Effect of indium on microstructure, phase stability, anisotropic lattice behavior and mechanical performance of Sn-58Bi solder alloy

Four different Sn–Bi–In solder alloy compositions were prepared by varying the In content while maintaining Bi at a constant of 58 wt.%. The first composition contains 42 wt.% of Sn with no In, serving as a reference of Sn-58Bi eutectic alloy. The following alloys contain 4, 8, and 12 wt.% of In, with reductions in Sn content to 38, 34, and 30 wt.%, respectively. For each composition, the Sn, Bi, and In were accurately weighed according to their composition and then were put in a graphite crucible that was coated with boron nitride. The crucible was then put in an electric furnace where the temperature was set to 350 ºC to melt the raw materials. Prior to 15 min of heating, the melted alloy mixtures were stirred lightly to impose homogeneity. Upon the removal of the dross, the molten alloys were then poured into a mold and were left to cool down at room temperature. All the solder alloys were prepared using the same methods to ensure comparability across all compositions.

The as-cast solder alloys were ground to ensure a flat surface and were then sent for an XRD analysis (model: Bruker D2 Phaser). The scan range was set from 20º to 90º of 2θ at a scan speed of 5º/min, with a step size of 0.02º, using CuKα radiation (λ = 1.54184 nm) at 10 mA and 30 kV. The phases were then identified by comparing the resultant pattern with the ICDD PDF-2 2003 by utilizing the PANalytical HighScore Plus software version 3.0.5. This laboratory-based X-ray diffraction (XRD) was conducted as a preliminary assessment prior to Synchrotron XRD analysis. It was performed to verify the quality of the sample preparation and to confirm the presence of expected phases. Additionally, the laboratory XRD data were also used to identify the important 2θ positions, allowing optimization of the scanning range for synchrotron measurements. This ensured that only the most relevant 2θ positions were selected, thereby reducing acquisition time without compromising data quality.

The XRD patterns at elevated temperature were obtained in situ at the Synchrotron Light Research Institute (SLRI), Thailand, using Beamline 1.1W. The measurement was conducted using synchrotron radiation with an energy of 12 keV where the wavelength was λ = 1.0332 Å. Samples of Sn-58Bi-xIn were analyzed at room temperature and at elevated temperatures which were 30, 70, 90, and 100 °C and were scanned at every stage of the temperature. An exposure time of 180 s was used for scanning at each temperature and the heating rate was controlled at 10 °C/min. The lattice parameters were then determined using PANalytical HighScore Plus software.

To further investigate the phase distribution of each element in the Sn-58Bi-xIn solder alloys, micro-X-ray fluorescence (XRF) analysis was employed. Synchrotron X-ray fluorescence experiments were carried out at beamline 7.2W, at the Synchrotron Light Research Institute (SLRI), without utilizing a monochromatic X-ray beam. The synchrotron micro-XRF end-station consists of four main components including an X-ray optical system, a visible-light microscope, a motorized sample holder with three degrees of freedom, and an energy-dispersive detection system. To achieve a micrometer-sized beam, a focusing system for X-rays is essentially needed. Here, a polycapillary full lens is employed to focus an X-ray beam to 30 × 30 µm² size. This polycapillary lens is installed on a motorized stage with four degrees of freedom. A CCD camera coupled with an objective lens is also used to specify and capture the measured area of a sample. Samples were placed at a 90° level between the X-ray and the CCD camera. A Vortex EM-650 silicon drift detector was used to collect the emitted fluorescence X-ray. A 0.03 mm scanning step with an exposure time of 30 s for each point was used. When the measurement is completed, a series of one-column XRF data files is generated. PyMca software was utilized to acquire an image showing the elemental distribution for analyzing purposes.

Solderability test is essential for evaluating a solder alloy’s ability to form strong, uniform and reliable joints. It evaluates how effectively the molten solder wets the surface, which reflects the interaction between the solder and the substrate. The solderability tests were evaluated using the Gen3 MUST System 3 based on the globule method, following the guidelines of the Japanese Industrial Standard (JIS) which is JIS Z 3198–4. Firstly, the 10 mm × 3 mm copper strips of 0.3 mm thickness were cleaned by dipping into an acetone bath, then in an acid solution (Acid Hydrochloric + Deionized Water) followed by rinsing with DI water to remove the oxides on the surfaces of the copper strips. No-clean rosin-based fluxes were applied onto the strip and the strips were then clipped to the holder (Type 6 Clip) of the solderability tester. The strips were then dipped at a 5 mm depth into a pot filled with molten Sn (280 ºC) for 20 s and were then withdrawn at a speed of 30 mm/s. These steps of coating the copper strip with pure Sn were done to simulate the HASL process.

All the Sn coated copper strips were then cleaned again with the acid solution to remove oxides or any contaminants on the surfaces before the solderability test. The tests were conducted on the same machine (Gen3) which was set into micro-wetting balance test mode, as shown in Fig. 1. Fluxes were again applied to the Sn coated copper strips before attaching the strips to the holder (Type 9 Clip). The clip will hold firmly the strips at an angle of 45º while the 4 mm globule block is elevating until it barely touches the Sn coated copper strips. During this process, the immersion and removal speed was set to 1.0 mm/s with the immersion time being 15 s at the immersion depth of 0.2 mm. The machine will start recording data by measuring the force on the strip as a function of time and the result of the solderability test was then evaluated by analyzing the result of wetting time and the maximum force.

Full size image

To investigate the interfacial reactions and intermetallic compound (IMC) formation at the solder joint, the as-cast sample was placed on a Cu substrate and was then reflowed in a reflow oven (F4N Reflow Oven), using the standard reflow profile of Sn-58Bi solder. Then, the samples were cold mounted, cross-sectioned ground, and polished by following the same steps as stated previously. Observations were conducted using SEM and from the micrographs, the thickness of the IMC layer was calculated using ImageJ analysis software.

Tensile test is conducted to evaluate the mechanical performance of the solder alloys, particularly their tensile strength and elongation, which are critical for assessing their reliability in electronic packaging applications. For tensile test, the mold used is as in Fig. 2. The mold which is made of tool steel where the final geometry of the tensile bars meets the standard requirements of ASTM E8/E8M. Details of the dimension requirement are as shown in Table 1. The tensile bars were subjected to tensile test using a 50 kN Universal Testing Machine (UTM) Instron model no. 5569 with a cross-head speed of 5 mm/min, gauge length of 30 mm, and the strain rate is 0.00278 s⁻¹.

Full size image

The thermal properties of the solder alloys were measured using differential scanning calorimetry (DSC, NETZSCH DSC 214 Polyma). Each sample with the weight of 10 mg was used for the DSC analysis. All of the samples were heated from 30 ºC to 200 ºC with the heating rate controlled at 10 ºC/min. Upon reaching the maximum temperature, samples were cooled down from 200 ºC to 30 ºC, at the same rate as the heating rate, 10 ºC/min. Both processes of heating and cooling were performed in the protective atmosphere of nitrogen where every sample was analyzed for 2 times.

The Sn–Bi–In ternary phase diagram simulated using Thermo-Calc software illustrates the phase stability and transitions over a range of Indium concentrations from 0 to 20 wt.% and temperatures up to 160°C, as presented in Fig. 3(a). At 0 wt.% In, the Sn–58Bi eutectic alloy exhibits a sharp melting point near 138 °C, consisting of β-Sn and Bi phases. With the addition of 4-wt.% In, the alloy remains in the β-Sn + Bi region, with a slight reduction in melting temperature and a relatively narrow solidus–liquidus gap, indicating minimal disruption to the original eutectic structure. At 8-wt.% In, the melting temperature further decreases, and a wider melting range develops, suggesting changes in phase stability and solidification behavior. At 12 wt.% of Indium, the melting temperature continues to decrease, and the melting range continues to broaden. These shifts in thermal behavior indicate that increasing Indium content leads to lower melting temperatures, which are advantageous for low-temperature soldering applications where thermal sensitivity is a concern.

Full size image

The simulated phase volume fractions as a function of temperature for Sn–58Bi–xIn solder alloys with 0-, 4-, 8-, and 12-wt.% Indium are shown in Fig. 3(b,c,d,e). The diagrams show the relative proportion of phases including β-Sn, Bi, BiIn, and γ-InSn₄ across a temperature range relevant to solder applications. As in Fig. 3(b), below 10 °C the β-Sn phase may undergo transformation into α-Sn. These simulations were based on the CALPHAD approach, providing an overview of the solid-state phase constitution and its evolution with temperature and composition. The results offer insight into the phase stability and distribution associated with different Indium contents in the ternary system.

Figure 4 shows the microstructures of Sn–Bi–In solders at various compositions with the corresponding EDX analysis. These results indicate that the microstructures are composed of dark BiIn intermetallic phase, white Bi phase, and black β-Sn phase, consistent with the results of a previous study [9, 12‐18]. The black region observed in the microstructure images corresponds to a Sn-rich matrix where it was found that an amount of Bi (14.49 wt.%) and In (9.66 wt.%) elements were dissolved in the β-Sn phase as depicted in Fig. 4(e). The white region primarily consisted of the Bi element with a minor incorporation of Sn (9.89 wt.%) element while the gray region was concluded to be the BiIn phase as shown in Fig. 4(f) and Fig. 4(g), respectively. These findings were consistent with previous studies by other researchers [12, 13, 19]. XRD analysis further confirmed the existence of all these phases as shown in Fig. 6. In the eutectic Sn-58Bi alloy, the microstructure exhibits a fine lamellar arrangement of alternating β-Sn and Bi phases as shown in Fig. 4(a), which is typical for solder at eutectic composition. As shown in Fig. 3(b), the volume fraction plot confirms that only β-Sn and Bi phases are present, and the phase fractions are nearly equal up to the melting point. Adding 4 wt.% of Indium showed that the microstructure remains largely similar, with no evidence of new intermetallic phase formation, as shown in Fig. 4(b). This is consistent with the ternary phase diagram presented in Fig. 3(a) where it showed that there was no formation of BiIn at this Indium concentration. The volume fraction plot in Fig. 3(c) further confirms the co-existence of only β-Sn and Bi phases at temperatures below 130 °C, suggesting that the addition of 4-wt.% Indium does not significantly modify the phase constitution of the Sn-58Bi solder alloy. Figure 4(c) shows that with an 8 wt.% addition of Indium, a notable change in microstructure occurred, where a third phase that has been identified as BiIn is clearly observed alongside the existing β-Sn and Bi phases. This observation is strongly supported by the ternary phase diagram, which demonstrates the appearance of BiIn at Indium contents above 7 wt.%. The corresponding volume fraction plot in Fig. 3(d) displays that BiIn exists as a stable phase at room temperature but begins to dissolve as the temperature increases. The presence of BiIn has altered the microstructure by increasing the interlamellar spacing and reducing the uniformity of the fine eutectic structure that is commonly observed in Sn-58Bi solder alloy. A further increase of Indium content to 12 wt.% showed that the BiIn phase becomes more pronounced and widely distributed in the microstructure, as shown in Fig. 4(d). It is also observed that the Bi-rich regions become coarser and more isolated. The simulation result of the volume fraction shown in Fig. 3(e) confirms that BiIn comprises a larger fraction of the solid phase at room temperature. However, the BiIn phase is demonstrated to dissolve at around 70 °C, in agreement with the simulated ternary phase diagram. The reduction of the Bi phase volume fraction with increasing Indium content, as shown in Fig. 3(d,e), suggests that a portion of Bi is consumed to form BiIn. Consequently, the residual Bi phase becomes coarser and less continuous, which may contribute to altering the solder alloy’s mechanical performance. In summary, with more Indium added, the Bi phase coarsened and the interlamellar spacing became thicker.

Full size image

However, at lower temperature, there existed an intermetallic phase of γ-InSn₄, as depicted from the phase diagram. It is believed that the phase diagram generated using Thermo-Calc software represents equilibrium conditions, which may not always be achieved in practical processing [8, 16]. Furthermore, the formation of γ-InSn₄ may require fast cooling rates. In this research work, the cooling rate was around 40–50 °C; it could have suppressed the nucleation and growth of γ-InSn₄. The occurrence could also contribute to In having a higher affinity for Bi, so instead of forming γ-InSn₄, In could have reacted with Bi, thus forming InBi phase [7, 10].

The results of micro-XRF mapping analysis for Sn-58Bi solder alloy are as shown in Fig. 5(a–b). The image mapping represents the distribution of the detected elements which are Bi and Sn. The micro-XRF elemental mapping of the Sn-58Bi lead-free solder alloy reveals a distinct phase segregation between Sn and Bi, consistent with the expected eutectic microstructure of the alloy. The Sn map reveals a moderately dispersed pattern, suggesting that Sn is distributed throughout the region without forming a continuous phase. On the other hand, the Bi map shows that the Bi distribution appears more interconnected and concentrated in certain regions, confirming the presence of Bi-rich phases distributed between the Sn-rich phases. This shows that during solidification, Bi and Sn segregated into distinct regions. This phase separation is a well-known characteristic of Sn-58Bi eutectic alloys, where the Sn-rich phase forms dendritic structures, while the Bi-rich phase solidifies in the spaces between the dendrite regions. The results also align with the results of the microstructure for Sn-58Bi alloys as shown in Fig. 4(a), where Bi-rich regions typically form due to phase segregation during solidification, reinforcing the fact that Sn and Bi do not mix homogeneously due to their limited solubility in each other.

Full size image

Figure 5(c–e) shows the results of synchrotron micro-XRF mapping analysis for Sn-58Bi-12In solder alloy. The results show elemental distribution maps illustrating the presence of Sn, Bi, and In elements, revealing distinct regions of segregation. These results are aligned with the microstructural result as shown in Fig. 4(d), where the microstructure is composed of three major phases which are the β-Sn phase, Bi-rich phase, and the BiIn intermetallic phase. The micro-XRF analysis of Sn indicates that it is distributed throughout the matrix, while the Bi distribution from micro-XRF is more isolated, indicating the presence of Bi-rich phases in the microstructure and the In distribution suggests its presence in the BiIn intermetallic phase.

The segregation of Bi in the microstructure indicates limited solubility in Sn, leading to phase separation. This phenomenon is commonly observed in Sn–Bi-based solders, where Bi tends to segregate at grain boundaries and within the matrix. The micro-XRF results confirm that Bi does not form a uniform distribution, which further supports the presence of Bi phases within the microstructure. This segregation can negatively impact the mechanical integrity of the solder, as localized Bi-rich regions may act as crack initiation sites under mechanical stress. The In distribution in the micro-XRF results demonstrates a unique interaction with both Sn and Bi. The co-existence of In and Bi in localized regions suggests that In has a strong affinity for Bi, leading to the formation of BiIn intermetallic compounds. This can be seen from the microstructural result with the presence of the BiIn intermetallic phase, which appears as dark regions. This is significant because the presence of BiIn intermetallic may alter the mechanical properties of the solder by improving ductility and reducing the brittle nature of Sn–Bi alloys. In summary, the micro-XRF results validate the microstructural findings by confirming the elemental composition and distribution within the Sn-58Bi-12In solder alloy.

Figure 6 shows the results of XRD analysis at different compositions. There are three phases identified which are rhombohedral Bi (PDF number: 01–085–1329, space group R-3m, a = 0.4546 nm, c = 1.1862 nm), tetragonal β-Sn (PDF number: 01–089–2958, space group I41/amd, a = 0.58327 nm, c = 0.31825 nm) and tetragonal BiIn (PDF number: 01–085-0344, space group P/nmm, a = 0.5024 nm, c = 0.4787 nm). The XRD analysis results are well aligned with the results of the microstructures as shown in Fig. 4. It is evident that the addition of In to Sn-58Bi solder alloys leads to the formation of the BiIn phase, which is the only new phase observed compared to the original Sn-58Bi alloy. There are no other new intermetallic compounds that could be found in the microstructures of the investigated samples at different ratios of In element added.

Full size image

The in situ synchrotron XRD patterns show the phase evolution of Sn-58Bi and Sn-58Bi-12In solder alloys at temperatures ranging from 30 °C to 100 °C, as can be seen from Fig. 7. In the Sn-58Bi alloy, only Bi and Sn phases are detected throughout the temperature range, with no significant phase changes. The peak intensities remain relatively consistent, indicating thermal stability and no formation of new phases up to 100 °C. However, adding In to Sn-58Bi leads to the formation of a new intermetallic phase, BiIn, which is clearly present at room temperature (30 °C), as presented in Fig. 7(b). As the temperature increases to 70 °C, the intensity of the BiIn peaks begins to decrease, and by 90 °C and 100 °C, the BiIn phase is no longer detectable. This phenomenon is well aligned with the Sn–Bi–In ternary phase diagram simulated by Thermo-Calc as shown in Fig. 3(a), which indicates that the BiIn phase only exists up to approximately 70 °C under equilibrium conditions. The disappearance of BiIn at higher temperatures suggests that it dissolves as the alloy approaches and surpasses this stability limit. The XRD patterns also reveal the existence of Bi₂O₃ in the Sn-58Bi-12In solder alloy, with diffraction peaks becoming noticeable starting at 70 °C and becoming more apparent at 90 °C and 100 °C. This indicates that the oxidation of Bi becomes more pronounced at elevated temperatures. The formation of Bi₂O₃ is likely due to increased surface reactivity and oxygen diffusion at higher temperatures, promoting the oxidation of Bi in the presence of ambient air.

Full size image

Previously, Zhang et al. reported that Bi has a strong affinity for In, leading to the formation of the BiIn phase due to their large negative mixing enthalpy [17]. The BiIn phase was found to coexist alongside Sn-rich and Bi-rich phases, which can be seen from the microstructural images as shown in Fig. 4(d). Bi is a key element in both the Bi phase and the BiIn intermetallic phase. However, only the Bi phase remains stable across the full range of the temperature that has been studied. Therefore, it serves as the most reliable reference for structural analysis, where unlike the BiIn phase, which dissolves at elevated temperatures.

The Bi phase has a rhombohedral crystal structure, in which the a and b axes are equal in length (a = b). As the temperature increases, the lattice parameters along these axes in the Sn-58Bi solder alloy show a decreasing trend, as illustrated in Fig. 8(a). This reduction in lattice parameter is attributed to thermal contraction of the Bi phase, which reflects its anisotropic expansion behavior. This characteristic is commonly observed in layered crystal structures such as rhombohedral Bi. However, in Sn-58Bi-12In, the a-axis lattice parameter remains nearly constant from 30 °C to 100 °C. This behavior suggests that In addition stabilizes the a-axis lattice parameter, possibly using interstitial or substitutional sites that resist lattice contraction. Due to In’s larger atomic radius (167 pm) compared to Bismuth (156 pm), In atoms introduce lattice distortion, which helps mitigate thermal effects and reduce compressive strain along the a and b axes [20].

Full size image

In contrast, the c-axis lattice parameter for Sn-58Bi shows a continuous expansion with temperature as depicted in Fig. 8(b), indicating normal thermal expansion. For Sn-58Bi-12In, the c-axis lattice parameter decreases significantly, suggesting that In addition alters the anisotropic thermal expansion behavior of Bi. The contrast between the trends of the a- and c-lattice parameters in Sn-58Bi-12In suggests that In induces anisotropic lattice distortion, likely due to interactions between In and Bi atoms, which result in greater compression along the c-axis compared to the a-axis.

The unit cell volume, derived from the lattice parameters, decreases sharply in Sn-58Bi with temperature due to simultaneous contraction in the a-axis and slight expansion in the c-axis as shown in Fig. 8(c). In Sn-58Bi-12In, the volume remains more stable across the temperature range, showing just a slight decrease. This relative stability in Sn-58Bi-12In suggests that In helps to retain structural integrity at elevated temperatures, likely by hindering thermally induced distortions. This stabilization effect is important for solder applications, where thermal cycling can induce fatigue and microstructural damage [21].

Thermal properties of the Sn-58Bi-xIn solder alloys, including melting temperature, pasty range, and undercooling, were evaluated using DSC, with values calculated from the obtained heating and cooling curves. The DSC results for the Sn-58Bi and Sn-Bi-xIn solders at different compositions are shown in Fig. 9. From the heating curves shown in Fig. 9, a minor endothermic peak around 80 °C can be found on each curve with In additions. This peak is attributed to the solid-state phase transformation of the BiIn phase. The reaction involves the Bi and Sn phases along with the BiIn compound, where BiIn eventually decomposes, resulting in the presence of only Bi and Sn phases according to vertical sections in the Bi–Sn–In system as shown in Fig. 3. With increasing temperature, the BiIn phase begins to disappear and Indium atoms dissolve into the Bi and β-Sn phases. This ultimately leads to an enthalpy change. A notable endothermic peak at around 102 °C can be found for each of the compositions, and the reaction of Sn-Bi-In solders can be described as the transformation of the Bi and Sn solid phases into a mixture of liquid, Bi, and Sn phases during heating. This reaction indicates that the solders start to melt at these temperatures, which can be viewed as the onset melting temperatures (solidus temperatures), which are desired for low-temperature soldering. Moreover, all these temperatures are significantly lower than the Sn-Bi eutectic point due to the addition of Indium. There is another major peak at around 116 °C to 128 °C on each DSC curve, which is attributed to the melting of the Sn phase in the solders during heating where the solders become fully melted. The DSC results obtained in this study are consistent with the findings of previous research on Sn-Bi-In solder alloys, demonstrating similar endothermic and exothermic peaks associated with phase transformations and melting behavior [12, 14, 15]. The solder alloys melting temperature (T_m) and the results for the calculations of the pasty range and its undercooling are as presented in Table 2. The melting temperature decreased progressively from 143 °C to 116 °C for Sn-58Bi and Sn-58Bi-12In, respectively. This indicates that In is effective in lowering the melting point, making the ternary alloys more suitable for low-temperature electronic packaging applications. The observed reduction in melting temperature and the presence of multiple thermal events align well with previously reported studies, confirming the influence of In addition on lowering the melting temperatures [12]. However, it is also observed that the undercooling and pasty range become widened when compared to the Sn-58Bi solder alloy. The occurrences are primarily attributed to the disruption of the eutectic composition and the formation of a ternary, non-eutectic system upon In addition. For a eutectic alloy such as Sn-58Bi, solidification occurs at a single temperature, resulting in a narrow undercooling and pasty range. However, the introduction of a third element such as In modifies the phase equilibrium and creates a composition that no longer solidifies isothermally. Instead, solidification occurs over a range of temperatures due to the sequential formation of multiple phases such as BiIn, β-Sn, and Bi which leads to a wider solidus–liquidus gap, resulting in an expanded undercooling and pasty range.

Full size image

The cooling curve of the Sn-Bi-4In alloy, as shown in Fig. 9, reveals several critical thermal events, indicated by distinct peaks as the temperature decreases. Classically, the temperatures recorded on cooling lie a few degrees lower than the temperatures measured on heating. The first peak at 117 °C signifies the onset of solidification, likely involving the transition of the liquid alloy to its first solid phase, which is Bi. A more prominent peak existed at 112 °C, which emerged from the first peak, during the cooling process. These apparent two peaks belong to the same thermal effect because the onset temperatures are pointed at around the same value. However, this occurrence does not exist in the other alloy. Following this, smaller peaks at 92 °C, 83 °C, and 68 °C reflect subsequent solid-state transformations which may involve the formation of secondary phases or further solidification of remaining liquid components. These transitions illustrate the complex cooling behavior of the Sn-Bi-xIn alloy, characterized by multiple phase changes as the temperature drops. For 8In and 12.0In alloys, the cooling curves are quite similar except that, for both alloys, a peak emerged at around 102 °C. All three smaller peaks are observed in all the alloys. However, it is clearly seen that the last peak, which is around 68 °C to 70 °C, has an increasing peak strength with increasing In addition. This peak signifies the BiIn phase solidification, which suggests that, with more In added, more BiIn phases are formed.

The effect of the In element on the solderability of Sn-58Bi solder alloy was also researched, where the wetting time and maximum wetting force of Sn-58Bi-xIn alloys are shown in Fig. 10. The wetting efficiency of a solder is commonly evaluated through parameters such as its wetting time and the maximum wetting force. Wetting time is commonly defined as the moment when the solder’s contact angle with the specimen reaches 90°, or alternatively, when the measured wetting force drops back to zero [22]. From Fig. 10(a), the wetting time of the solders with 4% and 8% In element is 1.463 s and 1.505 s respectively, where it is higher than the wetting time of the base solder Sn-58Bi, which is 1.374 s. When In is introduced into the Sn-58Bi solder matrix, the presence of In atoms can modify the liquid-state viscosity and even increase the surface tension, which consequently slows down the wetting process. Indium is an element that oxidizes easily. Even a small amount of In can cause surface oxidation to the liquid alloy, which can hinder the wetting process. However, for Sn-58Bi-12In, the wetting time is shorter than the base solder, which is 1.356 s, inferring that the wetting time is better. This may be due to the BiIn-rich phase contributing to enhanced diffusion of molten solder on the substrate. The presence of BiIn phases at this composition may modify the interface properties and promote better wetting kinetics, ultimately reducing the wetting time [8]. Specifically, the BiIn phase can reduce the liquid–vapor surface tension (γ_LV) and potentially lower the solid–liquid interface surface tension (γ_SL), both of which will contribute to a decrease in the contact angle as described by Young’s equation as in Eq. (1), where γ_SV is the solid–vapor surface tension [10]. The occurrence promotes enhanced spreading of the molten solder on the substrate and accelerates wetting kinetics, thereby reducing the overall wetting time.

Full size image

It is important for a solder to have good wettability, which means that obtaining a shorter wetting time for the solder is the most desirable. Good solderability is typically indicated when it can reach two-thirds of the maximum wetting force within 2 s [23], where it is shown that these criteria are met by all four solder alloys. From the results, it is worth noting that, the wetting times for 0, 4 and 8 wt.% of In addition were within a similar range but then slightly reduced with 12 wt.% of In. It is also revealed that Sn-58Bi-12In solder alloy exhibited the shortest wetting time among all compositions.

For the maximum force, the results are as presented in Fig. 10(b), where it is showing a decreasing trendline with increasing In element. The lowest maximum force is 1.816 mN (Sn-58Bi-12In) and the highest is 1.988 mN (Sn-58Bi), while Sn-58Bi-4In and Sn-58Bi-8In resulted in 1.970 mN and 1.915 mN respectively. All the solder alloys with In element have lower maximum force than the Sn-58Bi-based solder; however, the values remained within a similar range and were comparable to the based solder. This may be due to as more In is added, the mechanical strength of the solder joint may decrease, reducing the resistance against external force and, consequently, the wetting force. When a material that can be wetted is dipped into molten solder, the solder tends to spread across its surface. As the specimen is pulled upwards from the solder bath, an extra force, besides its own weight, resists the motion due to the solder’s adhesion. This opposing force is referred to as the wetting force. Obtaining a high wetting force is important for a solder as it deduced that the solder has good wettability properties. The maximum wetting force (F_max) of a solder alloy can be defined as in Eq. (2) [24]:where \({\text{P}}\) represents the perimeter of the specimen γ, denotes the surface tension between the solder and the flux, θ is the contact angle, ρ indicates the density of the solder, \({\text{g}}\) is the acceleration due to gravity, and \({\text{V}}\) refers to the immersed volume.

During a soldering process, an IMC layer develops as a result of the reaction between the molten solder and the Cu substrate. This layer plays a crucial role for a solder joint by establishing metallurgical bonding at the interface between the solder and the Cu substrate. However, if the IMC layer becomes too thick, it can negatively affect the long-term reliability of the solder joint. Figure 11 shows results of soldering interfaces at different compositions with the corresponding EDX analysis at point A. From the EDX analysis results, it indicates that the IMC layer (point A) was mainly composed of Sn and Cu elements, and it can be inferred from the results that Cu₆Sn₅ is the major inter-metallic compound at the interface. The intermetallic compound Cu₆Sn₅ is one of the most important phases formed at the interface between Sn-based solder and the Cu substrate. It exists in two structures, the high-temperature hexagonal η-phase and the low-temperature monoclinic ηʹ-phase, which transforms around 186 °C. This transformation causes about 2% volume expansion, which may induce interfacial stress and cracking in solder joints. Cu₆Sn₅ has a high hardness (5–7 GPa) and a Young’s modulus between 90 and 140 GPa, indicating that it is a hard and brittle phase. The coefficient of thermal expansion (CTE) is relatively high (about 18–32 × 10–6 K-1), which can create a thermal mismatch with Cu during temperature cycling. Doping with small amounts of Ni, Zn, In, or Au can stabilize the hexagonal η-phase at room temperature, reduce internal stress, and improve reliability. In solder joints, Cu₆Sn₅ acts as the main metallurgical bonding layer, but excessive growth or transformation to η′ can lead to brittle failure. Therefore, controlling the thickness and stability of the Cu₆Sn₅ layer is essential for maintaining joint strength and long-term reliability [25, 26].

Full size image

The thickness of the IMC layer is presented in Fig. 12. These results show that, with increasing In content, the total thickness of the IMC layer slightly decreases. It has been reported that nanosized Bi particles may segregate at the Cu₃Sn/Cu interface in solder joints, and the Bi segregation at the Cu₃Sn/Cu interface can accelerate the Cu₆Sn₅ IMC growth rate in microelectronics interconnects [10, 27]. Bi segregates by migrating toward the Cu₃Sn/Cu interface during thermal aging. This segregation, driven by the Kirkendall effect, facilitates interfacial atomic diffusion and lowers the energy barrier for IMC formation [28]. The presence of Bi at the interface enhances reaction kinetics, contributing to a faster growth rate of the IMC layer.

Full size image

Figure 13 shows the tensile properties of Sn-58Bi-xIn solder alloys, with stress–strain curves presented in Fig. 13(a) and the relationship between tensile strength and elongation rate in Fig. 13(b). The results indicate a clear trend where the tensile strength decreases from 70 MPa to approximately 50 MPa as In content increases from 0 to 12 wt.%, while elongation improves from 10 to 16%. Notably, at 4 wt.% and 8 wt.% In, tensile strength remains relatively stable, whereas elongation experiences a significant increase, reaching nearly 14%. At 12-wt.% In, tensile strength stabilizes around 50 MPa, while elongation continues to rise, reaching approximately 16%. This inverse relationship between tensile strength and elongation is attributed to the microstructural evolution caused by In addition. Indium, due to its atomic similarity to Sn and its tendency to dissolve in the Sn matrix, alters the microstructural distribution of Bi-rich phases. Previous studies suggest that In addition promotes the formation of BiIn intermetallic phases, which interrupt the continuity of the Sn–Bi eutectic structure, thereby reducing strength but improving ductility [29]. Furthermore, it has been observed that BiIn phases tend to coarsen with increasing In content, contributing to the observed mechanical property shifts. Shalaby et al. reported that In and Ag additions enhance mechanical performance in Sn–Bi-based alloys due to the solid solution strengthening effect of Bi in the Sn matrix and the role of intermetallic compounds as grain refiners [11, 13, 14]. However, in this study, Bi primarily functions as a solid solution strengthening element, while In addition counteracts this effect by softening the Sn matrix. Consequently, tensile strength exhibits minimal variation, whereas ductility improves significantly. Overall, the findings highlight the trade-off between strength and ductility in Sn-58Bi-xIn solder alloys. At 0 wt.% of In, the alloy shows the highest tensile strength but lowest elongation, indicating a brittle nature. When 4-wt.% In is added, both tensile strength and elongation decrease but more In addition, the elongation significantly improves while strength stabilizes, reflecting a transition to more ductile behavior.

Full size image

Figure 14 shows the SEM images of fracture mode analysis for the Sn-58Bi-xIn solder alloys after the tensile test. For Sn-58Bi, it can be seen that the fracture surface exhibits a predominantly brittle fracture mode, characterized by a relatively smooth and cleaved surface with large cracks. This observation is consistent with the inherent brittleness of the eutectic Sn-58Bi alloy due to the presence of Bi-rich phases, which contribute to its low ductility. With the addition of 4-wt.% In, the fracture surface shows a more ductile failure mode compared to the Sn-58Bi alloy. The morphology indicates a combination of brittle and ductile characteristics, with regions exhibiting rough, fibrous structures and some microvoid coalescence. At 8-wt.% In, the fracture surface becomes more porous and rough, with an increased number of dimples indicative of ductile fracture. The increasing existence of microvoid coalescence suggests improved plastic deformation. Adding 12 wt.% of In shows that the fracture mode appears highly ductile, exhibiting extensive microvoid formation and dimple rupture features. The increased presence of ductile tearing suggests enhanced toughness and reduced brittleness, attributed to higher In content. Overall, the progressive addition of Indium transforms the fracture mode from brittle in Sn-58Bi to quasi-brittle in Sn-58Bi-4In, and predominantly ductile in Sn-58Bi-8In and Sn-58Bi-12In.

Full size image