Journal of Alloys and Compounds 486 (2009) 242–245 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jallcom Effect of surface ﬁnish on the fracture behavior of Sn–Ag–Cu solder joints during high-strain rate loading Taehoon You a , Yunsung Kim a , Woogwang Jung a , Jeongtak Moon b , Heeman Choe a,∗ a School of Advanced Materials Engineering, Kookmin University, Chungneung-dong, Songbuk-ku, Seoul 136-702, South Korea b MK Electron Co., Ltd., 316-2 Geumeoh-ri, Pogok-eup, Cheoin-gu, Yongin-si, Kyeonggi Province 449-812, South Korea article info Article history: Received 28 May 2009 Received in revised form 14 July 2009 Accepted 15 July 2009 Available online 23 July 2009 Keywords: Solder joint Sn–Ag–Cu Surface ﬁnish Surfaces and interfaces Mechanical properties Intermetallics abstract This study assesses the reliability of eutectic Sn–Pb, Sn–1.0Ag–0.5Cu, Sn–3.0Ag–0.5Cu and Sn–4.0Ag–0.5Cu solder bumps on three different pad surface ﬁnishes (ENIG, electrolytic Ni/Au and Cu- OSP) with and without an aging treatment at 150 ◦ C for 100 h. This study focused primarily on how the pad surface ﬁnish and solder alloy composition affects the reliability of solder joints using a high-speed ball pull test method. The fracture forces and failure mechanisms were also examined. The results showed that the electrolytic Ni/Au surface ﬁnish had the highest fracture forces for all four different solder alloys with and without the aging process. © 2009 Elsevier B.V. All rights reserved. 1. Introduction The reliability of a solder joint in area array packages, i.e. chip- scale package (CSP), ball-grid array (BGA) and ﬂip-chip ball-grid array (FC-BGA), has attracted considerable attention for the fol- lowing reasons. First, the ball pitch and size have been reduced due to device miniaturization, causing the solder joint to be more prone to failure during service [1]. Second, an increasing number of packages are often subject to harsh environments, such as for use in mobile devices and electronic devices in vehicles. Third, the recent transition from a conventional Sn–Pb solder to lead- free Sn–Ag–Cu, due to the increasing environmental concerns, has created uncertainty on the mechanical reliability of Sn–Ag–Cu sol- der joints, which is supported by a large volume of experimental data [2–4]. Among the many reliability issues, the impact resis- tance of lead-free Sn–Ag–Cu solder joints is of considerable concern because high-strain loading conditions are encountered frequently in portable electronic devices [1,5]. The weakness of the Sn–Ag–Cu solder joint under an impact loading condition originates from the brittleness of the IMC (intermetallic compounds) layers, such as (Cu,Ni) 6 Sn 5 or (Cu,Ni) 3 Sn, formed at the interface between the sol- der bulk and substrate pad ﬁnish, and the relatively higher stiffness of the solder ball [5,6]. To the best of knowledge, there are no reports ∗ Corresponding author. Tel.: +82 2 910 4417; fax: +82 2 910 4320. E-mail address: heeman@kookmin.ac.kr (H. Choe). on the effects of different combinations of solder ball and surface ﬁnish types under high-strain rate loading conditions. Therefore, in this study, a high-speed ball pull test method was used to simulate the actual impact resistance of Sn–Ag–Cu solder joints with three different compositions; the testing speed of the high-speed ball pull method can range from several hundred to thousand millime- ters per second, as opposed to the speed of a traditional ball pull test typically lower than 5 mm/s [7–9]. In addition, three different surface ﬁnish types in common use were also applied to examine their effect on the strength of the solder joint under high-strain rate loading conditions because the type of applied surface ﬁnish is closely related to the fracture behavior at the solder joint IMC under impact loading conditions [2]. 2. Material and methods Four types of commercially available ball-grid array (BGA) solder balls (MK Elec- tron, Korea) with a diameter of 400 m were examined. Their compositions were Sn–1Ag–0.5Cu, Sn–3Ag–0.5Cu, Sn–4Ag–0.5Cu, and eutectic Sn–37Pb (henceforth denoted as SAC105, SAC305, SAC405, and Sn–Pb, respectively). The substrate was a bismaleimide triazine (BT) with Cu pads of 250 m diameter. The following three commonly used solder pad ﬁnishes were applied: Cu-organic solderability preser- vative, electroless nickel immersion gold, and electrolytic Ni/Au. These ﬁnishes are henceforth referred to as Cu-OSP, ENIG, and Ni–Au, respectively. The solder balls were attached to the substrates in a seven-zone convective reﬂow oven (1706 EXL, Heller) under a nitrogen atmosphere. The samples were reﬂowed either twice or ﬁve times. For some samples, isothermal aging was also performed for the high-speed pull test after reﬂow at 150 ◦ C for a storage time of 100 h. The results were compared with those of the multiple-reﬂowed samples. The soldering proﬁle had a 150 ± 2 ◦ C preheat temperature with a peak temperature of 0925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2009.07.085