The Influence of Resin Coverage on Reliability for Solder Joints Formed by One- Pass Reflow Using Resin Reinforced Low Temperature Solder Paste

Transcription

1 2017 IEEE 67th Electronic Components and Technology Conference The Influence of Resin Coverage on Reliability for Solder Joints Formed by One- Pass Reflow Using Resin Reinforced Low Temperature Solder Paste Atsushi Yamaguchi, Ph.D., Yasuo Fukuhara, Andy Behr, Hirohisa Hino, Yasuhiro Suzuki, Naomichi Ohashi Electronic Materials Business Division, Automotive & Industrial Systems Company Panasonic Corporation Kadoma, Osaka, Japan Abstract SAC305 solder paste is commonly used electronic assembly. This solder alloy consists of 96.5% tin, 3% silver, and 0.5% copper and melts at 219 C. The peak reflow temperature range is typically C. With electronic devices such as smartphones, notebook PCs, and tablets becoming thinner, packaging substrates, such as ultra-thin flip chip ball grid arrays (FCBAs), and the printed circuit boards (PCBs) on which they are mounted are becoming thinner. The growing use of thinner substrates is creating manufacturing and reliability challenges. For example, it is increasingly difficult to control the warpage of CPU packages in notebook PCs during the solder reflow process. The result is greater numbers of solder joint defects, including Non-Wet Open (NWO) and Head-on-Pillow (HOP) defects caused by warpage of package substrates and PCBs. These issues have created a demand for low-temperature solders to help reduce warpage and improve SMT assembly yields by adopting lower soldering temperatures. Tin Bismuth (SnBi) eutectic solders have a desirably low melting point of 139 C. However, the brittleness of the alloy limits commercial use. This situation prompted us to develop a solder paste material that combines low temperature SnBi solder with epoxy resin. This approach enables the concurrent formation of SiBn solder joints and a reinforcing polymer collar via a one pass reflow process. This paper describes the solder joint properties and reliability of this low temperature joint reinforced solder paste (JRP) developed by us, which consists of SnBi solder compounded with epoxy resin. We evaluated the influence of the epoxy resin component in the developed material (JRP) on solder joint reliability. We compared the joint properties of samples made SAC305 solder paste, unreinforced SnBi solder paste and JRP solder paste. The evaluation revealed that the JRP technology alleviates issues associated with the brittleness of SnBi solder by encasing the formed solder joints with a fully cured epoxy resin. Ball joint shear testing, BGA solder joint strength testing, temperature cycle testing, and drop shock testing revealed that low temperature JRP solder paste demonstrated equivalent or better joint properties than those made with SAC305 solder paste. Keywords-solder solder joint; low temperature solder; resin coverage; reliability; temperature cycle test; drop shock test; low temperature joint reinforced solder paste; Tin Bismuth solder; BGA reliability; solderball reliability; underfill; I. INTRODUCTION Solder pastes containing an alloy of 96.5 % tin, 3% silver, and 0.5% copper, commonly referred to as SAC305, are frequently used in electronic assembly to interconnect surface mount technology (SMT) packages like ball grid arrays (BGAs), Land Grid Arrays (LGAs) and quad-flat noleads (QFNs) onto printed circuit boards (PCBs) for many types of products. While this alloy is widely used, the high processing temperatures create thermally induced stresses which can result in reliability challenges. SAC305 has a 219 C melting point and the peak reflow temperature range is typically in the range of 240 C to 260 C. As printed circuit boards decrease in thickness, warpage issues resulting from elevated reflow temperatures like those used for processing SAC305 are exacerbated. Reflow induced PCB warpage has been demonstrated to result in a number of solder joint defects such as Head-in-Pillow, Non-Wet Open and solder bridging [1]-[3]. Low reflow temperature Bismuth (Bi) based alloys are being evaluated to as an alternative to SAC305 in order to reduce thermally-induced circuit board warpage and resultant solder joint defects. Bismuth based solder alloys are well recognized as the leading low temperature solder alloy candidates [4]-[6] for SMT assembly. Tin Bismuth (SnBi) solders have much lower melting points than SAC305. For example, a eutectic solder alloy of 58% Tin and 42% Bismuth (58Bi/42Sn) has a 139 C melting point. This is a difference of 80 C. However, the inherent brittleness of Bibased solder alloys has limited their commercial adoption for high volume electronic products like mobile phones and portable computers. Polymer resin reinforcement of solder joints formed with SnBi solder is one approach for addressing for brittleness of interconnects made with these alloys. This joint reinforcement scheme is particularly attractive for mobile devices that are commonly subjected to drop-shock type stresses. Traditional methods of resin reinforcing solder joints such as underfilling and corner bonding generally occur after reflow processing. The additional materials and process /17 $ IEEE DOI /ECTC

2 comparing by comparing solder joints formed with JRP Bismuth-Tin-Silver and SAC305. II. EXPERIMANTAL DESIGN AND EXPERIMENTAL PROCEDURE steps associated with additional post reflow processing increase manufacturing costs. Additionally, the more components assembled on each printed circuit board requiring resin reinforcement, the greater the cost impact. Researchers are also conducting studies on the addition of dopants to SnBi solders to reduce the alloy brittleness and improve solder joint reliability. However, solder joints formed with doped Bi-based solder alloys achieving the reliability level of SAC 305 joints have not yet been reported [7]-[9]. Rather than using underfills, corner bonds or improving the brittleness of Bi-based solder alloys via dopant additions, we are developing joint reinforced solder paste (JRP) technology combining a polymer resin and a low temperature Bismuth based solder. For this technology, a specialized epoxy resin is blended with SnBi particles and other compounds to form a solder paste. Conventional solder pastes are generally comprised of solder, rosin, acid and other additives. The JRP material consists of solder, epoxy resin, hardener, acid and additives. As shown Figure 1, the solder joints are reinforced by a collar of fully cured epoxy resin after a single reflow cycle at 160 C. At the beginning of the reflow process, the resin migrates to the exterior of the liquidus solder and coalesces around the circumference of the solder joint while the solder fillet is in the process of forming. As the reflow process progresses, the epoxy resin polymerizes. By the end of the reflow process, the resin is fully cured and no additional processing is required. This polymer reinforcement improves reliability of solder joints without the need for additional materials or processes. JRP technology is a viable candidate for replacing SAC 305 solder paste in many surface mount assembly applications. This approach offers reduced printed circuit board warpage as a result of significantly reduced reflow temperatures and improved solder joint reliability from epoxy resin reinforcement[10][11]. In this study, we investigated the influence of JRP epoxy resin coverage on solder joint strength and reliability Ball Joint Shear Test Joint strength and the fracture morphology were investigated employing a ball shear test to compare JRP solder joints with SAC305 solder joints. Five test specimens were created for each test condition. The solder paste was printed onto an organic surface protection (OSP) treated Cu pad on the PCB test vehicle. A 300 micron diameter SAC305 solder ball was then placed on the pad and the assembly was reflowed. For printing the JRP samples, various stencil thickness and aperture size were used to provide a range of epoxy resin coverage heights around the reflowed solder joint. For the SAC305 samples, a single printing condition was used. The JRP assemblies were reflowed using a one step, 160 C peak reflow profile. The SAC305 assemblies were reflowed using a conventional, two step, 240 C peak reflow profile, as shown in Figure 2. Paste printing and solder reflow conditions of JRP and SAC305 specimens are listed in Table 1. After reflow, each solder joint in the evaluation was subjected to a ball joint shear test. The shear rate was 0.1mm/sec and the vertical gap or stand-off distance between shear test tool and the Cu pad was 50 microns, as shown in Figure 3. Additionally, after the ball shear test, we examined the morphology of the solder joint fracture by cross-sectional analysis. 1399

3 BGA Solder Joint Strength Test Package level joint strength and fracture morphology were investigated comparing JRP and SAC305 solder pastes using 0.5mm pitch, 400 I/O BGAs assembled onto a PCB and subjected to a solder joint strength test. The SAC305 BGA solder ball height was 116 μm and 0.25 mm. Samples were prepared by forming 9 solder joints at the 4 corners of the BGA for a total of 36 solder joints per package as shown in Figure 4. Each solder paste was printed on the 36 OSP coated copper pads on the PCB test vehicle. Then, the BGA was oriented and placed on the printed paste. After BGA placement, the assemblies were reflowed. As with the ball shear test sample preparation, various stencil thickness and aperture size were used to build the JRP specimens, while a single condition was used for the SAC305 specimen. The range of JRP stencil conditions were expected to yield varying epoxy resin solder joint coverage heights after the reflow process. Printing and reflow conditions of the BGA solder joint strength test samples are listed in Table 1. After reflow, the test assembly was subjected to a BGA solder joint strength test. The shear rate was 0.1mm/sec and the vertical gap between the shear test tool and the PCB Cu pad was 50 microns, as shown in Figure 5. Additionally, we examined the solder joint fracture morphology by crosssectional analysis. Package Level Temperature Cycling and Drop Shock Tests For temperature cycle test, a test matrix combining two surface mount packages and three solder pastes was used to evaluate the temperature cycling and reliability of these materials. The packages consisted of a 0.65 mm pitch LGA and a 0.5 mm pitch BGA. The three solder pastes used for this evaluation were SAC305, SnBiAg and SnBi JRP. Each paste was printed using the same conditions and the packages were assembled on the test PCBs. Five samples were created for each package and paste combination. The samples were used for the thermal cycling test. The temperature cycle condition was -40 C and 85 C with a 30min soak at each minimum and maximum temperature. For drop shock test, 0.4 mm pitch BGA and three solder pastes were used to evaluate mechanical reliability of these materials. The drop shock test were performed in accordance with JESD22-B11 (Board Level Drop Test Method of Components for Handheld Electronic Products) which direct test conditions of 1500G, 0.5ms, with a half-sine pulse. A BGA with daisy chain design was mounted on each PCB test vehicle. The location of BGA in the center of the test board and an image of the testing apparatus are shown in Figure 6. The BGA was monitored electrically for solder joint failures during the drop shock test. 1400

4 III. RESULTS AND DISCUSSION Ball Joint Shear Test Results 1) Ball Joint Shear strength Figure 7 shows the results of the ball joint shear test. For JRP sample printing conditions #1, #2 and #3 the stencil aperture size remained constant and stencil thickness was altered. The results demonstrate that a thicker stencil yielded a higher ball shear value. In sample sets #3 and #5, the stencil aperture and thickness for the JRP paste and SAC305 were identical. Under these printing conditions, the shear strength of JRP was higher than that of SAC305. Solder paste print volume generally grows with increases in aperture size and stencil thickness. The shear strength of JRP rises with an increased print volume, and exceeds that of SAC305 when the print amount is equivalent to or higher than that of SAC305. 2) Analysis of fracture morphology Figure 8 illustrates the fractured solder joints after the shear test. In sample sets #1, #2, #3 and #4, all made with JRP, the fracture occurred through solder alloy matrix. In sample #6, made with SAC305, the fracture took the form of pad detachment. The epoxy resin in the JRP samples not only reinforced the solder joint, but also strengthened the bond between the copper pad and the PCB. The SAC305 samples, which did not contain resin, could not adequately reinforce the pad and prevent delamination. These results indicate that coverage of pads by the JRP resin reinforces the bond between the cupper pad and the underlying laminate. We also noted that as the amount of resin in the collar covering the JRP solder joint is increased, the resin fillet around the solder ball joint becomes larger, and the shear strength rises because the both the solder and the epoxy resin must fail before the solder joint fractures. BGA Solder Joint Strength Test Results 1) Solder joint strength Resin coverage heights of the BGA solder joint test assemblies under various conditions were measured and compared in terms of resin height ratio, as shown in Figure 9. The resin height ratio is defined as follows. Resin height ratio = Resin coverage height / Gap between PCB and BGA 100% Figure 10 shows the resin height ratio and the measurements of the BGA solder joint strength tests. The joint strength of JRP was the greatest in sample set #4 which had solder joints with a 100% resin height ratio. The other JRP sample sets (#1, #2 and #3) with less than 100% resin height ratio showed similar joint strength to that of the unreinforced SAC305 sample (#5.) 1401

5 2) Analysis of fracture morphology Figure 11 shows a typical cross sections of a JRP and SAC305 solder joints. Figure 12 shows fractured solder joints from the BGA solder joint strength test. The fractures always occurred at the interface between the solder and the BGA pad, irrespective of the solder paste composition (JRP or SAC305) or resin height ratio. As shown in Figure 11, for the JRP samples, an intermetallic joint is formed with the solder ball. The SnBiAg alloy contained in the JRP melts, diffuses, and forms a joint with the SAC solder ball. This joint forms even without melting the SAC solder on the BGA side of the joint. Therefore, the SAC and copper pad joint interface layer of the pad on the BGA side and the solder is the same for both JRP and SAC305 paste samples. This is why the JRP and SAC305 solder pastes specimens demonstrated the same fracture morphology. In both conditions, the shear stress was exerted on the BGA side of the solder joint structure. As described above, JRP exerts its greatest reinforcement effects on resin coverage at solder joint structures with 100% resin height ratio, achieving higher solder joint strength than that of SAC305. Reliability Test Results 1) Temperature cycle test Figure 13 shows the results of the temperature cycle test. Unsurprisingly, SnBiAg solder showed the greatest increase in resistance after 1000 cycles for both BGA and LGA packages. Figure 14 shows a cross section of 0.65 mm-pitch LGA. This assembly demonstrated the greatest increase in resistance value. After 1000 cycles, cracks were discovered in the SAC305 solder joints as well as in SnBiAg solder joints. No cracks were found in the solder joints formed with JRP. The JRP solder joints had the resin collar around them, and a resin height ratio of 100%. For 0.5 mm BGA samples, cracks were found in SAC305 solder joints even though resistance values did not increase, No cracks were found in BGA JRP solder joints despite their resin height ratio of 40%. As indicated above, the temperature cycle test confirmed the beneficial effect of resin coverage around JRP solder joints. 2) Drop shock test Figure 15 shows the results of the drop shock testing. As anticipated, SnBiAg solder joint samples showed the lowest resistance to drop shock. The JRP samples showed drop shock resistance of 4.5-fold better performance than the 1402

6 SnBiAg solder paste alone due to the reinforcing effect of the resin coverage around the solder joints. The JRP resistance to drop shock was equivalent or higher than that of the SAC305 solder joint samples. Current status and future development JRP technology has been used commercially in products such as camera modules and mobile devices. Next, we will conduct product-level reliability tests for higher reliability products, such as PCs, tablets, and smartphones. JRP can be reworked after soldering. Historically, there tends to be a tradeoff between improving reinforcement performance and improving rewokability characteristics. We are therefore working on improving materials, starting with the design of the epoxy resin compound, to achieve simultaneously more reliable and more reworkable JRP materials. IV. CONCLUSIONS In this study, we investigated the influence of epoxy resin collars on solder joints created with JRP solder paste. Joint strength and reliability were compared to Tin-Bismuth-Silver solder alloy (SnBiAg) and SAC 305. For ball bump shear test, when the amount of resin that covers the JRP solder joint is increased, the resin fillet around the ball solder joint becomes larger, and the shear strength rises. The fracture must propagate through both the resin as well as solder. As a result, the shear strength of JRP exceeds that of SAC305 when the print volume is equivalent or greater. For the BGA solder joint strength test, the strength of interconnects made with JRP was highest in solder joints with a 100% resin height ratio. Below 100% resin height 1403

7 ratio, the joint strength of the JRP was nearly the same as SAC305. For temperature cycle test, in LGA solder joints after 1000 cycles, cracks were found in unreinforced SAC305 solder joints as well as in the unreinforced SiBiAg solder joints. The JRP solder joints had resin filling around them, and a resin height ratio of 100%. As a result, no cracks were found. As for BGA solder joints, cracks were also found in SAC305 solder joints where resistance values did not change, whereas no cracks were found in JRP solder joints despite their resin height ratio of 40%. JRP > SAC305>SnBiAg For drop shock test, JRP solder joint samples showed drop shock resistance 4.5-fold that of unreinforced SnBiAg as a result of the reinforcing effect of the resin collar around the solder joints. The resistance to drop shock was equivalent to or higher than the SAC305 solder joint samples. JRP > SAC305 >> SnBiAg As demonstrated above, the strength and durability of resin reinforced solder joints created with JRP technology is better than joints created with conventional SAC 305 solder paste. REFERENCES [1] D.Amir, R. Aspandir, Buttars, W.W.Chin, and P.Gill, Head-on- Pillow SMT failure Modes, Proceedings of SMTA International Conference, [2] L.Kondrachova, S.Aravamudhan, R.Sidhu, D.Amir, and R.Aspandir, Fundamentals of the Non-Wet Open BGA Solder Joint Defect formation, Proceedings of the international Conference on Soldering and Reliability(ICSR), [3] Amir. S.Walwadkar, SAravamudhan, and L.May, The Challenge of Non Wet Open BGA Solder Defect, Proceedings of SMTA international Conference, [4] R.Aspendir, K.Byrd, Kok Kwan Tang, L. Campbell and S. Mokler Investigation of Low Temperature Solders to Reduce Reflow Temperature, Improve SMT Yields and Realize Energy Savings, Proceedings of the 2015 APEX Conference, February [5] S.Mokler, R.Aspendiar, K.Byrd, O.Chen, S.Walwadkar, K.K.Tang, M.Renavikar and S.Sane. The application of Bi-based solders for Low Temperature Reflow to Reduce Cost, while Improving SMT Yields in Client Computeing Systems, Proceedings of the 2016 SMTA international Conference, 2016, Chicago, IL. [6] M. Holtzer, Low Temperature SMT Process Conditions, Proceedings of the 2013 IPC APEX Conference, 2013, San Diego, CA. [7] K.Okamoto, K.Nomura, S.Doi, T. Akamatsu, S. Sakuyama, and K. Uenishi, Effect of Sb and Zn addition on Impact Resistance Improvement of Sn-Bi Solder Joints, IMAPS2014,September,2013, Orlando,FL. [8] Omid Mokhtari, Hiroshi Nishikawa, Effect of Minor Alloying Additive on the Shear Strength of Sn-58Bi Solder Joint, IMAPS2013,September 2013, Orlando, FL. [9] Morgana Ribas, Sujatha Chegudi, Anil Kumar, Sutapa Mukherjee, Siuli Sarkar, Ranjit Pandher, Rahul Raut, Bawa Singh, Low temperature Alloy Development for Electronics Assembly-Part II, SMTA International Conference, 2013, Fort Worth, TX. [10] Olivia.H.Chen, A.Molina R, R. Aspandair, K.Byrd, S.Mokler and K.K. Tang, Mechanical Shock and Drop Reliability Evaluation of the BGA Solder Joint Stank-Ups formed by Reflow Soldering SAC Solder Ball BGAs with BiSnAg and Resin-reinforced Solder Pastes, Proceedings of SMTA International Conference, 2015, Rosemont,IL. [11] Olivia.H. Chen, James Gao, Tim C.C.Pan, Kok Kwan Tang, Raiyo Aspandiar, Kevin Byrd, Bite Zhou, Scott Moker, and Al Molina, Solder Joint Reliability on Mixed SAC-BiSn Ball Grid Array Solder joints Formed with Resin Reinforced Bi-Sn Metallugy Solder Pastes, Proceedings of SMTA Intenational, September 2016, Rosemont, IL 1404