Characterization of electroplated bismuth-tin alloys for electrically conducting materials

Journal of Electronic Materials, Oct 2000 by Kang, S K, Buchwalter, S, Tsang, C

EXPERIMENTAL PROCEDURES AND RESULTS

Materials Preparation

This research did not use coated copper powders in its experiment because of the difficulties in analyzing micron-sized particles. Instead, flat copper foils were coated under the same conditions to simulate the powder form. The flat surface allowed for examination with several different analysis techniques. For example, resistivity measurements would not have been possible with the particulate form. In addition, to avoid surface interference, non-conductive silicon wafers coated with copper were used in the resistivity measurements because copper foils would have contributed to a higher conductivity.

Sample Set A were Cu foils (2.54 cm x 2.54 cm) which were beaker-plated on both sides with BiSn. A commercial BiSn plating solution was used. Sample Set B consisted of BiSn thin films electroplated onto the Cu-coated side of a silicon wafer, 2.54 cm x 2.54 cm in area. Variables in both types of electrodeposition included current density and plating time. The thickness of BiSn was also varied since plating time is directly related to film thickness. The samples were plated for 1, 2,4, and 8 min. Selected samples from Set A and B, plated for 4 min, were dried with a nitrogen gun and heat treated in a temperature controlled oven in air atmosphere at 50*C, 1000C, 2000C for 0.25 h, 1 h, 4 h, and 24 h, respectively.

Differential Scanning Calorimetry (DSQ

Differential scanning calorimetry was used to determine the enthalpy changes in the material. In addition, endothermic peaks of the enthalpy changes, along with the BiSn binary phase diagram, identified the composition of the BiSn thin film. This analysis was conducted on a DSC220C from Seiko Instruments, Inc. Small circular samples of BiSn coated Cu foil (Set A) were cut with surgical scissors and sealed in aluminum sample holders. The samples were then heated in the DSC at a rate of 10*C/min from 30'C to 300'C.

A typical thermogram for the 4 min sample is shown in Fig. 1, where a strong endothermic peak is observed at 1390C matching with the BiSn eutectic composition. The DSC data can be used to estimate the weight fraction and thickness of the plated BiSn thin film. The BiSn film thickness was also measured independently by using an AlphaStep 200.

X-Ray Photoelectron Spectroscopy (XPS)

In order to determine the surface oxide composition, the following Set A samples were examined by xray photoelectron spectroscopy: as plated for 4 min, 50*C for 1 h, 500C for 4 h, 50'C for 24 h. Curve fitting was applied to the peaks of Sn and Bi to determine the oxide to metal ratio of each element. XPS spectra were taken with a Perkin-Elmer Phi 5500 system with Mg Ka excitation. The electron take-off angle was 450 from the samples' surfaces. XPS analysis of BiSn plated on Cu foils was successful in identifying the type of surface oxide growth. Figure 2 depicts the general spectrogram of the as-plated sample. Sn and Bi peaks were identified at approximately 500 and 540 eV. The Sn concentration was much higher than that of Bi, 25.5% and 3.7%, respectively. With the curve fitting techniques, it was determined that the oxides on the surface are primarily tin oxides, in particular SnO or SnO,. This spectrogram is presented in Fig, 3, where 87.5% of the Sn was related to oxides while 12.5% came from the metal. Curve fitting analysis of the Bi peaks showed that the bismuth associated with oxides versus metals was in a ratio of approximately 50:50.