RETRACTED: Copper Electrodeposition in Through-Silicon Via under Galvanostatic Condition

Abstract

We present a tertiary current-distribution model for copper electrodeposition under galvanostatic conditions, with detailed surface chemistry kinetics for the model system of copper electrodeposition. The values of the kinetic parameters are extracted from results of statistical chronoamperometry and linear sweep voltammetry experiments using a potential-dependent, diffusion-adsorptiondesorption model of the experimental system. The resulting surface chemistry description is combined with fundamental conservation laws, including mass transport, potential distribution, and competitive adsorption, to form the tertiary current distribution model. Two-dimensional finite element simulations of this model provide information about the causes of the thickness of the copper electrodeposition layer, including potential variations of the cathodic surface, fluctuations in the cupric ion concentration and coverage of the accelerator. The model is used to predict the filling process with different current densities and SPS concentrations and is verified by copper electrodeposition experiments. The "optimal zone" was determined for the SPS concentration and the inward current density to achieve good TSV filling. Sample points in the outer zone were selected and the failure mechanism was studied to help to infer the right direction toward the "optimal zone" in copper electrodeposition. The models are potentially useful tools for measuring the properties of electrolyte solutions and for the optimization of copper electrodeposition processes. Through-Silicon via (TSV) is one of key techniques for 3D integration with many advantages. The electroplated copper is a major interconnect material in TSV due to its high electrical conductivity, lower cost than alternative deposition methods.1,2 Copper electrodeposition as an efficient method of metallization is widely accepted, and it has been proposed by many researchers. [3][4][5] According to the feature of TSV on chip, some special additives such as accelerator (Bis-(3-sulfopropyl) disulfide, or SPS), suppressor (Polyethylene glycol, or PEG), leveler (Janus Green B) and Cl − , are usually applied in electrolyte solution to assist copper filling with void-free.6-8 An appropriate distribution of additives provides a fast growth at the bottom and the growth rate at via opening is suppressed, as improper distribution of additives result in unfilled feature.9 Hence, behaviors of additives on the cathode surface are very important for TSV filling process. However, there are still some challenges for establishing the optimal process for copper filling because the influence of feature of TSV, complicated mechanisms of additives behaviors and complicated fabricated process.2 A predictive numerical model of the copper electrodeposition process is a very helpful tool to investigate the optimal process with some advantages, such as low cost and efficiency. A typical numerical model should have the ability to predict feature evolution of copper electrodeposition under different operating conditi...