooth growth, simulations have been done at temperatures varying between 150K and 450K and deposition rates from 0.1ML/s to 100ML/s. Taking into account the funneling mechanism[2] and Ehrlich-Schwoebel barrier[3,4], the simulation results are in good agreement with the previous experimental[5] and simulation ones. In what concerning the effect of multistep deposition process, the temperature was fixed at 350K and the initial deposition time has been chosen in such a way that the relaxation period starts before and after the 2D percolation of the deposited film. Different deposition rates (1ML/s and 10ML/s) and relaxation times (0.5s and 1.0s) have been also considered. The simulation results show that the variation of the thickness of continuity function of the moment when the interruption of the deposition process is done, is in a good qualitative agreement with the experimental results mentioned before. Our model being a rigid lattice model, a better agreement with the experimental data corresponding to Cu deposition on SiO2 (amorphous substrate) is expected by simulating a heteroepitaxial growth.

[1] C. Furgeaud, PhD Thesis, Univ. of Poitiers, 2019
[2] J. W. Evans, D. E. Sanders, P. A. Thiel, A. E. DePristo, Phys. Rev. B, 1990, 41, 5410
[3] H. C. Kang, J. W. Evans, Surf. Sci., 1992, 271, 321
[3] R. L. Schwoebel, E. J. Shipsey, J. Appl. Phys., 1966, 37, 3682
[4] G. Ehrlich, F. G. Hudda, J. Chem. Phys. 1966, 44, 1039
[2] C. E. Botez, P. F. Miceli, P. W. Stephens, Phys. Rev. B, 2001, 64,125427
[3] G. Costantini , F. Buatier de Mongeot, C. Boragno, U. Valbusa, Surf. Sci., 2000,459, L487–L492
[4] M. Hancock, C. Fein, R. G. Tobin, J Chem Phys, 2010, 133, 164701
[5] J. Seo, H.-Y. Kim, J.-S. Kim, J. Phys. Cond. Matt., 2006
Acknowledgments: This work has been done in the framework of the INTEGRAL project (INTerface reactivity, microstructure and stress Evolution during thin film GRowth: multi-scALe modelling and experimental validation (INTEGRAL)” - Reference ANR: # ANR-19-CE08-0024-01