Study On Microvia Filling Performances And Action Mechanismsof EPE Inhibitors In Copper Electroplating Process

Microvia metallization is not only an effective method for achieving electrical interconnection of the adjacent layers in print circuit board (PCB), but also a key technology for the development of high density interconnection (HDI) boards. To ensure the reliability of electrical interconnection, the microvias should be fully filled by deposited copper. During the process of microvia filling, the performances of the plating solution can be evaluated by the sizes of some parameters, such as the the plating time, thickness of the deposited copper film on the board, and the microvia filling performance. Up to now, the copper plating solution for microvia filling in the home market is dominated by the foreign brands, which go against the development of PCB industry of our country. So, it is significant to develop the copper electroplating solution with our own intellectual property and high efficieny for microvia filling.Generally speaking, it is impossible to fully fill microvia by deposited copper in the plating solution without additives, because the current density in the bottom of the microvia is relatively lower. When chloride ions, accelerators, suppressors as well as levelers are present in the plating solution, the differences of current density between in the bottom of the micrvia and at the surface of the board can be changed by the synergistic effect among the additives, which finally lead to the microvia being perfectly filled. This phenomenon can also be called “superfilling”.In this work, EPE2900was finally chosen to be the optimal suppressor after comparing with other ethylene oxide (EO)-propylene oxide (PO)-ethylene oxide (EO) triblock copolymers by studying the relationship between the suppression strength of the suppressors and the corresponding filling performances. EPE2900, one of EO-PO-EO (EPE) triblock copolymers, contains40%EO and has a molecular weight of2900. The results of micovia filling experiments indicated that EPE2900, as a suppressor in copper plating solution, showed much better micovia filling performance than PEG6000. So, EPE2900was determined to be the suppressor used in this work. According to the microvia filling performances, the copper plating formula using EPE2900as the suppressor was optimized and the optimized formula was:220g/L CuSO4·5H2O,54g/L H2SO4,60mg/L Cl-,6mg/L SPS,200mg/L EPE2900and4mg/L JGB. Using this optimized formula, the microvia filling experiment was carried out at a current density of2A/dm2with the plating time of60min. The temperature of the copper plating solution was25℃. After the filling experiment finished, the thickness of the deposited copper on the board was about16μm and the filling performance was up to95%.In the optimized copper plating formula, cyclic voltammetry and galvanostatic measurements were employed to systematically study the function of each additive as well as their interactions. The results indicated that there was an obvious synergistic effect between EPE2900and Cl-. In other words, the suppression strength of EPE2900on inhibiting copper deposition can be greatly enhanced in the presence of Cl-. In addition, given a fixed EPE2900concentration, there existed an optimal Cl-concentration to achieve the maximum synergistic effect. EPE2900showed quick adsorption on the cathode, whereas SPS showed a slow adsorption process. When they were simultaneously present in the plating solution, competitive adsorption would occur between EPE2900and SPS. Strong forced convection could facilitate the adsorption of EPE2900on the cathode surface. After the introduction of JGB into the plating solution, the suppression strength of EPE2900was enhanced. The results of cyclic voltammetry indicated that compared to other additives, the synergistic effect between EPE2900and Cl-was the most obvious, which had a decisive effect on the potential of copper deposition.The steric configuration of EPE2900molecule in water solution and the electronegativity of oxygen atoms in EPE2900molecule were simulated and calculated by Material Studio (MS) and Gaussian softwares, respectively. Based on the filling plating results obtained at different conditions and the interactions between EPE2900and other additives, a simple adsorption model of EPE2900on the cathode was proposed. Then, some electrochemical measurements, such as linear sweep voltammetry, galvanostatic measurement and electrochemical impedence spectroscopy were employed to demonstrate the proposed adsorption model.To get more insight into the adsorption mechanism of EPE2900, cyclic voltammetric stripping (CVS) method was employed to systematically study the effect of EPE2900and Cl-concentrations on the integral values of copper stripping peaks. It indicated that when the concentration of EPE2900in the plating solution was fixed, Q values would decrease with Cl-concentration increasing in a certain range. This result demonstrated that the increase of Cl-concentration can improve the suppression strength of EPE2900. Importantly, it was noted that when the Cl-concentration was increased from0to10mg/L, the Q values decreased sharply. However, this situation was greatly changed when Cl-concentration was larger than10mg/L. Because at this point the Q values decreased slowly as the increase of Cl-concentration. According to this phenomenon, the abscissa of the knee point was calculated by using linear fitting method, which corresponded to the lowest Cl-concentration needed for EPE2900to achieve the critical adsorption on the cathode. With the similar method, the lowest EPE2900concentration needed for critical adsorption can also be calculated in the plating solution containing fixed Cl-concentration. The results of several parallel experiments all demonstrated that, to ensure EPE2900achieving critical adsorption on the cathode, there at least11.46mg/L EPE2900and4.85mg/L Cl-were present in the plating solution. According to the trend of Q values varied with EPE2900and Cl-concentrations, the adsorption model of EPE2900at its critical adsorption status was proposed. At the same time, the transformation of the adsorption model caused by the increase of EPE2900and Cl-concentrations was described in detailed. Specially, the testing techniques using rotating ring-disk electrode (RRDE) was introduced into this text to demonstrate the interaction between EPE2900and Cu+, which proved the reasonability of the adsorption model of EPE2900.