This report focuses on the use of the additive MTT and presents a much faster process for plugging and filling 50 µm through-holes of AR 6 in comparison with studied elsewhere TNBT and NTBC additives. Extension of this development to AR 10 TGHs is also discussed. Main points of interest are optimized plating conditions, like pH, potential and current density, and their impact on the quality of the superconformal filling of TGHs. The potential incorporation of the additive itself or its reduction products into the deposit has also been discussed. Best results for filling TGHs with AR 6 and 10 have been achieved at greatly reduced times of about 1h and 4h, respectively. The DC electroplating is carried out in the presence of MTT in acidic (pH 2 to 4) copper methane sulfonate (CuMSA)/Cl¯ formulations. The applied plating conditions enable the ‘butterfly’ mechanism that nucleates hemispherical or X-shaped Cu plugs in the TGH’s centers while keeping the thickness on external surfaces low. Issues with imperfections trapped in the plugs have been addressed by decreasing the current around the time of plug formation then increasing it for the rest of plating duration. In order to fill AR 10 holes much longer times at lower current density are needed. Larger diameter AR 4 TGH’s have been also successfully filled with Cu. The reported approach also constitutes a cost-effective solution as lab scale quantities of MTT are 8 to 14 times less expensive than the competing additives NTBC and TNBT, respectively.
Finally, the reduction behavior of MTT at different pH values and operational times in a typical bath for Cu superconfomal plating but in the absence of Cu2+ ions has been discussed. A comparison has been made also with the reduction of NTBC and TNBT. The reduction is a two-electron process that forms formazan (MTT case) and diformazan (TNBT and NTBC). Cyclic voltammetry, Chronoamperometry, and controlled-potential electrolysis (CPE) and UV-Vis spectroscopy results are presented. The curves shown in CV work indicating two peaks appearing at - 0.2 V and - 0.6 V (vs Ag/AgCl) where reduction likely occurs. The potential of – 0.22 V was chosen to perform CPE because it is close to the potential where Cu superconformal plating takes place. A key observation during CPE is the formation of purple product both on cathode surface and in the electrolyte (in the MTT case) as purple flaky precipitate after 16 h. The additive reduction progress assessment by UV-Vis spectroscopy, shows establishment of a trend suggesting a linear increase of the reduction product accumulation with time. Also, comparison of the chronoamperometry and UV Vis results suggests highest rate and solubility of the reduction product at lowest pH and decreasing solubility and surface accumulation at higher pH. These observations imply higher rate of additive’s degradation and, likely, lower propensity of its (or its reduction product) incorporation in the Cu deposit upon real superconfomal Cu plating scenarios.
Figure. (left) Hole filling with optimized conditions; dense and defect-free Cu fill and virtually no fill on the flat surface. (right) a higher resolution image.