Copper was supplied to Sn–3.5Ag by electroplating Cu/Ni double under-bump metallization (UBM), and the amount of Cu was controlled by varying the Cu UBM thickness. Supposed Cu contents in the solder were; 0.2, 0.5, and 1.0 wt%, respectively, and the solder joint microstructure was investigated after 1, 5, and 10 reflows. In the case of specimens with 0.2 and 1.0 wt% Cu, only one type of intermetallic compound (IMC) formed, either (Cu,Ni)6Sn5 or (Ni,Cu)3Sn4, while two types formed in specimen with 0.5 wt% Cu. No correlation could be found between the solder joint microstructure and the ball shear test. However, drop test results showed two opposite trends. The drop resistance of 0.2 and 1.0 wt% Cu specimens was quite good initially but degraded dramatically with multiple reflows, in contrast to that of the 0.5 wt% Cu specimen, which was very poor after one reflow but improved substantially later on. The former was ascribed to thickening of IMC during reflow, while the latter was related to (Ni,Cu)3Sn4 thickening beneath (Cu,Ni)6Sn5 and subsequent spalling of (Cu,Ni)6Sn5 from (Ni,Cu)3Sn4.
Correction to: Influence of Substrate Surface Finish Metallurgy on Lead-Free Solder Joint Microstructure with Implications for Board-Level Reliability
