An inexpensive, fast, selective, and sensitive technique for surface area measurement of metallic nanoporous materials (MNPM) is developed, systematically tested, and validated. The approach employed is based on underpotential deposition (UPD) of metals on foreign substrates. In this work, Pb UPD on Au is chosen to illustrate the applicability of and reveal the advantages and limitations of the proposed method. Experiments are designed for surface area measurement of nanoporous gold (NPG) electrodes with pore sizes in the range of 5−15 nm, prepared by electrochemical dealloying of single phase AupAg1−p (atomic fraction p = 0.1, 0.2, and 0.3). Dealloying is performed galvanostatically at a current density of 1 mA cm−2 in a AgClO4 solution, acidified to pH 1. The experimental results suggest a linearly increasing charge in the Pb UPD layer with NPG thickness. This finding hints at (i) uniformity of the NPG structure and (ii) the general ability of this method to work for analysis of bulk materials. The proposed approach is tested by studying the dependence of the NPG surface area upon the original alloy composition and correlating the results with the NPG structure and morphology imaged by high-resolution scanning electron microscopy. An anomalously high surface area is registered in dealloyed Au0.1Ag0.9 samples and is attributed to the lack of a pre-existing percolation backbone. Unlike the instantaneous Pb UPD process on a flat metal surface, the slow and thickness-dependent kinetics of Pb layer formation on NPG is associated with hindered mass transport through pores. Further validation of the Pb UPD method is made by experimental monitoring of heat treatment-enforced coarsening and the basic modeling of the correlation between surface area and ligament size in NPG. Finally, a critical comparison with Brunauer−Emmett−Teller (BET) analysis reveals important advantages of the developed method for surface area measurement in MNPM specimens.

Interfacial voiding in solder joints formed with Sn-Ag-Cu solder alloys and electroplated Cu was examined as a function of the plating solution chemistry and parameters. Galvanostatic Cu plating of *10 lm thick Cu films was performed in a commercially available plating solution, and in model generic plating solutions. Analysis of the current voltage behavior along with Secondary Ion Mass Spectrometry studies of organic impurity content of two plated and a wrought copper samples, yielded a conclusion that for certain chemistry solutions (e.g., H 2 SO 4 + CuSO 4 + Cl -+ PEG) and current density ranges above 2.5 mA cm -2 , organic impurities were incorporated into the growing Cu. Solder joints were produced with a variety of electroplated Cu samples. These joints were, then, annealed at a temperature of 175°C for 1 week, cross sectioned and examined. In general, it was observed that interfacial voiding in laboratory electroplated Cu layers was qualitatively similar to the unexplained voiding observed in some industrially plated Cu products. More specifically, it was found that the propensity for voiding could be correlated with specific electroplating parameters that in turn were associated with significant incorporation of organic impurities in the Cu deposit.

The sporadic voiding phenomenon in Cu 3 Sn intermetallic compound (IMC) formed during thermal aging, sometimes referred to as "Kirkendall voiding", has been found to lead to degradation of solder joint reliability in board level shock testing. It was suggested that the voiding phenomenon resulted from the incorporation of specific impurities in the copper during electroplating. In this study, Cu samples were electroplated from a generic suppressor-brightener additive system using a rotation disk electrode (RDE) apparatus. Overpotential during plating, surface morphology and the propensity for voiding of plated samples were investigated. Galvanostatic (constant current density) plating was conducted at 10 mA/cm 2 sequentially up to 18 hours. The solution exhibited dependences of overpoential and voiding propensity on bath aging, due to the breakdown of the organic additives. Cu samples were also plated in the current density range of 0.8-40 mA/cm 2 . In the 10-20 mA/cm 2 , current density range, a fine-grain, smooth, deposit surface was observed, accompanied by an especially low voiding level of samples plated in that range.

A quantitative study of the impact of key Cu plating parameters on the voiding propensity of solder joints with Cu electroplated in a commercially available plating solution (CAPS) is performed first on 0.3 cm 2 Cu rotating disk electrode. It is shown that similar to samples plated in a generic plating solution (GPS) containing bis(3-sulfopropyl) disulfide, polyethylene glycol, and Cl -ions, void-prone samples are deposited predominantly at higher overpotentials, in the range from positive to -0.20 V. In the second part, a Hull cell with 46 cm 2 cathode is used to scale up the voiding study in both, GPS and CAPS. It is demonstrated that plating conditions could be chosen in a way to generate both, void-prone and void-proof Cu on the same cathode panel. Thus, the controlled voiding propensity illustrated for the first time in a prototype of industrial Cu plating helps in realizing the sporadic nature of the voiding phenomenon.

No abstract