Microstructures and shear properties of Sb and In strengthened Sn5Bi/Cu joints

doi:10.21203/rs.3.rs-3550931/v1

Download PDF

Research Article

Microstructures and shear properties of Sb and In strengthened Sn5Bi/Cu joints

https://doi.org/10.21203/rs.3.rs-3550931/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The effects of x(Sb/In) (x = 1, 2, 3 wt.%) added to Sn5Bi solder alloy on the melting properties, microstructures and the shear behavior of solder/Cu joints were investigated combined with the corresponding phase diagram. It is found that the addition of Sb reduces the melting range and supercooling with the increase of Sb content caused by SnSb formation, while the In addition can reduce the melting point resulted by In solid solution. Both Sn5BixIn/Cu and Sn5BixSb/Cu are mainly composed of β-Sn and two types precipitates, i.e. Bi particles and Cu₆Sn₅ compound. The difference lies in that In doping only can dissolve into β-Sn and Cu₆Sn₅ while Sb doping mainly form tiny SnSb. Increase In content also inhibit precipitation of Bi particles and reduce the phase fraction Cu₆(SnIn)₅, resulting in that precipitation strengthening becomes weaker and solid solution strengthening becomes stronger. In contrast, the phase fraction of Bi particles and Cu₆Sn₅ increases with increasing Sb content, leading to that precipitation strengthening becomes stronger. These strengthening mechanism causes that their ultimate shear force monotonically increases with increasing Sb and In content, and the strengthening effect of the Sb element is better. Sb and In addition also can improve the ductility of Sn5Bi/Cu solder joints and both 2 wt.% In/Sb addition present the best shear fracture work. Thus, the optimal mechanical property was achieved by 2 wt.% Sb addition.

Tin solders

Solution strengthening

Shear property

Fracture

Phase diagram

Since the eutectic Sn-3.0Ag-0.5Cu (SAC305) alloy has become the mainstream solder of surface mount technology, the emerging trend for cost and high reliability driven it developed into the third generation of Sn-Ag-Cu (SAC) based alloys by doping elements [1]. The SAC305 alloy is a typical precipitation strengthening alloy, which is strengthened by insoluble and incoherent second phase particle, i.e. Ag₃Sn and Cu₆Sn₅ intermetallic (IMC) [2–5]. These strengthening effects meet the thermal reliability requirement in consumer electronics applications within − 25°C ~ 125°C [6; 7] and high temperature storage at maximum 150°C [8, 9]. However, the development of industrial electronic devices and automotive electronic products have brought the need for solder joint interconnection to operate under harsh conditions [10]. For instance, the operating temperature of service near the hood is about 150°C. More than that, the service life of industrial electronic devices is generally 10–20 years, which is about 1–5 years longer than that of consumer electronic appliances [1, 11]. That is, industrial electronic devices usually need to remain highly reliable for a long period of time under certain extreme harsh conditions, which are usually tested under isothermal aging around 170°C [11, 12]. This Pb-free alloy development is clearly from the automotive sector, but more demanding in medical, military, defense, and avionics [11].

The precipitation of non-coherent secondary phase particles in SAC solder alloys fail to satisfy the high thermal stability. Whereas, solid solution strengthening is an effective way to strengthening alloys at elevated temperatures due to that solubility of solute elements increases with increasing temperature. Even if the solubility decreases at low temperatures, precipitation strengthening can be achieved through element precipitation to strengthen the alloy. The industrial development of the Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni (Innolot) alloy has shown that the high-temperature deformation resistance and thermal resistance can be improved by doping Bi and Sb to achieve solid solution hardening [10, 13]. When service temperatures of the SAC305 solder alloys exceeds 150°C, the homologous temperature which is proportional to diffusion coefficient exceeds 0.8 as a result of its intrinsic low melting point. The non-coherent IMC particles, i.e Ag₃Sn and Cu₆Sn_5, easily coarsen [14, 15, 16]. Meanwhile, the β-Sn matrix tends to recrystallize and then the recrystallized grains facilitate crack nucleation and intergranular fracture in service [17–19]. During thermomechanical fatigue, the β-Sn matrix recrystallization and IMC particles coarsening have tendency to occur at the locations where the strain is concentrated [18]. Thus, introducing solid solution strengthening into precipitation hardening tin based alloys can effectively improve the high temperature resistance of the β-Sn matrix [20, 21]. For β-Sn matrix, solid solution strengthening can be achieved by Bi, Sb, In without obvious oxidation [4, 15, 22, 23]. Among them, Bi and In are used as main alloy elements (not micro alloy elements) to reduce the melting point, while the addition of Sb could increase the melting point [22, 24].

In fact, as early as 1997, the National Center for Manufacturing Sciences (NCMS) studied the formula of adding Bi, Sb, and In to Sn-Ag or SAC eutectic, and since 2002, National Institute of Standards and Technology (NIST) and the Colorado School of Mines have recorded their performance. However, the relationship between the amount of solid solution elements added and the interaction between solid solution elements, as well as their solder joints/Cu phase diagram are not well understood. Thus, in this paper, based on Sn5Bi solder/Cu substrate (introducing Cu₆Sn₅ precipitation strengthening), the addition of xIn and xSb (x = 1, 2, 3 wt.%) to study the effect of solution and precipitation strengthening elements on the microstructures and shear properties of solder joints. This study provides theoretical support for further development of a solder alloy system characterized by solid solution elements.

2.1 Sample preparation

The Sn5BixSb and Sn5BixIn (x = 0, 1, 2, 3 wt.%) solder alloys were prepared by mixing the Sn58Bi, Sn5Sb, Sn52In and pure Sn powders (3 # powder, Yunnan Tin Industry Co., Ltd.) and 12 wt.% flux (lab made) according to the nominal proportions. To mix them evenly, the stirring time was not less than 10 minutes. The solder paste was printed on a ceramic plate to form solder balls at 260°C, and then reflowed on FR4 Cu-OSP test boards with an opening of 0.8 mm in diameter for further microstructure observation and shear test. The remaining solder was melted to form an alloy block about 30 mg for the Differential Scanning Calorimetry (DSC) test.

2.2 DSC test

The DSC test (NETZSCH STA 449 F5) for melting properties was conducted in accordance with the JIS Z3198 industrial standard. All tests were carried out in an alumina crucible under high-purity argon with heating and cooling rates of 5°C per min between 30°C and 250°C (see Fig. 1a-b for the DSC sample preparation and the time-temperature-curve).

2.3 Microstructure observations of doped Sn5Bi/Cu BGA solder joints

The cross sections of the solder joints were cold mounted in acrylic resin, wet grounded using SiC grit paper in orders #320, #1200, and #3000, and then polished with 1.0 and 0.3 µm alumina suspension, respectively. Subsequently, the SnBixIn joint was etched with ferric chloride solution (FeCl₃ : HCl : H₂O = 10 g : 25 ml : 100 ml) ,while the SnBixSb joint was polished with silicon dioxide suspension polishing solution (OP-S type) without further etching required. Joint microstructures and elemental distributions were examined by Ultra-High Resolution Scanning Electron Microscope (FE-SEM, JSM-6480), Energy Dispersive Spectroscopy (EDS, IXRF 550ix) and Electron Probe Micro-Analysis (EPMA, EPM-810Q).

2.4 Ball shear test

Solder joint shear test was performed 72 h later after reflow using a PTR-1102 bonding tester with the shear height was fixed at 50 µm. The relatively low shear rate of 0.1 mm/s was chosen to ensure fracture propagation through the inside solder not the interface or IMC. At least six solder joints were sheared to failure under same condition for each alloy composition. After shear test, the fracture surface was analyzed using SEM.

3.1 Effect of Sb and In on the melting properties of Sn5Bi Solder

The DSC curves of Sn5BixSb and Sn5BixIn alloy are shown in Fig. 1c, which illustrates that only one phase transformation occurs in both melting and solidification processes. As the Sb content increases, the overall DSC curve moves towards high temperature. In contrast, as the In content increases, the overall DSC curve moves towards low temperature. The obtained DSC results are summarized in Tables 1. Compare to Sn5Bi, an increase in Sb content leads to an increase in solidus, peak values and liquidus, while an increase in In content results in an decrease in solidus, peak values and liquidus in general. The change trend of these solidus, peak values and liquidus are displayed in Fig. 1d. The trend of blue dots indicates that for both alloys, addition of Sb and In result in more significant changes in the solidus than in the peak values and liquidus. Accordingly, the melting range of the Sb doped solder decreases from 26°C of 1 wt.% to about 17°C of 2 and 3 wt.%, whereas the melting range of the In doped solder widens, which increases from 23.8°C of 1 wt.% to about 27.6°C of 3 wt.%.

Table 1

DSC results of Sn5BixSb and Sn5BixIn solder alloys
Alloy	Solidus (℃)	Peak (℃)	Liquidus (^oC)	Melting range (℃)	Supercooling (℃)
Sn5Bi	215.6	228.1	235.6	20.0	9.1
Sn5Bi1Sb	221.0	237.1	245.3	24.3	17.8
Sn5Bi2Sb	228.6	239.3	245.6	17.0	13.2
Sn5Bi3Sb	229.1	239.6	246.0	16.9	8.0
Sn5Bi1In	218.3	232.3	242.1	23.8	16.8
Sn5Bi2In	211.4	229.9	237.0	25.6	19.75
Sn5Bi3In	208.9	227.7	236.5	27.6	15.4

Furthermore, the red pentagram Fig. 1d illustrated the variation of undercooling with the type and content of elements added. The degree of undercooling was calculated from the difference between the heating onset temperature and cooling onset temperature. The undercooling of Sb doped alloys presents a monotonous downward trend, from 17.8°C of 1 wt.% to 8°C of 3 wt.%. While the undercooling of In doped alloys presents the trend of first increasing and then falling as In content increases, with the maximum value of 19.8°C at 2 wt.%. The degree of undercooling is important in the solidification process of solder alloys. Thus, these thermal properties differences may be related to the phase composition and phase change during the heating and cooling procedure.

3.2 Microstructure of Sn5BixSb/Cu

Figure 2a shows that Sn5Bi is composed of the gray β-Sn, bright white Bi particles, and Cu₆Sn₅ compound. When 1 wt.% Sb element is added (Fig. 2b), the compound phase in the solder joint is still Cu₆Sn₅ and Bi phase, and no SbSn compound is observed. When the content of Sb element is added to 3 wt.%, a square gray SbSn compound appears as shown by the arrow in Fig. 2c. As for SbSn compounds, the determination of them is controversial, because of three types SbSn compounds in the current literature, namely Sb₂Sn₃, Sb₃Sn₄, and SbSn phases. The latest version of Sb-Sn binary phase diagram published by Okamoto H et al. indicates that Sb₂Sn₃ should be a high temperature phase [25]. It forms due to the presence of an L + β(SbSn) → Sb₂Sn₃ peritectic reaction at 515 K (214.85°C) and decompose at 598 K (315.85°C), but the crystal structure of Sb₂Sn₃ is still unknown. However, Chen et al. [26] showed that the phases of SbSn and Sb₂Sn₃ could not be distinguished according to XRD experiments. Schmetter et al. [27] discussed Chen's results and considered that Sb₂Sn₃ did not exist in the SbSn system, but they proposed a room temperature stable binary phase Sb₃Sn₄ originated from a peritectic reaction of L + SbSn = Sb₃Sn₄ at 599 K (316.85°C). In addition, Schmetter et al. [27, 28] used the XRD to characterize the crystal structure of the SbSn phase as Sb₃Sn₄ phase with additional Sb layers. As for the stability of these compounds, Gierlotka et al. [29] calculated that the enthalpy of formation of Sb₃Sn₄ was − 29.600 KJ/mol based on the alternative solid solution model, and its crystal structure was rhombic. The structure of the SbSn phase is confirmed as NaCl type lattice, which is face-centered cubic. Based on the above reasons, the square Sb-Sn compound observed in this paper should be the SbSn phase.

Moreover, lower solubility of Sb in Sn at room temperature [11, 25–28], however, no Sb signal are observed in Sn5Bi1Sb/Cu and a few SnSb compound are observed in Sn5Bi3Sb/Cu solder joints in Fig. 2(b) and (c). Thus, to better understand the phase formation of the Sn5BixSb alloy, vertical section of the Sn-Bi-Sb ternary phase diagram at 5 wt.% Bi was drawn using Pandat software as shown in Fig. 2d. The dashed vertical line represents the sample composition, i.e. Sn5Bi1Sb and Sn5Bi3Sb. It can be seen that SnSb phase can form even though 1 wt.% Sb addition. Presence of Bi phase and SnSb phase which agrees well with experimental results. But, unlike primary IMC Cu6Sn5 with large size, tiny SnSb phase precipitates due to solid-state transformation. Combined with the inability of SEM's built-in EDS to distinguish elements Sn and Sb, small and a few SnSb phase are difficult to observe under SEM.

Because of EDS limitation, EPMA was conducted to see the distribution of Sb. The EPMA BSE images Sn5Bi3Sb/Cu joints and the elemental mapping at different magnifications are shown in Fig. 3. From Fig. 3a-e, the compounds of Cu (Cu₆Sn₅) are distributed in a network pattern at grain boundaries. Sb particles can be divided into large particles and smaller sized particles. The large Sb particles exhibit a certain corresponding distribution with Bi. There is often the formation of SbSn phase around Bi particles, which should be related to the solid formation of both phases. The excess Sb will inevitably be discharged in the area formed by Bi, creating conditions for the formation of SbSn phase with high concentration of Sn (Fig. 3k). By zooming in on the microstructure (Fig. 3f) and its corresponding element distribution (Fig. 3g-j), it finds that small SbSn phases with high density are more uniformly distributed on the Sn matrix. Within the field of view of the enlarged EPMA element distribution map, SbSn (Fig. 3h) and Bi (Fig. 3j) can only be seen some scattered particles, while Cu₆Sn₅ (Fig. 3i, Cu element) still maintains a net structure with a mesh size of approximately 5–10 µm. Small SbSn phase sizes are mostly 1–3 µm. Particle spacing is around 1–3 µm. If Cu₆Sn₅ phase is used to determine the grain size, the small SbSn phase should be distributed in β-Sn within the Sn crystal.. From the table in Fig. 3m, it shows that Sb element can be measured in the Cu₆Sn₅ phase at the interface and in the solder, with a content of 0.41 at.%, indicating that some Sn atoms have been replaced by Sb atoms to form a Cu₆ (Sn, Sb)₅ structure. The amount of Sb in the tin matrix is 2.96 wt.%, which is basically consistent with the amount added. Sb was also detected in the Bi phase. The precipitation of Bi phase is accompanied by the formation of SbSn phase, so there will inevitably be local aggregation of Sb phase inside the Bi phase. Or dissolution, the exact form of existence requires further and more detailed characterization work.

The above phase diagram drawing in Fig. 2d is based on the phase diagram of traditional bulk alloys, without considering the impact of the Cu substrate. Actually, the joint is quite different from the bulk alloy due to the incorporation of Cu atoms from the substrate during the joint formation process (see Fig. 4a). To further understand the behavior of solder joints, especially 3 wt.% with high Sb content, it is necessary to construct the Sn5Bi3Sb-Cu phase diagram to investigate the effect of the Cu substrate on the phase formation. Considering the peak reflow temperature is 250°C, the Cu equilibrium solubility (C_{Cu equ}.) of various Sn alloys are calculated ranging from 1.02 wt.% to 1.14 wt.% as listed in Table 2. Based on these values, the Cu content range of Sn5Bi3Sb-Cu phase diagram was set to 1.4 wt.% as shown in Fig. 4b. Compare with Fig. 2d, the shape of different phase areas are changed by the introduction of Cu. Furthermore, C_{Cu equ}. variation results in difference of the Cu₆Sn₅ phase fraction. With the addition of 5 wt.% Bi, the content of the Cu₆Sn₅ compound decreases from 3.7–3.54%, which may be related to the increase in solubility of Bi in the Sn lattice, making it easier for the concentration of Cu to reach saturation. When 1 wt.% Sb is added, C_{Cu equ}. in the solder joint alloy increases, and the content of Cu₆Sn₅ increases to 3.68%, which is equivalent to that of the pure tin solder joint. When the content of Sb further increases to 3 wt.%, C_{Cu equ}. has a large increase, reaching 1.57 wt.%, thus increasing the Cu₆Sn₅ content to 5.57%. Table 2 also presents that the SbSn phase fraction increases synchronously with the increase of Sb, while the Bi phase remains unchanged. This implies that, after solid solution in Sn, the excess Bi can precipitate independently due to not forming compounds with Sn, Cu, and Sb, thus being unaffected by the Sb content.

Table 2

Calculated Cu equilibrium solubility at 250 ^oC and calculated phase fraction of precipitates at 25 ^oC for variou Sn based solder joints
Composition	Cu equilibrium solubility at 250 ℃ (wt.%)	Cu₆Sn₅ phase fraction (mol/mol)	Bi phase fraction (mol/mol)	SnSb phase fraction (mol/mol)
Sn	1.4	3.7	-	-
Sn5Bi	1.02	3.54	1.63	-
Sn5Bi1Sb	1.06	3.68	1.62	2.12
Sn5Bi2Sb	1.09	3.78	1.61	4.26
Sn5Bi3Sb	1.12	3.89	1.60	6.41

3.3 Microstructure of Sn5BixIn/Cu

Figure 5 shows the microstructure, EDS spot analysis and the corresponding Sn5Bi-In alloy phase diagram. By comparing the Sn5Bi/Cu microstructure in Fig. 2a, it can be seen that, after adding 1, 3 wt.% of the In element (Fig. 5a-b), the main phases stay the same, but In partially replace Sn in β-Sn and Cu₆Sn₅ as shown in Fig. 5c. These phase types without considering the influence of Cu substrate conforms to the results shown in the Sn5Bi-In phase diagram (Fig. 5d). As the In content rising, the percentage of In replacing Sn in β-Sn and Cu₆Sn₅ increases, while the number of Bi phase decreases significantly. This is because the addition of In can increase the solid solubility of Bi in Sn, thereby inhibiting the separation of Bi. Compared to Cu₆(Sn, In)₅ in the solder joint, the content of In in Cu₆(Sn, In)₅ at the interface is relatively low as shown in Fig. 5c. EPMA of Sn5Bi3Sb/Cu was conducted for more accurate characterization as shown in Fig. 6a-f. In was found mainly dissolved in the β-Sn with 2.86 wt.% In, followed by Cu6Sn5, and finally in Bi (Fig. 6g). The In content in interfacial Cu₆Sn₅ is 0.66 at.%, which is significantly higher than the substitution amount of Sb in Cu₆Sn₅ (Fig. 6g).

Furthermore, the diagram of the Sn5Bi3In-Cu solder joint is drawn as shown in Fig. 7. The Cu equilibrium solubility at 250 ^oC and phase fraction of precipitates at 25 ^oC for In doped solder joints are calculated as presented in Table 3. The C_{Cu equ}. of In doped solder joints is lower than those of their Sb counterparts. This can be verified by comparing the Cu distribution of Fig. 3d and Fig. 6d. Obviously, less Cu6Sn5 phase are found in In doped solder joints. In doping also results in an increase in the size of Cu₆Sn₅, near the interface, which agree well with previous research [30]. Similarly, for Sn5Bi2In/Cu and Sn5Bi3In/Cu solder joints, their calculated phase fraction of Bi is much lower than those of their Sb counterparts. This is consistent with the Bi distribution of Fig. 3e and Fig. 6e. The experimental results and calculation results have good consistency, indicating that combined CALculation of PHAse Diagrams (CALPHA) and microstructure characterization can better understanding the microstructures difference of various composition solder joints.

Table 3

Calculated Cu equilibrium solubility at 250 ^oC and calculated phase fraction of precipitates at 25 ^oC for Sn5BixIn (x = 1, 2, 3)/Cu joints
Composition	Cu equilibrium solubility at 250 ℃ (wt.%)	Cu₆Sn₅ phase fraction (mol/mol)	Bi phase fraction (mol/mol)
Sn5Bi1In	0.90	2.97	1.52
Sn5Bi2In	0.88	2.89	1.24
Sn5Bi3In	0.85	2.81	0.91

3.4 Mechanical behavior of Sn5Bix (Sb, In)/Cu solder joints

3.4.1 Shear behavior

Figure 8 shows the two typical shear curves of the Sn5Bix(Sb, In)/Cu joints. According to the fracture behavior after ultimate shear force (USF) point, a sharp drop in force, or resistance to crack propagation with more smooth dropping before fracture, namely fracture mode 1 and mode 2 respectively. For mode 1, the cutting area after shear is generally greater than 50%, while it is less than 50% for mode 2. The former has good plasticity, while the latter has weak resistance to deformation (Fig. 8a). As shown in Fig. 8b, Sn5Bi fractures in both mode, consisting of 60% mode 1 and 40% mode 2. After adding x (Sb, In) elements, the fracture ratio of mode 1 decreased to at least 80%. Among them, when the amount of Sb added is 2 wt.% and the addition amount of In is 1 wt.% and 2 wt.%, all the joints fracture in mode 1. The fracture behavior of alloys has a good correlation with the work of crack growth. The total shear work (TW) refers to the area under shear force vs displacement graph from initial shear force to the fracture. The TW (blue triangle in Fig. 8b) provided that solder joints with 100% fracture ratio of mode 1 exhibit the highest TW, while solder joints with 60% fracture ratio of mode 1 has the minimum TW. These results reveal that In or Sb addition can improve the ductility of Sn5Bi/Cu solder joints and both 2 wt.% In/Sb addition present the best TW.

Addition In or Sb of also can enhance their USF as the red circle shown in Fig. 8b. With the increase of Sb/In addition, the USF of the solder joint presents a monotonous increase trend, and 3 wt.% Sb doping reached the maximum USF value of 42.11 N which increase by 23.06%. Combining with the results of USF and TW, the Sn5Bi2Sb/Cu solder joints demonstrate optimal mechanical properties.

3.4.2 Mechanical performance corresponded to Sb/In additions

To better understand the relationship between the microstructures and shear properties of Sn5Bix(Sb, In)/Cu solder joints, Fig. 9 illustrates the calculated phase fraction corresponding to the shear results of xSb/In doped joints. The proportion of particles in Sn5BixSb/Cu joints shows that the phase fraction of SbSn, Cu₆Sn₅ and the total second phases containing Bi particles increase linearly with the increase in the Sb content, and the USF also shows a linear relationship with Sb content (Figs. 9a-d). The crack propagation work (CPW) is defined as the area under the force vs displacement from USF to the end of the shear displacement. The CPW increases parabola, which first rises with increasing Sb, and reaches the maximum when the content of Sb is 2 wt.% (Fig. 9f). The TW for all the ductile fracture joints shows a similar trend in Fig. 9g.

For Sn5BixIn/Cu joints (see Fig. 9b1-g1), the phase fraction of Bi particles, Cu₆Sn₅, and the sum of the two of them decreased with the In content increasing from 1 to 3 wt.%, showing a linear decline in slope. This trend is consistent with the conclusion of Belyakov et al. [4]. Increasing the content of In, on the one hand, increases the solid solution strengthening part caused by the solution of Bi and In in the matrix, on the other hand, the addition of In reduces the precipitation of Bi particles and Cu6(Sn, In)5, making the precipitation strengthening part weak. However, when In continues to increase, the solution strengthening and precipitation strengthening combine together. This result can be verified by the USF curve versus In content, almost linear growth in USF from 0 ~ 2 wt.% In, with increasing content of In to 3 wt.%, the curve deviates from the straight line and grows in a parabola shape. The CPW for In doped solder joints in Fig. 9f1 exhibit the same trend as the Sb doped solder joints. The TW for the ductile fracture joints shows that 1 wt.% In doped Sn5Bi1In joints displays a poor performance (Fig. 9g1).

Masaki Moriuchi et al.[20] first theoretically investigated the solution strengthening mechanism of SAC doped with Bi and Sb. In the solution strengthening mechanism (we suppose that it is precipitate strengthening here), the solute atmospheres that are formed around solute atoms distort the crystalline structure and become obstacles to dislocations. When the atomic radius between the solute atom and the matrix phase atom is different, the size effect caused by them can be expressed by equations (1) and (2) [31].

$$\begin{array}{c}{F}_{\text{m}}=\mu b{r}_{\text{s}}\left|\epsilon \right|\text{ }\text{(}\text{1}\text{)}\end{array}$$

$$\begin{array}{c}\epsilon =\left({r}_{\text{s}}-{r}_{\text{m}}\right)/{r}_{\text{m}}\text{ }\text{(}\text{2}\text{)}\end{array}$$

Where ${F}_{\text{m}}$is the maximum force of the obstacle acting on a dislocation. $b$ is the Burgers vector, ${r}_{\text{m}}$ is the atomic radius in the matrix phase, and$\epsilon$ is the misfit strain. In Eq. (2),${r}_{\text{s}}$ and ${r}_{\text{m}}$ are the solute atomic radius and the atomic radius in the matrix phase, respectively. For In and Sb in Sn5Bi matrix, the $\epsilon$ and the ${F}_{\text{m}}$ could be expressed in the following.

$${\epsilon }_{\text{I}\text{n}}=\left({r}_{\text{s}}-{r}_{\text{m}}\right)/{r}_{\text{m}}=\left(167-140\right)/140=0.19$$

$${F}_{\text{m},\text{I}\text{n}}=\mu b167\text{*}0.19=31.73 {\mu }_{\text{I}\text{n}}{b}_{\text{I}\text{n}}$$

$${\epsilon }_{\text{S}\text{b}}=\left({r}_{\text{s}}-{r}_{\text{m}}\right)/{r}_{\text{m}}=\left(128-140\right)/140=0.08$$

$${F}_{\text{m},\text{S}\text{b}}=\mu b128\text{*}0.08=10.24{\mu }_{\text{S}\text{b}}{b}_{\text{S}\text{b}}$$

Thus, the atomic-induced lattice distortion caused by the In solid solution is greater and that of Sb is smaller, assumed that ${\mu }_{\text{S}\text{b}}{b}_{\text{S}\text{b}} \text{a}\text{n}\text{d} {\mu }_{\text{I}\text{n}}{b}_{\text{I}\text{n}}$ are generally identical. According to the above results from microstructure and the calculated phase diagram, Sb addition will form fine SbSn IMC, which is no longer solute in Sn matrix. Therefore, the Sb addition made the atomic misfit no longer computed further for the critical resolved shear stress. The critical resolved shear stress dependent on ${F}_{\text{m}}$ and concentration of solute atoms in the solid solution alloy C can be expressed by Labusch limit [31] as shown in Eq. (3). It is considered that the multiple solute atoms are involved in the elementary step of a one-time dislocation motion, when the solute atomic concentration is high.

$$\begin{array}{c}{\tau }_{\text{a}}={\left[\left({F}_{\text{m}}^{4}{C}^{2}w\right)/\left(4\mu {b}^{9}\right)\right]}^{1/3} \text{(}\text{3}\text{)}\end{array}$$

where, ${\tau }_{\text{a}}$ is the critical shear stress, C is the concentration of the solute atoms, w is a parameter that describes the range to which the interaction between the dislocations and the solute atoms extends, which is approximately 5b [20]. Thus, the critical resolved shear stress caused by In addition can be further expressed by Eq. (4).

$$\begin{array}{c}{\tau }_{\text{a}，\text{I}\text{n}}=108.21{{\mu }_{\text{I}\text{n}}{b}_{\text{I}\text{n}}}^{-4/3}{C}^{2/3} \text{(}\text{4}\text{)}\end{array}$$

In this paper, the proportion of elements in Sn5BixIn (x = 1, 2, 3) is calculated by the weight percentage, while $C$ is the atomic percentage of solute atoms, which can be obtained by calculation:

$${C}_{1\text{I}\text{n} }=1.06\%$$

$${C}_{2\text{I}\text{n} }=2.11\%$$

$${C}_{3\text{I}\text{n} }=3.17\%$$

Take the value of $C$ into Eq. (4) to get ${\tau }_{\text{a}，\text{I}\text{n}}$:

$${\tau }_{\text{a}，1\text{I}\text{n}}=\text{5.22}{{\mu }_{\text{I}\text{n}}{b}_{\text{I}\text{n}}}^{-4/3}$$

$${\tau }_{\text{a}，2\text{I}\text{n}}=\text{8.26}{{\mu }_{\text{I}\text{n}}{b}_{\text{I}\text{n}}}^{-4/3}$$

$${\tau }_{\text{a}，3\text{I}\text{n}}=\text{10.84}{{\mu }_{\text{I}\text{n}}{b}_{\text{I}\text{n}}}^{-4/3}$$

As for Sb, it is actually an IMC particle existent in tin matrix, the critical shear stress should be related to Eq. (5) related to the precipitation strengthening caused by IMC particles [32].

$$\begin{array}{c}\varDelta {\tau }_{p}=a{f}^{1/2}{p}^{-1} \text{(}\text{5}\text{)}\end{array}$$

where a is a material-related constant, f is the volume percentage of IMC particles, and p is the average radius of particles. The greater the particle density and the smaller the particle size, the more obvious the effect of this strengthening mechanism. This ensures that the Sb added Sn5BixSb alloy is stronger than the In doped one. For $\varDelta {\tau }_{p}\propto$ ${f}^{1/2}$, the In/Sb doped joints reinforced via IMCs can be shown in Fig. 10a, which shows an opposite trend, differently from the USF results shown in Fig. 9e1. Further plot the solution induced strengthening curve shown in Fig. 10b, it is increasing with In addition which may compensate the weaker precipitate strengthening compared to Sb induced precipitate strengthening. Making a parabolic increase not linear curve resulted by Sb addition, see Fig. 9e1. This kind of USF change may result from changes in the strengthening mechanism by the inner characteristics of the element.

4.1 The relationship between the thermal properties and phases of Sn5Bix(Sb, In)/Cu solder joints

As shown in Fig. 11a-c, the latent heat, △H, for phase transformation shows an increasing trend with the increase of the Sb content in Fig. 11c. This phenomenon should be related to the formation of SbSn. The existence of these high-temperature phases changes the specific heat capacity and thermal conductivity of the alloy and then changes the microscopic heat transfer system during the phase transformation process.

Due to the formation of the high temperature phase of the SbSn phase, the specific heat capacity decreases (peak height of the endothermic peak [33]) and the average thermal conductivity k in the solid-liquid two-phase region decreases, resulting in a longer heat input time for phase transformation. Thus the phase transformation peak widens and full width at half maxima (FWHM) increases (Fig. 11a) with increase in system entropy [33], which is consistent with the results of Sn5BixIn alloy. However, since In is in a solid solution state in the alloy and the thermal conductivity of pure indium is higher than that of tin, the atomic mismatch between In and Sn improves the lattice distortion energy and then reduces the solidus of the alloy (Fig. 11b). The latent heat in for the phase transformation process eventually decreases with an increase in the In content (Fig. 11a).

The effects of Sb and In addition (x = 1 wt.%, 2 wt.%, 3 wt.%) on the melting behavior, microstructure, shear and fracture of Sn5Bi/Cu joints can be summarized as follows. The solidus, peak temperature and liquidus of the alloy solder showed an upward trend with the increase of the Sb content. When the addition of Sb increased from 1 wt.% to 3 wt.%, the solidus temperature of the alloy increased by 8.1°C. However, the liquidus temperature only increases by 0.7°C, which shows an upward trend in the melting range of the alloy. The solidus, peak temperature, and liquidus decrease with the In increasing. When In content increased from 1 wt.% to 3 wt.%, the liquidus temperature decreased by 5.6°C and the solidus decreased by 9.4°C, leading to an increase of the melting range from 23.8°C to 27.6°C.

The phase composition of Sn5BixIn/Cu and Sn5BixSb/Cu consists of the gray β-Sn matrix, bright white Bi phase particles, and the Cu₆Sn₅ compound generally. However, the solder joint containing In has the In distribution in each phase, while the solder joint containing Sb shows a dark gray SbSn square compound, clearly seen in Sn5Bi3Sb/Cu solder joints with 3 wt.% Sb addition.

The ultimate shear force (USF) of the solder/Cu joints increases linearly with the increase of Sb and parabolically with increasing In content. Solution strengthening is the main strengthening mechanism, and dispersion strengthening is the secondary strengthening mode, in which dispersion strengthening consists mainly of compound phase (SbSn, Cu₆Sn₅ or Cu₆(Sn, In)₅) and the solid precipitate phase (Bi phase) in the solder joint. The addition of In inhibits the segregation of the Bi particles, and the amount of the Bi phase precipitated in the solid state decreases, which makes the USF of solder joints increased more obviously when the same content of In and Sb is added. However, because of the existence of certain solid solubility of Bi and In in Sn, both of them have the effect of solid solution strengthening.

The content of Sb and In is 2 wt.%, the anti-crack propagation ability is the strongest, and the solder joint has the best anti-crack propagation ability. The work of resisting crack propagation is the maximum value when the content is 2 wt.%, which has a great relationship with the instability propagation of solder joints.

Acknowledgments:

The authors would like to thank Professor H. L. Zhou from Jiangsu University of Science and Technology for her support and assistance in sample production and circuit board design. The researchers would like to acknowledge Deanship of Scientific Research, Taif University for funding this work.

Funding:

The authors gratefully acknowledge the supports from the following funds:

The Yunnan Fundamental Research Projects (grant No. 202301BC070001-001) funded by the Yunnan Provincial Department of Science and Technology;

The Jiangsu Province Industry-university-research cooperation project (No. BY2022832) funded by the Jiangsu Provincial Department of Science and Technology.

The researchers would like to acknowledge Deanship of Scientific Research, Taif University for funding this work.

Author contributions:

Jiahen Zhang: Experiment, Original draft, Data curation, Formal analysis. Yongheng zhao: Experiment, Data curation, Formal analysis, Writing - original draft. Xiaojing Wang: Writing - review & editing, Discussion, Conceptualization. Shanshan Cai: Writing -review & editing, Discussion. Jubo Peng: Supervision, Writing - review. Chen Liu: DSC test. Islam H. El Azab and Mohamed M. Ibrahim: Writing - review, Discussion. Shuai Wang: Phase diagram drawing. Baosheng Liu: Review and Discussion. Weibin Zhang: Phase diagram preparation and microstructure discussion. Cai Chen: Review & editing. Fengyi Guo: EPMA test, Yingwu Wang: Review & editing. Salah M. El-Bahy: English polish.

Data and code availability:

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of interest or competing interests:

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Ethical approval:

Experiments in this paper don’t involve in any vivo testing on animal subjects, human subjects, or human tissue.

Lau JH, Lee NC (2020) Assembly and Reliability of Lead-Free Solder Joints, Springer Nature Singapore Pte Ltd., Singapore.
Cui Y, Xian YW, Zois A, et al(2023) Nucleation and growth of Ag3Sn in Sn-Ag and Sn-Ag-Cu solder alloys. ACTA MATERIALIA 249 (2023) 118831. https://doi.org/10.1016/j.actamat.2023.118831
Zhang L, Long WM, Wang FJ (2020) Microstructures, interface reaction and properties of Sn-Ag-Cu and Sn-Ag-Cu-0.5CuZnAl solders on Fe substrate. J MATER SCI-MATER EL 31:6645-6653. https://doi.org/10.1007/s10854-020-03220-1
Belyakov SA, Nishimura T, Akaiwa T (2019) Role of Bi, Sb and In in microstructure formation and properties of Sn-0.7Cu-0.05Ni-X BGA interconnections. 2019 International Conference on Electronics Packaging (ICEP):235-239. https://doi.org/10.23919/icep.2019.8733493
Emadi F, Vuorinen V, Ross G et al (2023) Co, In, and Co–In alloyed Cu6Sn5 interconnects: Microstructural and mechanical characteristics. MATER SCI ENG A 881 (2023) 145398. https://doi.org/10.1016/j.msea.2023.145398
Belyakov SA, Gourlay CM, Coyle R, et al (2020) Microstructure evolution during thermal fatigue of a SAC solder containing 5.5%Sb, Proceedings of SMTA International.
Lai YS, Wang TH, Tsai HH, et al (2020) Cyclic bending reliability of wafer-level chip-scale packages. MICROELECTRON RELIAB 47:111-117. https://doi.org/10.1016/j.microrel.2006.04.010
Choudhury P, Ribas Mda, Pandher RS, et al (2015) Development of Lead-Free Alloys with Ultra-High Thermo-Mechanical Reliability, Proceedings of SMTA International.
Wai LC, Wei SW, Yuan HH, et al (2014) High temperature die attach material on ENEPIG surface for high temperature (250DegC/500hour) and temperature cycle (−65 to +150DegC) applications. EPTC, Singapore, pp:229-234. https://doi.org/10.1109/EPTC.2014.7028376
Coyle R, Johnson C, Hillman D, et al (2019) thermal cycling reliability and failure mode of two ball grid array packages with high reliability pb-free solder alloys. SMTA International, Sept:22-26.
Li C, Chen S, Cai S et al(2023) Microstructure and shear properties evolution of Mn-doped SAC solder joint under isothermal aging. J IRON STEEL RES INT 30 (2023) 1650-1660. https://doi.org/10.1007/s42243-023-01027-y.
Dalton E, Ren G, Punch J, et al (2018) Accelerated temperature cycling induced strain and failure behaviour for BGA assemblies of third generation high Ag content Pb-free solder alloys. MATER DESIGN 154:184-191. https://doi.org/10.1016/j.matdes.2018.05.030
Skwarek, Agata, Illes, et al (2018) Reliability studies of InnoLot and SnBi joints soldered on DBC substrate. SOLDER SURF MT TECH 30:205-212. https://doi.org/10.1108/SSMT-10-2017-0029
Xian JW, Belyakov SA, Gourlay CM (2020) Time-Lapse Imaging of Ag3Sn Thermal Coarsening in Sn-3Ag-0.5Cu Solder Joints. J ELECTRON MATER 50(3): 1-10. https://doi.org/10.1007/s11664-020-08498-9
Belyakov SA, Xian J, Zeng G, et al (2019) Precipitation and coarsening of bismuth plates in Sn–Ag–Cu–Bi and Sn–Cu–Ni–Bi solder joints. J MATER SCI-MATER EL:1-13. https://doi.org/ 10.1007/s10854-018-0302-8
Wang X, Zhu Q, Wang Z, et al (2009) MODELING OF Ag3Sn COARSENING AND ITS EFFECT ON CREEP IN Sn-Ag-Cu SOLDER. ACTA Metall Sin 45(8): 912-918. https://doi.org/ 10.1016/j.jallcom.2009.03.176
Xu H, Liu F, Zhou L, et al (2020) Failure mechanism of lead-free component boards in thermomechanical test based on recrystallization enhanced cracking. MICROELECTRON RELIAB 110:113623. https://doi.org/10.1016/j.microrel.2020.113623
Yin L, Wentlent L, Yang L, et al (2012) Recrystallization and precipitate coarsening in Pb-free solder joints during thermomechanical fatigue. J ELECTRON MATER 41:241-252. https://doi.org/10.1007/s11664-011-1762-2
Dutta I, Kumar P, Subbarayan G (2009) Microstructural coarsening in Sn-Ag-based solders and its effects on mechanical properties. JOM-US 61:29-38. https://doi.org/10.1007/s11837-009-0085-8
Moriuchi M, Kariya Y, Kondo M, et al (2022) Class I creep deformation of Sn–Ag–Cu containing solid solution elements and its effect on thermal fatigue life of solder joints. MATER TRANS 63(6):805-812. https://doi.org/10.2320/matertrans.MT-MC2022015
Belyakov S, Nishimura T, Akaiwa T, et al (2019) Role of Bi, Sb and In in microstructure formation and properties of Sn-0.7 Cu-0.05 Ni-X BGA interconnections. 2019 International Conference on Electronics Packaging (ICEP):235-239. https://doi.org/10.23919/ICEP.2019.8733493
Luo X, Peng J, Zhang W, et al (2022) CALPHAD-guided alloy design of Sn–In based solder joints with multiphase structure and their mechanical properties. MATER SCI ENG A 860:144284. https://doi.org/10.1016/j.msea.2022.144284
Cai S, Luo X, Peng J, et al (2021) Deformation mechanism of various Sn-x Bi alloys under tensile tests. ADV COMPOS HYBRID MA 4:379-391. https://doi.org/10.1007/s42114-021-00231-2
Wu X, Wu J, Wang X, et al (2020) Effect of In addition on microstructure and mechanical properties of Sn–40Bi alloys. J MATER SCI 55:3092-3106. https://doi.org/10.1007/s10853-019-04148-6
Okamoto H (2012) Sb-Sn (Antimony-Tin). Journal of Phase Equilibria and Diffusion 33(4): 347-347.
Chen SW, Chen CC, Gierlotka W, et al (2008) Phase equilibria of the Sn-Sb binary system. J ELECTRON MATER 37:992-1002. https://doi.org/10.1007/s11664-008-0464-x
Schmetterer C, Polt J, Flandorfer H (2018) The phase equilibria in the Sb-Sn system–Part II: Experimental results. J ALLOY COMPD 743:523-536. https://doi.org/10.1016/j.jallcom.2017.11.367
Schmetterer C, Polt J, Flandorfer H (2017) The phase equilibria in the Sb-Sn system–Part I: Literature review. J ALLOY COMPD 728:497-505. https://doi.org/10.1016/j.jallcom.2017.08.215
Gierlotka W (2020) On the binary Sb–Sn system: ab initio calculation and thermodynamic remodeling. J MATER SCI 55(1):347-357. https://doi.org/10.1007/s10853-019-03934-6
Labusch R (1970) A statistical theory of solid solution hardening. PHYS STATUS SOLIDI B 41(2):659-669. https://doi.org/10.1002/pssb.19700410221
Courtney T(2005) Mechanical Behavior of Materials, McGraw-Hill Science.
Wang C(2014) Phase diagrams and its application, Higher Education Press.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
19 Dec, 2023
Reviews received at journal
11 Dec, 2023
Reviewers agreed at journal
11 Dec, 2023
Reviewers invited by journal
11 Dec, 2023
Editor assigned by journal
17 Nov, 2023
Submission checks completed at journal
12 Nov, 2023
First submitted to journal
03 Nov, 2023

You are reading this latest preprint version

Microstructures and shear properties of Sb and In strengthened Sn5Bi/Cu joints

Status:

Version 1

Abstract

Figures

1. Introduction

2. Experimental procedures

2.1 Sample preparation

2.2 DSC test

2.3 Microstructure observations of doped Sn5Bi/Cu BGA solder joints

2.4 Ball shear test

3. Results and discussion

3.1 Effect of Sb and In on the melting properties of Sn5Bi Solder

3.2 Microstructure of Sn5BixSb/Cu

3.3 Microstructure of Sn5BixIn/Cu

3.4 Mechanical behavior of Sn5Bix (Sb, In)/Cu solder joints

3.4.1 Shear behavior

3.4.2 Mechanical performance corresponded to Sb/In additions

4.1 The relationship between the thermal properties and phases of Sn5Bix(Sb, In)/Cu solder joints

Conclusions

Declarations

References

Additional Declarations

Status:

Version 1