Evolution of surface roughness during copper electrodeposition in the presence of additives: Comparison of experiments and Monte Carlo simulations

The effect of solution additives on surface roughness evolution during copper electrodeposition on flat copper surfaces was investigated by comparing experimental measurements with numerically simulated results. Deposition was carried out on Cu(111) surfaces for 300 s at 10-20 mA/cm2 with solutions of 0.5 M CuSO4 + 1 M H2SO4 with various combinations of 1.5 mM Cl-, 500 ppm polyethene glycol MW 3400, and either 2 mM or 2 *mM 3-mercapto-1-propane sulfonic acid. Potential transients and surface roughness data were measured and compared with multiscale numerical simulations carried out with a noncontinuum Monte Carlo model of the surface linked to a continuum model of the diffusion layer. Hypotheses for electrodeposition of copper in additive-free solutions as well as those containing one, two, or three additive components were selected from the literature and encoded into a Monte Carlo model. The simulations included up to 32 parameters of which some were available from independent measurements and the remaining were selected to match experimental data. Simulations were carried out on the Condor high-throughput system. Comparison of numerical results with experimental data led to recognition of several areas for improvement. These included: recognizing the need for high-throughput experimental methods for obtaining a sufficient quantity of data to establish statistically significant trends; improving numerical procedures to handle pseudoparticle sizes significantly smaller than the 100 nm used here; linking multiscale codes with external scripts to facilitate the use of different modules at one scale without affecting the operation of the module used at another scale; and automating the process for obtaining values for the most sensitive parameters.