A Comparison of Corrosion Behavior of a Super Duplex Stainless Steel and an Austenitic Stainless Steel in a Molten Sn 3.0Ag 0.5Cu Lead-Free Solder

Transcription

1 Materials Transactions, Vol. 53, No. 6 (2012) pp to The Japan Institute of Metals A Comparison of Corrosion Behavior of a Super Duplex Stainless Steel and an Austenitic Stainless Steel in a Molten Sn 3.0Ag 0.5Cu Lead-Free Solder Huei-Sen Wang 1,+, Ken-Do Hsu 1, Mei-Hui Wu 2 and Yan-Zong Su 1 1 Department of Materials Science and Engineering, I-Shou University, Kaohsiung 84001, Taiwan, R. O. China 2 New Materials Research and Development Department, China Steel Corporation Kaohsiung, Taiwan, R. O. China To determine the endurance of a super duplex stainless steel (SDSS) used for wave soldering bath materials, the corrosion behaviors of a SDSS, SAF2507, and a comparative austenitic stainless steel, SUS304L (conventional material for tin-lead soldering container) in a Sn 3.0Ag 0.5Cu molten lead-free solder were investigated. After testing, the samples were analyzed by optical microscopy (OM), scanning electron microscopy (SEM), and semi-quantitative phase identification under an energy dispersive spectrometer (EDS) to evaluate the effects of the composition of test materials and immersion conditions on their microstructure evolution and corrosion behaviors. As results show, when compared to the SUS304L, SAF2507 has better corrosion resistance to lead-free solder after immersion at the assigned temperatures (350, 450, and 550 C) and times (from 250 up to 1500 h). When the test temperatures of 350 and 450 C were employed, no obvious dissolution occurred for SAF2507, whereas SUS304L exhibited severe dissolution. However, if the immersion temperature of 550 C was used, the dissolution rates of SAF2507 increased significantly. It was found that the failure type of both materials was related to atom diffusion, formation of the reaction layer (RL), and finally dissolution, which is a typical failure type of Liquid Metal Corrosion. Moreover, SEM and EDS results reveal that the major intermetallic phases in the RL for both stainless steels are Fe/Sn and Cr/Sn compounds. [doi: /matertrans.m ] (Received December 20, 2011; Accepted March 23, 2012; Published May 16, 2012) Keywords: lead-free solder, super-duplex stainless steel, reaction layer, liquid metal corrosion 1. Introduction The use of lead-free solder in the electronics industry has increased dramatically over the past decade due to environmental protection legislation. Most research projects related to lead-free solder have focused on the physical properties 1,2) (e.g., thermal or mechanical properties) of new solders as well as their reliability and performance, since this is the most important issue at present. When changing to a lead-free solder, one has to consider that a higher percentage of tin in an alloy will enhance the operation temperature 1,2) (normally C higher than that of tin-lead solder) and this will increase potential corrosion problems in relation to the solder machine parts, i.e., solder bath or impeller. Therefore, the equipment s compatibility with a lead-free solder becomes an important concern for all related equipment manufacturers. For equipment manufacturers, corrosion protection for solder bath components can be split into two categories: (a) homogenous solder corrosion resistant materials (HSCRM) without any surface treatment and (b) those with surface treatment (e.g., coating) on the homogenous materials (STHM). 1) For HSCRMs, conventional 300 series stainless steel (e.g., SUS304L) is generally used for the tin-lead solder baths with excellent longevity results. 3) However, when used for a lead-free solder, high dissolution rates may occur. 4 7) To combat this problem, many alternative materials, such as titanium or gray cast iron, have been investigated. 8) However, the use of such materials has inherent limitations, for example, high tooling costs in the case of gray cast iron and high material costs in the case of titanium. Providing a possible alternative choice of HSCRM in the context of solder containers becomes a major work in this study. Earlier studies suggested that iron based materials with a higher Cr (e.g., SUS 309S) 2) and Mo (e.g., SUS 316), 4) and less Fe + Corresponding author, huei@mail.isu.edu.tw contents may retard the dissolution rate from the molten leadfree solder. Taking into account the above considerations, a super duplex stainless steel (SDSS), SAF2507, which has both a higher Cr (7% more) and Mo (4% more) content when compared to SUS304L, was selected for evaluation for this study because of its feasibility as an HSCRM. The selected SDSS was tested along with an SUS304L for comparative purposes. The materials were immersed in a commercial Sn 3.0Ag 0.5Cu molten lead-free solder with a variety of immersion times and exposure temperatures. After immersion testing, microstructure evolution and corrosion behaviors of the tested samples were investigated. From the test results, the endurance of both stainless steel materials as an HSCRM was determined. 2. Materials and Experimental Methods Material samples for molten lead-free solder endurance tests were cut from 3.0 mm thick plates and machined into strips approximately 20 mm wide, 50 mm long, and 1.5 mm thick. The chemical composition of selected materials used in this study is shown in Table 1. Before testing, all of the sample plates were polished with 1000-grit SiC paper to remove surface contamination and then cleaned with pure alcohol. To determine the mass ratio of specimen/solder used for each test, mass of the polished plates and the solder were measured. The mass of SUS304L plate is similar to the mass to SAF2507 plate which is approximately 12 g. The solder used for each sample is approximately 515 g. The samples were then immersed vertically in a crucible with molten Sn 3.0Ag 0.5Cu lead-free solder, which is the most promising lead-free solder for a wave soldering process, at various temperatures: 350, 450 and 550 C. The temperature ranges selected in the study are extremely high compared to industrial standardized lead-free solder bath temperatures (around 280 C). 7) However, when considering

2 A Comparison of Corrosion Behavior of a Super Duplex Stainless Steel and an Austenitic Stainless Steel 1149 Table 1 Chemical composition of the stainless steels used for the dissolution tests (mass%). Cr Ni Mo N Mn Si Fe SUS304L bal. SAF bal. Fig. 1 Definition of the maximum dissolution depth (D max ) for inhomogeneous dissolution plates. the surface temperature of the sheathed heater of a solder bath at the initial stages of the power-on, the temperature ranges were chosen accordingly. 2) After immersion for the assigned times, from 250 to 1500 h, tested samples were removed from the molten solder and cross-sectioned. The tested samples were observed in order to measure the dissolution depth. The definition of maximum dissolution depth refers to that described by Takemoto et al., 2) shown in Fig. 1. Furthermore, to investigate the corrosion mechanism after immersion tests, microstructure evolution of tested samples was evaluated under OM, SEM and EDS. 3. Results and Discussion 3.1 Corrosion depth analysis Figure 2 shows the effects of immersion time and temperature on the maximum dissolution depth (D max ) of SUS304L and SAF2507 in Sn 3.0Ag 0.5Cu molten solder. For both materials, it is clear that D max increases with the soaking temperature and time. However, the SUS304L has a higher dissolution rate when compared to SAF2507. For SUS304L, diffusion and mirror dissolution occurred after immersion for 1000 h at 350 C (Fig. 3(a1)). When the immersion time is longer (e.g., Fig. 3(b1) 1500 h at 350 C) or soaking temperature is higher, dissolution increases. As the temperature was elevated to 550 C (Fig. 3(c1)), massive dissolution was observed. For SAF2507, no dissolution was observed after immersion for the assigned times at 350 (Fig. 3(a2)) or 450 C (Fig. 3(b2)). Dissolution was only observed as the immersion temperature rose to 550 C (Fig. 3(c2) is an example of immersion for 1000 h at 550 C). Moreover, as massive dissolution occurs, D max and immersion time have a parabolic relationship, which indicates a slight reduction in the dissolution rate with immersion time. Earlier studies suggest 4) that the driving force for the dissolution rate of solid metals into molten liquid metal is (Cs ¹ C); here Cs is the saturation concentration of the solute in liquid metal and C is the concentration of the solute in liquid after reaction time. As the RL around the test plate forms, the concentration of the solute in the molten solder, C, increases, especially at the Fig. 2 D max of SUS304L and SAF2507 after immersion testing. interface of the solder/stainless steel. The increase of C reduces the concentration difference, (Cs ¹ C), which directly reduces the dissolution rate. This can be considered the dominant factor for the slight reduction in the dissolution rate with immersion time. 3.2 Microstructure observation of SUS304L after immersion tests Figure 4(a1) shows an example of the cross-sectional photomicrographs at earlier stages of dissolution which can be observed at a lower immersion temperature after immersion for an intermediate time (e.g., 350 C, after immersion for 1000 h) or shorter periods of immersion at a higher immersion temperature (e.g., 450 C, after immersion for 250 h). From the EDS analysis (Fig. 4(a2)), it was observed that the inhomogeneous Sn diffusion layer was initially formed in the SUS304L matrix. That is, Sn diffuses into the stainless steel matrix and Fe, Cr and Ni from the surface of matrix diffuse into the solder; however, no significant dissolution occurred at this stage. As the immersion temperature and time were increased, large and irregular dissolution was observed. Figures 4(b1), 4(c1) and 4(d1) show three examples after exposure to molten solder for longer periods of immersion time at 450 C. These samples show very similar and relatively thick RLs. Chemical analysis of the RLs was performed in an SEM equipped with EDS. Figures 4(b2) to 4(d2) and 4(b3) to 4(d3) show the local compositions in the vicinity of SUS304L matrix/rl interface and RL/solder interface, respectively. As seen in Figs. 4(b2) to 4(d2), the EDS semi-quantitative analysis indicates that the RL consists of Sn, Fe and Cr. In Figs. 4(b3) to 4(d3), the RL also consists of Sn, Fe, but Cr concentration in this area is hardly detected. Figures 5(a) and 5(b) show a higher magnification of the RL. EDS analysis reveals the Fe/Sn and Cr/Sn (see Table 2) 2,3) compounds formed in the RL. 3.3 Microstructure observation of SAF2507 after immersion tests Figures 6(a) and 6(b) show the cross sectional images after 1500 h (longest immersion time) at 350 and 450 C, respectively. These figures show the typical microstructures of duplex stainless steel which consists of a mixture of Face-

3 1150 H.-S. Wang, K.-D. Hsu, M.-H. Wu and Y.-Z. Su Fig. 3 Test results for SUS304L after immersion for (a1) 1000 h (b1) 1500 h at 350 C and (c1) 1000 h at 550 C, and for SAF2507 after immersion for 1500 h at (a2) 350 C (b2) 450 C and (c2) 1000 h at 550 C. Table 2 EDS analysis of the reaction layer in SUS304L (atomic%). Zone 1 Elements Cr Fe Ni Cu Ag Sn Zone Zone Centered Cubic (FCC) austenite ( ) islands in Body-Centered Cubic (BCC) ferrite ( ) grains. No obvious RL was observed under those temperatures after 1500 h of immersion. However, after 1500 h of immersion at 450 C, grain sizes of the ferrite ( ) phases grew obviously. Furthermore, the EDS line scan indicated that the earlier stages of a Sn diffusion layer may initially form after a 1500 h immersion test at 450 C (see Fig. 7). As the temperature was elevated to 550 C, a thicker RL was observed, especially in the specimens undergoing a longer test period. Figures 8(a1) to 8(b1) show the two examples after exposure the longer periods in molten solder at 550 C. From these figures, it was observed that the dissolution morphology of these samples is very similar to the dissolution morphology of SUS304L samples. However, the SAF2507 dissolution rate was much lower when compared to SUS304L. From EDS analysis (Figs. 8(a2) and 8(b2)), in the vicinity of SAF2507 matrix/rl interface, it was observed that the main constituents in the RL are Sn and Fe. The presence of small amounts of Cr in the RL was also confirmed. However, in the vicinity of SAF2507 RL/ solder interface, Fe and Cr contents in the RL are further reduced. EDS line scan analysis (Fig. 9) also reveals Fe/Sn and Cr/Sn compounds formed in the RL of the SAF2507 test plates. 3.4 Comparison of the corrosion mechanism of the immersion samples Regarding their chemical compositions, SAF2507 has a higher content of Cr than SUS304L and an additional

4 A Comparison of Corrosion Behavior of a Super Duplex Stainless Steel and an Austenitic Stainless Steel 1151 Fig. 4 Cross-sectional photomicrographs of a SUS304L reaction layer (RL): (a1) earlier stages of immersion tests, 450 C, 250 h; (b1) higher immersion temperature and longer periods of time, 450 C, 500 h; (c1) 450 C, 1000 h and (d1) 450 C, 1500 h. (a2) to (d2) are EDS analyses in the vicinity of SUS304L matrix/rl interface obtained from (a1) to (d1), respectively. (b3) to (d3) are EDS analyses in the vicinity of RL/solder interface obtained from (b1) to (d1), respectively. Fig. 5 Higher magnification of the RL in SUS304L. constituent of Mo. The main advantage of adding Cr to steel is to improve the corrosion resistance through the formation of a passive film. This advantage can be further enhanced in the presence of Mo.9) Numerous examples in the literature2,3) indicate that the stable passive film also offers extremely poor wetting with solder and aids against the attack from molten solder. This may be explained insofar as SAF2507 has a greater endurance to molten lead-free solder. Nevertheless, for SAF2507, dissolution still occurred at higher immersion temperatures (550 C).

5 1152 H.-S. Wang, K.-D. Hsu, M.-H. Wu and Y.-Z. Su It has been established9) that if the SDSS operates within certain temperature ranges, Cr and Mo have negative effects due to their encouragement of the formation of a precipitation phase, such as alpha prime ( A) within the duplex stainless steel. Alpha prime ( A) is the lowest temperature decomposition phase in a duplex stainless steel, occurring between 300 and 525 C; this is the main cause of hardening and 475 embrittlement. To avoid A precipitation, it has been suggested that duplex stainless steel only be used at Fig. 6 Cross-sectional images of SDSS after 1500 h testing at: (a) 350 C and (b) 450 C. temperatures under 300 C.9) However, SDSS failure caused by A precipitation does not seem to be the case in this study. Based on experimental observations, it is found that failure mechanisms for SAF2507 are very similar to SUS304L, summarized as follows: Step 1 Sn atoms diffuse into the stainless steel. Meanwhile, Fe, Cr, and Ni diffuse into the solder and a RL is formed. Step 2 Fe, Cr, and Sn form compounds in the RL. Step 3 Sn keeps passing through the unprotected HSCRM and the RL becomes thicker. The dissolution of the RL begins. Step 4 Part of the RL pieces break down and, simultaneously, Sn keeps diffusing into the unprotected HSCRM. A new RL is formed. The dissolution rate is then reduced, due to the concentration of the solute in the molten solder (e.g., Fe, Cr, Ni, increases). Fig. 7 EDS line scan of SDSS: earlier stages of Sn diffusion layer initially formed in the SAF2507 matrix after 1500 h of immersion testing at 450 C. Fig. 8 Two samples of SDSS after exposure to molten solder for assigned times at 550 C: (a1) 1000 h, (b1) 1500 h. (a2) and (b2) show EDS analyses in the vicinity of SAF2507 matrix/rl interface obtained from (a1) and (b1), respectively. (a3) and (b3) show EDS analyses in the vicinity of RL/solder obtained from (a1) and (b1), respectively.

6 A Comparison of Corrosion Behavior of a Super Duplex Stainless Steel and an Austenitic Stainless Steel 1153 Fig. 9 The above failures are related to atom diffusion, formation of RLs, and finally dissolution. These are typical results of liquid metal corrosion. In addition, although severe dissolution for SAF2507 did occur at higher immersion temperatures (550 C) in this study, for most lead-free solders the process temperatures are lower than 280 C. Additionally, for real soldering processes, high temperature duration at initial stages of the power-on would not be sustained as long as the process occurs within the parameters of this study. Therefore, from the experimental results, it is found that SAF2507 is still a feasible option in the context of HSCRM. 4. Conclusions (Fe/Sn) and (Cr/Sn) compounds formed in the RL. To determine the endurance of a super duplex stainless steel (SAF2507) used for wave soldering bath materials in a molten lead-free solder (Sn 3.0Ag 0.5Cu), dissolution tests on the SAF2507 and a comparative material, SUS304L, were performed. From the test results, conclusions can be made as follows: (1) SAF2507 has better corrosion resistance to lead-free solder than SUS304L. With test temperature below 450 C, no obvious dissolution occurs for SAF2507, whereas SUS304L results in severe dissolution. (2) For SUS304L, an earlier stage of the Sn diffusion layer may initially form after 1000 h of immersion at 350 C. After a longer test period, dissolution occurs. (3) For SAF2507, an earlier stage of the Sn diffusion layer may initially form at 450 C after 1500 h of immersion. As the temperature is elevated to 550 C, a massive RL and dissolution are observed, especially in specimens with longer test periods. (4) EDS analysis reveals that Fe/Sn and Cr/Sn compounds are formed in the RLs of both selected materials. (5) In this study, the failure of both materials was related to atom diffusion, formation of a RL and finally, dissolution. It was confirmed that these were all results of liquid metal corrosion. REFERENCES 1) J. Morris, M. J. O Keefe and M. Perez: Glob. SMT & Pack. Mag. 7 (2007) ) T. Takemoto and M. Takemoto: Sold. Surf. Mt. Tech. 18 (2006) ) T. Takemoto and M. Takemoto: Proc. Eco Design 2005: Fourth International Symposium on Environmentally Conscious Design and Inverse Manufacturing, (2005) pp ) T. Gyemant: Glob. SMT & Pack. Mag. 4 (2004) ) K. Sweatman, S. Suenaga, M. Yoshimura and T. Nishimura: APEX (2004) pp. S S ) H. Nishikawa, S. Kang and T. Takemoto: Q. J. Japan Weld. Soc. 27 (2009) 214s 218s. 7) K. Sweatman: Glob. SMT & Pack. 4 (2006) ) J. Morris and M. J. O Keefe: Appliance 61 (2004) ) R. N. Gunn (ed.): Duplex stainless steels: microstructure, properties and applications, (Woodhead Publishing Ltd., England, 1997).