Interfacial reaction between Sn-3.0Ag-0.5Cu solder/Co-P plating and Ni-Co-P plating

Abstract

In response to health and safety concerns, lead-free soldering has become a popular technology in electronics packaging. Compared with the lead-containing solders, Sn- 3.0mass%Ag-0.5mass%Cu (SAC, all mass% unless specified otherwise) solder widely used in Japan has a relatively low impact reliability owing to the solder alloy hardness that induces a high stress concentration at the interface. In general, there is a correlation between the impact reliability and the morphology and thickness of the reaction layer formed at the solder/under bump metallurgy (UBM) interface. The most common UBM is electroless Ni-P plating over copper pad. Electroless Ni-P acts as a diffusion barrier layer between the copper and the solder. However, due to nickel diffusion, P-rich layers form at the interface between the solder and electroless Ni-P. Solder joint failure is related to the growth of these layers and to their brittleness and affects the mechanical reliability of joints. Recently, a new composition of UBM is proposed as diffusion barrier 1-4) . For instance, Magagnin et al. reported that electroless Co-P strongly limits interdiffusion and intermetallic compounds formation as compared with the electroless Ni-P with Sn-Ag-Cu alloy. Furthermore, in the Co-P samples, P-rich layers did not form at the interface 4) . It is important to investigate the relationship between morphology of reaction layer and UBM. This study aims to clarify the effect of Co-P and Ni-Co-P on the morphology of reaction layer formed at the solder/UBM interface. 2. Experimental SAC solder (0.3 g) was used in this study. Electroless Co-P(Au) (3.1 Pm) and electroless Ni-Co-P(Au) (5.2 Pm) finished Cu plates on FR-4 PCBs (25.0×25.0×1.6 mm) were prepared as UBM. Electroless Ni-P(Au) (5.0 Pm) substrate was also used as a reference substrate. These substrates were plated with gold to avoid oxidation of the cobalt and nickel surface. The experimental procedure is shown in Fig. 1 .T he substrate was immersed in 4% HCl solution for 120 s and then rinsed with deionized water. Then, solder was put on the center of the substrate and activated flux (0.01 ml) was dropped on the solder. The test specimen was put into a radiation furnace in a nitrogen atmosphere and heated according to the temperature rise profile shown in Fig. 2. The reflow peak temperature was 513 K with the sample above 490 K for 115 s. After soldering, the spreading area of the solder on the UBM was measured by using the optical microscope (OM).�Three tests were conducted to obtainan average value for each specimen. Then, specimens were cut and the cross-section of the specimens was polished to observe the interface between the solder and UBM. The reaction layer at the interface was observed by scanning electron microscope (SEM).