Effect of Dimethylglyoxime on Cobalt Bottom-up Filling

Abstract

Copper (Cu) damascene processes have been used to produce back end of line (BEOL) interconnect structures1. As the critical dimension of BEOL structures approaches the electron mean free path of Cu or below, the copper resistivity exponentially increases, posing significant challenges on scaling. Metals with shorter electron mean free path, for example cobalt (Co), have been explored as the alternative material to replace Cu in the finest metal levels2.Cu electrodeposition for trench filling has been extensively investigated. The use of multi-component additive packages leads to void-free filling and the deposition process is explained with the synergistic interaction between the so called suppressor and accelerator3, 4. However, since the standard potential of Co2+/Co is more negative than that of H+/H2, hydrogen evolution reaction (HER) is inevitable during Co deposition, which may pose additional challenges on Co void-free filling. Defect-free filling of Co has been reported in fine features using a single derivative of mercapto-benzimidazole, which suppresses Co deposition but breaks down upon the metal deposition and additive incorporation5. Co filling in extremely fine structures has also been reported6, 7 and HER has been shown to play an important role in such processes8, 9, creating a contrast in current efficiency, and thus in deposition rates, between the feature bottom and field regions. However, proprietary chemistries were used in such studies and no chemical information is available. In our previous study, additives with a conjugated pair of oxime groups such as dimethylglyoxime (DMG) not only strongly suppress Co deposition, and the suppression breaks down upon the reduction and incorporation of adsorbed Co-dioxime chelates10, but also catalyze HER11 and have the potential in tailoring the Faraday efficiency. In this talk, the effects of electrolyte pH, concentration of DMG, current density and agitation will be systematically discussed in a context of current efficiency. A mechanism is proposed to explain the Co bottom-up filling using DMG.Figure 1 (a) shows the cyclic voltammograms of Co deposition in presence of different DMG concentrations. It is clear to see that the suppression effect becomes stronger as more DMG is added into the electrolyte. Moreover, the suppression breaks down at a negative potential and a hysteresis is resulted, potentially enabling different deposition rates between field and feature. Figure 1 (b) shows the effect of agitation on Co deposition, where such hysteresis loops gradually shift toward more negative potentials as the rotation rate increases, in a similar way as the DMG concentration increases. Figure 1 (c) shows the current efficiency of Co thin film deposition at different DMG concentration and different rotation rates. For example, 200 ppm DMG at 215 rpm was used to emulate Co deposition in the field region and 25 ppm DMG at 29 rpm to mimic the situation at the bottom of a feature with an aspect ratio of about 1:3. The significant difference in Co deposition rates at a low current density of 4 mA/cm2 leads to a successful void-free Co filling in the trench shown in Figure 1(d). References P. C. Andricacos, C. Uzoh, J. O. Dukovic, J. Horkans, and H. Deligianni, IBM Journal of Research and Development 42, 567 (1998).D. Gall, Journal of Applied Physics 119, 085101 (2016).T. Moffat, D. Wheeler, W. Huber, and D. Josell, Electrochemical and Solid-State Letters 4, C26 (2001).T. P. Moffat, J. Bonevich, W. Huber, A. Stanishevsky, D. Kelly, G. Stafford, and D. Josell, Journal of The Electrochemical Society 147, 4524 (2000).C. H. Lee, J. E. Bonevich, J. E. Davies, and T. P. Moffat, Journal of The Electrochemical Society 156, D301 (2009).F. Wafula, J. Wu, S. Branagan, H. Suzuki, A. Gracias, and J. van Eisden, in Electrolytic Cobalt Fill of Sub-5 nm Node Interconnect Features, 2018 (IEEE), p. 123.J. Wu, F. Wafula, S. Branagan, H. Suzuki, and J. van Eisden, Journal of The Electrochemical Society 166, D3136 (2019).M. A. Rigsby, L. J. Brogan, N. V. Doubina, Y. Liu, E. C. Opocensky, T. A. Spurlin, J. Zhou, and J. D. Reid, ECS Transactions 80, 767 (2017).M. A. Rigsby, L. J. Brogan, N. V. Doubina, Y. Liu, E. C. Opocensky, T. A. Spurlin, J. Zhou, and J. D. Reid, Journal of The Electrochemical Society 166, D3167 (2019).T. Lyons and Q. Huang, Electrochimica Acta 245, 309 (2017).Y. Hu and Q. Huang, Journal of The Electrochemical Society 166, D3175 (2019). Figure 1