Investigation Of Electrochemical Mechanism About The Additives In Copper Interconnect

Nowadays copper interconnect technology is adopted widely in the semiconductor industry because of the excellent performance of copper. The electro-deposition of copper is economical, selective, suitable for mass production and only deposited in necessary parts. The good covering power ability and anti-metal performance make it the main approach in copper interconnect filling. However, the perfect filling in hole is not easy to be achieved for silicon vias with high aspect ratio. Adding additives to make the current inverts to gain ―bottom-up‖ effect is a popular method today.In the process of super filling, additives are playing really important role. In the first place, we invested the properties and behaviors of the 4 commercial additives(GW,GN1,GN2,GN3) by electro-chemical measurements, such as linear sweep voltammetry(LSV), cyclic voltammetry(CV) and electrochemical impedance spectroscopy(EIS). Compared the GW additives and the GN series additives from the electro-chemical properties. Moreover, we choose one group additives with good property from GN series additives and the GW group additives, using them do deposition experiment in Damascence chip. Compared the electro-deposition effect between the two group additives. We can get that the properties of GW-ASL additives are better than the GN series additives both in electrochemical aspect and the deposition effect aspect.In the second place, we researched the mechanism of the additives used in copper interconnect based on SPS-PEG additive group, got an evaluation method according to the value of the peak-valley performance. The scan rate, initial scan potential and low scan potential all have influence on the peak-valley performance. According the evaluation method: the bigger of ΔE, the deeper of the via can be deposited; the bigger of ?I, the better of the super-filling; the bigger of Ip, the faster of the filling rate; the lower of Iv, the better of the inhibition effect, we evaluate the additives of GW-ASL. It can explain the competitive absorption relationship between GW-A and GW-L. In addition, the impurity peak which appeared in the CV curve is studied in this paper. The impurity peak appeared only when the accelerator and suppressor are both added into the plating solution. The electric quantity needed when dissolving the impurity increasing with the electric quantity of dissolving increasing. IR and XPS testing are measured in the situations of before and after the impurity peak appeared. The amount of the remain elements are has a little different between the two situations. In the same time, the existence of the impurity affect the type of the remain copper.