Interfacial Properties of Sn-Cu-xCr Alloy using Laser-Assisted Bonding

Shin-Il Kim; Dong-Yurl Yu; Yun-Chan Kim; Junhyuk Son; Dongjin Byun; Junghwan Bang

doi:10.5781/JWJ.2023.41.4.8

Abstract

Conventional hot air reflow soldering is one of major bonding technology when bonding is proceeded on automotive electronics applications. However, here we focus on modifying the bonding process and composition of solder to enhance mechanical and thermal properties. This study experimentally investigated the mechanism of how laser-assisted bonding (LAB) form a small size of initial intermetallic compounds (IMC), and 0.2 wt% Cr element was added to Sn-0.7Cu eutectic solder to suppress the IMC growth. As a result, Sn-0.7Cu and Sn-0.7Cu-0.2Cr initial grain refinement was achieved by LAB. Isothermal aging was conducted to observe the mechanical and thermal properties depending on the IMC growth under 100, 125, 150 °C respectively. Sn-0.7Cu-0.2Cr tended to inhibit the growth of IMC at solder matrix and interface compared to Sn-0.7Cu throughout all aging temperature and time. The shear test was proceeded and Sn-0.7Cu-0.2Cr shear force was achieved 18% higher than Sn-0.7Cu. Even after isothermal aging, Sn-0.7Cu-0.2Cr showed lower reduction rate of shear force compared to Sn-0.7Cu.

Key words: Laser-assisted bonding (LAB), Intermetallic compounds, Lead-free solder, Shear test, Isothermal aging test

1. Introduction

Automotive electronics are exposed to harsh environments, such as high temperature, low temperature, and vibration, compared to general home appliances. Since electronics around the engine are exposed to a high- temperature environment of more than 125 °C during driving, the high reliability of solder joints around the engine room is required to ensure the safety of the driver as well as the function of electrical connection¹⁾. The representative Sn-Pb solders of the past have excellent thermal and mechanical properties²⁾. The use of Pb solders that is harmful to the human body, however, has been restricted due to environmental regulations, such as the End of Life Vehicles (ELVs) implemented since 2007³⁾. Various lead-free solder alloys have been developed to replace such Pb solders, and Sn-Ag-Cu and Sn-Cu solders have been typically used. A representative solder among Sn-Ag-Cu solders is Sn-3.0Ag-0.5Cu (SAC305), and Sn-Cu solders include Sn-0.7Cu⁴^,⁵⁾. The SAC305 solder shows a reduction in reliability at high temperature due to creep characteristics and coarsening of Ag₃Sn intermetallic compounds (IMCs)⁶⁾. The Sn- 0.7Cu has relatively low wettability and easily growing IMC properties compared to other solder and it causes concerns about brittle property and degradation of joint strength. To address these problems, studies on improving high-temperature reliability by adding trace elements, oxides, and polymers have been actively conducted⁷⁾.

Conventional hot air reflow bonding requires the process time of approximately five to seven minutes for heating, maintenance, and cooling. Since the maximum temperature is higher than the melting point of the solder, the substrate and package are inevitably subjected to thermal stress over an extended period of time. This may cause the warpage of the substrate, package and future problems with high-temperature reliability due to the elongation and contraction of joints caused by the difference in the coefficient of thermal expansion (CTE) between heterogeneous materials⁸⁾. To solve above problems, research on laser-assisted bonding (LAB) has been actively conducted. LAB can minimize the thermal stress of the substrate and package through local heating of joints. It also has a benefit of high-temperature reliability by inhibiting the growth of IMCs in solder and joints through short-term bonding⁹^-¹¹⁾.

As such, in this study, the properties of the Sn-0.7Cu solder and the solder containing 0.2 wt% Cr were evaluated using LAB. The growth and mechanical properties of IMCs at joints according to isothermal aging were then compared and analyzed.

2. Experimental Method

2.1 Test coupon preparation

In this study, LAB was applied to solders with trace elements, and the mechanical properties of joints and the behavior of IMCs according to isothermal aging were compared and evaluated. The Sn-0.7Cu solder and the Sn-0.7Cu-0.2Cr solder containing Cr were used in the experiment, and bonding was performed using 300 ㎛ solder balls. The size of the substrate opening was 230 ㎛, and a printed circuit board (PCB) with organic solderability preservative (OSP) surface treatment on a Cu pad was used. 1,064 nm continuous wave (CW) fiber laser (INYA 1000W, INLASER Co., Ltd.) was used in the experiment. In previous study which is using hot air reflow method, the melting points of the Sn-0.7Cu, Sn-0.7Cu-0.2Cr solder were found to be 228 and 231 °C respectively, and a peak temperature of 260 °C was applied when the reflow soldering. In this experiment, lasers were irradiated at various powers. The experimental condition was set at 380W, which the peak temperature of 260 °C was detected by thermo- couple. When the laser power was set to 380W and irradiated for 1.5 seconds, bonding joint was not formed. However, it was observed that the bonding joint started forming from 2 seconds of irradiation. Therefore, the laser irradiation time was set to 2 seconds, to minimize the formation of IMCs. A laser beam size of 20×20 (mm²) were set. Fig. 1 shows the laser bonding conditions and temperature profile.

Fig. 1

The schematics of (a) LAB condition and (b) setting temperature profile during LAB

2.2 Analysis of joint properties

2.2.1 Isothermal aging test

To examine the internal change of the solders and the growth of IMCs at interface according to aging, the isothermal aging test was conducted for up to 500 hours under high-temperature conditions of 100, 125, and 150 °C. The low-speed shear test was then conducted to investigate mechanical properties due to the deterioration of the joint.

2.2.2 Microstructure and fracture surface analysis

The specimens were mounted with epoxy resin and hardener to observe changes in microstructure and IMCs under solder composition and isothermal aging conditions, and micro-polishing was performed using abrasive paper and alumina. To investigate the geometry of IMCs, etching was performed using an Sn etching solution (95 vol% C₂H₆O, 3 vol% HNO₃, 2 vol% HCl). As for the cross-sectional analysis, the thickness and size of the solder and IMCs according to the isothermal aging time were observed using a scanning electron microscope (SEM, Inspect F, FEI Co., USA). The size of IMCs was measured based on the ASTM E112 standard. In addition, an energy dispersive spectrometer (EDS) was used to analyze the generated IMCs.

Table 1

Process optimization data according to laser conditions

Real peak temperature (°C)	360 W	380 W	400 W
1.5 s	228	235	242
2.0 s	228	235	243
2.5 s	229	236	245

An electron probe micro analyzer (EPMA, SX-100) was used to observe how Cr is distributed and in what form it exists within and at the interface of the solder.

2.2.3 Low-speed shear test

To examine the mechanical strength of the joint, the shear test was conducted at a shear height of 50 ㎛ and a shear rate of 300 ㎛/s using a shear tester (Dage 4000, Nordson Co., UK). The test was conducted 15 times to derive the average value. Fig. 2 shows the schematic of the shear test. To examine the effect of the IMC growth on the fracture behavior, the fracture surface was analyzed using SEM.

Fig. 2

The schematic of shear test condition

3. Results and Discussion

The microstructures of the solders according to the isothermal aging temperature and time are shown in Fig. 3. The β-Sn and Cu₆Sn₅ regions were observed in the microstructures. For both solders, the size of the Cu₆Sn₅ IMC increased after 500 hours of isothermal aging. The fine Cu₆Sn₅ IMC generated in the early stage of bonding grew inside the solder due to long- term high-temperature aging, and its size tended to increase as the temperature increased. In addition, the coarsening of IMCs was more obvious in the Sn-0.7Cu solder compared to the Sn-0.7Cu-0.2Cr solder. This is in agreement with the result of previous studies that the addition of trace elements into the process solder inhibits the IMC growth by interfering with the movement of Sn and Cu atoms inside the solder¹²^-¹⁴⁾.

Fig. 3

Microstructure of Sn-0.7Cu-0.2Cr, Sn-0.7Cu solders at (a), (b) As reflow and after 500 hours aging test with (c), (d) 100 °C, (e), (f) 125 °C, (g), (h) 150 °C respectively

Fig. 4 shows the SEM images of the solder joint according to the isothermal aging temperature and time. Good joints are observed from both solders at the beginning, but the growth of IMCs in small quantities can be seen from the Sn-0.7Cu-0.2Cr solder. Since bonding was performed within a short period of time due to the nature of LAB, initial IMCs grew to a thickness of less than 1 ㎛. Under the 100 °C condition, Cu₆Sn₅ grew gradually in the Sn-0.7Cu solder as the isothermal aging time increased, and Cu₃Sn grew at 500 hours. In the Sn-0.7Cu-0.2Cr solder, however, only Cu₆Sn₅ grew for up to 500 hours. Under the 125 °C condition, Cu₃Sn began to be observed at 300 hours in the Sn-0.7Cu solder, but it was not observed in the Sn-0.7Cu-0.2Cr solder. Under the 150 °C condition, Cu₃Sn began to be formed from 100 hours in the Sn-0.7Cu solder, but it was observed at 300 hours in the Sn-0.7Cu-0.2Cr solder. In Fig. 4, the rapid change from Cu₆Sn₅ to the Cu₃Sn IMC can be seen due to the increase in the mutual diffusion rate of Sn and Cu atoms as the temperature increased. In addition, under the 100 °C condition, the Kirkendall void was not observed from both solders. Under the 125 and 150 °C conditions, however, the Kirkendall void was observed from 500 hours for the Sn-0.7Cu- 0.2Cr solder and 300 hours for the Sn-0.7Cu solder.

Fig. 4

SEM images of solder joint with aging test at (a) 100 °C, (b) 125 °C, (c) 150 °C

To explain the mechanism of interfacial IMC growth inhibition due to Cr addition , Fig. 6 presents a schematic of the joint interface. For the Sn-0.7Cu solder, it exhibits typical mutual diffusion of Sn and Cu, forming Cu6Sn5 IMCs at the interface. For Sn-0.7Cu-0.2Cr, although it forms Cu₆Sn₅ at the interface like Sn-0.7Cu, as confirmed in Fig. 5, the added Cr exists around the IMCs within the solder and at the interface in the form of Cr, CrSn₂, etc. According to prior literature, comparisons of the diffusion rates of Cu and Sn atoms in Sn-0.7Cu and Sn-0.7Cu-0.2Cr solders showed that the diffusion rate of Sn from inside the solder towards the substrate slightly increases, whereas the diffusion rate of Cu from the substrate into the solder greatly decreases¹²⁾. Therefore, under the same conditions, it is observed that adding Cr to the solder decreases the diffusion of Cu, which inhibits the growth of Cu6Sn5 IMCs at the interface. Additionally, this is believed to be advantageous for the joint’s thermal reliability as it inhibits the growth of Kirkendall voids, which grow in an interconnected form.

Fig. 5

(a) BSE image and EPMA element mapping of Sn-Cu-Cr composite solder. (a) BSE image, (b) overlayed, (c) Sn, (d) Cu, (e) Cr mapping images respectively

Fig. 6

The schematics of interdiffusion between Cu and Sn at the solder joint. (a) Sn-0.7Cu, (b) Sn-0.7Cu- 0.2Cr

Fig. 7 shows the total and Cu₃Sn IMC thickness graphs. It was observed that the IMC thickness of the Sn-0.7Cu-0.2Cr solder is lower than that of the Sn- 0.7Cu solder. The initial IMC thickness was found to be 0.50 ㎛ for the Sn-0.7Cu-0.2Cr solder and 0.63 ㎛ for the Sn-0.7Cu solder. After 500 hours, the IMC thickness of the Sn-0.7Cu-0.2Cr solder was 16.4% lower at 100 °C, 39.5% lower at 125 °C, and 39.6% lower at 150 °C compared to that of the Sn-0.7Cu solder. This confirmed that the IMC growth is inhibited in the Sn-0.7Cu- 0.2Cr solder compared to the Sn-0.7Cu solder¹⁷^,¹⁸⁾.

Fig. 7

Total and Cu₃Sn IMC thickness graph with aging test at (a), (b) 100 °C, (c), (d) 125 °C, (e), (f) 150 °C

Fig. 8 compares the grain size of IMCs according to the isothermal aging temperature and time. At the beginning, the size of IMCs was small for both the Sn-0.7Cu-0.2Cr and Sn-0.7Cu solders. This appears to be due to the nature of LAB in which bonding is performed instantaneously at high temperature and it is rapidly cooled at room temperature. Afterwards, the IMC growth according to the isothermal aging temperature and time could be confirmed, and the IMC growth was smaller in the Sn-0.7Cu-0.2Cr solder compared to the Sn-0.7Cu solder. According to a prior study, if a small grain forms, more diffusion paths with small diameters and high grain boundaries are produced, which leads to rapid IMC growth¹⁶^,¹⁹⁾. When the Sn-0.7Cu and Sn-0.7Cu-0.2Cr solders were compared based on the study, it was found that the IMC growth was inhibited in the Sn-0.7Cu-0.2Cr solder even though its initial IMC size was smaller. It is judged that the difference in IMC growth was caused by the dispersion of Cr and Cr₂Sn in the grain boundaries of IMCs.

Fig. 8

Top-view images of Sn-0.7Cu-0.2Cr and Sn-0.7Cu solders at (a), (b) As reflow and after 500 hours aging test with (c), (d) 100 °C, (e), (f) 125 °C, (g), (h) 150 °C respectively

Fig. 9 shows the shear force and shear force decreasing rate graphs of joints according to the isothermal aging temperature and time. At the beginning, the shear force of the Sn-0.7Cu-0.2Cr solder was approximately 18% higher than that of the Sn-0.7Cu solder. This appears to be because the added Cr element caused the grain refinement of the solder structure, thereby increasing the grain boundary and making dislocation movement difficult. For both solders, the shear force decreased as the isothermal aging temperature and time increased. The shear force of the Sn-0.7Cu solder continued to decrease and its shear force decreasing rate increased. The shear force and decreasing rate of the Sn-0.7Cu-0.2Cr solder, however, showed a tendency to be maintained after 100 hours. This appears to be because the Cr atoms inhibited the coarsening of IMCs inside the solder¹⁷^,¹⁸⁾.

Fig. 9

Shear force & decreasing rate of shear force under LAB with aging time at (a, b) 100 °C, (c, d) 125 °C, (e, f) 150 °C

Fig. 10 shows the fracture surface after the shear test under each condition. After the shear test, the internal fracture of the solder was observed under three temperature conditions (100, 125, and 150 °C). In the shear test, fracture occurred in the part with the lowest strength. In particular, the results were in agreement with a previous study that reported that ductile fracture occurs in the low-speed shear test because the shear stress inside the solder is relatively low⁵⁾.

Fig. 10

Fracture surface images of Sn-0.7Cu-0.2Cr and Sn-0.7Cu solders at (a), (b) As reflow and after 500 hours aging test with (c), (d) 100 °C, (e), (f) 125 °C, (g), (h) 150 °C respectively

4. Conclusions

In this study, solders containing trace elements were bonded using laser-assisted bonding (LAB). The effects of LAB and the addition of trace elements on the thermal reliability of the joint were observed.

1) When the microstructures of the solders were analyzed, both solders exhibited small grain sizes at the beginning. As isothermal aging progressed, coarsening of the grains and intermetallic compounds (IMCs) inside the solders was observed. In the solder containing Cr, however, the growth of grains and IMCs was inhibited.
2) When the IMC thickness was compared, the IMC thickness at the beginning and after isothermal aging was lower in the Sn-0.7Cu-0.2Cr solder compared to the Sn-0.7Cu solder. The Cu₃Sn IMC and Kirkendall void were also less observed in the Sn-0.7Cu-0.2Cr solder.
3) When the IMC grain size was compared, the initial grain size of the Sn-0.7Cu-0.2Cr solder was found to be smaller. The IMCs and grains inside the solder after isothermal aging were also found to be smaller compared to the Sn-0.7Cu solder.
4) The bonding strength measurement results revealed that the initial shear force of the Sn-0.7Cu-0.2Cr solder was approximately 18% higher than that of the Sn- 0.7Cu solder, and that the decreasing rate of the Sn- 0.7Cu-0.2Cr solder was also lower even after isothermal aging.