Loading...
Please wait, while we are loading the content...
Similar Documents
Intermetallics Formation and Growth at the Interface of Tin-Based Solder Alloys and Copper Substrates | NIST
| Content Provider | Semantic Scholar |
|---|---|
| Author | Madeni, Juan Carlos Liu, Stephen Siewert, Thomas A. |
| Copyright Year | 2003 |
| Abstract | Increasing concerns regarding environmental contamination is driving the soldering research community to develop lead-free solder alloys. Previous studies have shown that Sn-based alloys such as Sn-3.5Ag, Sn-0.7Cu, Sn-9Zn, and Sn-3.2Ag-0.8Cu have promising mechanical properties, and can be considered as serious candidates to replace the Sn-Pb alloy. However, for joint life and reliability predictions, information about the interaction between these alloys and the most used substrates is needed. In that sense, the formation and growth of intermetallic compounds at the interface between the Cu-plated substrates and Sn-3.5Ag, Sn-0.7Cu, Sn-3.2Ag-0.8Cu, and Sn-9Zn have been studied. Coupons of joints prepared with each solder alloy on a Cu-plated circuit board were subjected to thermal aging test for 20, 100, 200, 500 hours at 70, 100 and 150 degrees Celsius. Each sample was analyzed using metallographic techniques, light microscopy, SEM and EDS. The results indicate that the formation of intermetallic layers is a diffusion-controlled process. The thickness of the intermetallic compounds increased with increasing aging temperature and time. The Sn-3.5Ag alloy showed the smallest intermetallic growth and the Sn-9Zn alloy the highest. The results also suggest that at shorter aging times the chemical reaction between the substrate and the solder alloys result in the formation of small number of nucleation sites of intermetallic compounds. INTRODUCTION Lead-based solder alloys have been the joining material used for many years in electronic packaging. However, increasing concerns about the health and environmental hazards of lead drove the research community to find replacement solder alloys for the tin-lead alloys. The newly developed alloys must have mechanical, thermal, chemical, and electronic properties similar or better than the eutectic tin-lead alloy. Despite that the mechanical properties of alloys Sn-3.5Ag, Sn-0.7Cu, Sn-9Zn, and Sn-3.2Ag-0.8Cu are promising,[1] there are still other properties that need to be studied. This is the case of the formation and growth of intermetallic compounds between the substrate material and the solder alloy. It is known that these layers of intermetallic compounds are usually deleterious to joint reliability.[2] They are usually brittle and if present in large amounts at the solder/substrate interface, they may reduce the lifetime of the joints. In order to predict the extent of the effect of the intermetallic compounds between the substrate and the newly developed solder alloys, a more extensive study must be conducted. This work characterizes the formation and growth of intermetallic compounds between copper substrates and the Sn-3.5Ag, Sn-0.7Cu, Sn-9Zn, and Sn3.2Ag-0.8Cu solder alloys. EXPERIMENTAL PROCEDURE The experimental program consists of three parts: the production of copper/solder joints, the thermal aging tests, and the metallographic and microstructural characterization. Production of Cu/solder joints Two materials were used for the production of joints. One of them was a commercial Cu-plated printed circuit board (PCB). Copper was chosen as the substrate material because of its compatibility with the solder-joint systems, its frequent use in the electronics industry and its relatively low cost. The other material was one of the four lead-free solder alloys under investigation: Sn-3.2Ag-0.8Cu, Sn-3.5Ag, Sn-0.7Cu, and Sn-9Zn. Their respective melting temperatures are 217C, 221C, 227C, and 198C. The substrates for all the coupons were cut into pieces of 25.4 x 25.4 mm (1x1 in.). The Cu surface was cleaned with a water-based solution of ammonium hydroxide, trisodium phosphate and sodium tetraborate pentahydrate (M-prep neutralizer). After letting the Cu surface dry for a few seconds, a solution of water-based phosphoric acid (M-prep conditioner) was applied. These solutions removed any impurities, dirt, and oxide layers present on the surface, and promoted wetting. Next, the molten solder alloy at 50C above its melting temperature, would be deposited on the substrate and cooled in air to ambient temperature. In the case of alloy Sn-9Zn, the copper surface was prepared using commercial zinc chloride and ammonium chloride based paste, which provided better wetting than the phosphoric acid-based solution chosen for the other three solder alloys. Thermal aging test The thermal aging test of the Cu/solder joint coupons was done using furnaces with temperatures controlled to ± 1C. To avoid oxidation during thermal aging, a continuous flow of industrial argon gas was provided to the furnaces. The Cu/solder joints were thermally aged at 70, 100 and 150 C, and for 20, 100, 200 and 500 hours. The specimens were water-quenched as they were removed from the furnaces. Metallographic characterization The samples were cut in half and cold mounted using epoxy for metallographic analysis. The reason for using cold mounting instead of hot mounting is that cold mounting undergoes a lower curing temperature. The mounted samples were ground and polished using standard metallographic techniques, first to 0.5 μm with diamond slurry, and then to 0.05 μm using colloidal silica. Polishing the samples in a vibratory machine with colloidal silica resulted in a good etched surface. The samples were examined using light microscopy, scanning electron microscope (SEM), and energy dispersive spectroscopy (EDS). Quantitative measurements of the intermetallic compounds were made using a digital image analysis system. RESULTS AND DISCUSSION Cu/solder interface The Cu/solder interface after thermal aging was found to be very different from the interface before the thermal test. Before, the most noticeable features of the interface were the epoxy board, the copper substrate, and the solder. Only at high magnifications can very small amounts of Cu6Sn5 be found at the Cu/solder interface.[3] After the thermal aging test, the appearance of the microstructure at the Cu/solder interface changed significantly. Evidence of the changes is shown in Figure 1 in the micrographs of the Cu/Solder interface at 150C for 500 hours. Beginning from the bottom, in Figures 1(a), (b) and (c), the visible layers are the epoxy board, the copper substrate, the intermetallic compounds (IMCs) Cu3Sn and Cu6Sn5, and finally the solder matrix. It was observed that the layer of Cu3Sn forms at the Cu/solder interface, and the layer of Cu6Sn5 on top of Cu3Sn. This sequence is true for the three material systems, Cu/Sn-3.2Ag0.8Cu, Cu/Sn-3.5Ag, and Cu/Sn-0.7Cu. In the case of Cu/Sn-3.5Ag, a few small particles of Ag3Sn can also be observed at the Cu6Sn5/Sn-3.5Ag interface. Different from the previous joint systems, the Cu/Sn-9Zn system displays only one intermetallic layer, Cu5Zn8. The intermetallic compounds were identified using electron diffraction spectroscopy (EDS) analysis. Evidence of morphological differences at the interface of the two IMCs was determined. The Cu3Sn/Cu6Sn5 interface was observed to be irregular.[4] The variation in thickness of the Cu3Sn intermetallic layer was not as drastic as the thickness of the Cu6Sn5 intermetallic layer. In the case of the Cu6Sn5/solder-matrix, the interface is more irregular, with localized developments of elongated, needle-like, or elongated scalloped structures, growing far into the solder matrix. Sn-3.2Ag-0.8Cu Sn-3.5Ag Ag3Sn particle Cu6Sn5 Cu6Sn5 Needle-like Ag3Sn Cu3Sn Cu3Sn Cu substrate Cu substrate Epoxy board Epoxy board |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://ws680.nist.gov/publication/get_pdf.cfm?pub_id=851327 |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
