The possibility of electrodeposition of soft gold from a thiosulf ate-sulfite bath was explored for electronics applications. The bath does not contain cyanide, and it is operated at a near neutral pH and a mildly elevated temperature. The bath is stable, does not undergo spontaneous decomposition without the addition of any stabilizer, and yields gold deposits with a hardness sufficiently low for use as gold bumps on semiconductor devices. Factors affecting the hardness were investigated in detail. It is shown that the use of high concentrations of the complexing agents and/or the addition of thallium(I) ions decreases both hardness and sulfur content of the deposit. The lowest Vickers hardness values achieved were approximately 80 kg mm2 in the as-deposited state and 50 kg mm2 after annealing at 350°C for 30 mm in air. The relationship between hardness and microstructure of the deposit was also examined.* Electrochemical Society Active Member.

Electrodeposition of Soft Gold from a Thiosulfate-Sulfite Bath for Electronics Applications.-A noncyanide bath for electrodeposition of soft gold bumps on silicon wafers for operation at pH 6, 60 • C, 5 mA cm −2 , and 500 rpm rotating disk electrode speed is developed containing both thiosulfate and sulfide as complexing agents. The use of high concentrations of Na 2 SO 3 and Na 2 S 2 O 3 ·5H 2 O and/or addition of Tl 2 SO 4 decreases the hardness and the sulfur content of the gold deposits. The lowest Vickers hardness of 80 kg mm −2 achieved decreases to ≈50 kg mm −2 upon 30 min annealing at 350 • C. According to SEM images, the hardness depends on the sulfur content and the size of Au crystals growth during electrodeposition. -(OSAKA, T.; KODERA, A.; MISATO, T.; HOMMA, T.; OKINAKA, Y.; YOSHIOKA, O.; J. Electrochem.

No abstract

A new method for simulating the Cu removal rate in electrochemical mechanical polishing (ECMP) based on the dissolution-type polishing mechanism was developed. The effect of a protective layer on the Cu removal rate was considered in this method because the protective layer is a key element in the dissolution-type polishing mechanism. This method was used to simulate the removal rate in a rotary-type ECMP system. The simulations accurately provided the dependence of the Cu removal rate on the aperture ratio. Furthermore, the dependence of the Cu removal rate on the aperture ratio was described with respect to changes in the average protective layer amount with time. Regarding the dependence of the Cu removal rate on the aperture diameter, however, a discrepancy was observed between the simulation and experimental results because this method did not take into account the effect of the aperture diameter on the electrolyte-filling ratio in apertures.

No abstract