Introduction

Research and development in the electronics industry have been focusing towards improved or new high-technology products covering a wealth of business areas. In most cases, such specific consumer standards include miniaturization, high performance and good mechanical reliability. For this reason, materials and technologies which are involved in the industrial manufacturing and processes have been continuously developing. In case of the soldering process, one of the most investigated topics is related to an addition of various nano-sized particles into the commercial solder materials to develop a next generation of those. It should be noted that the main focus of these research is put on investigations and characterization of nanocomposite solders and their joints by minor additions of both ceramic and metal nanoparticles into the commercial Sn–Ag–Cu (SAC) solder paste (Tan et al. 2015; Zhang et al. 2019).

For instance, the authors have investigated an impact of various ceramic and metal nano-sized inclusions on microstructure and mechanical properties of the SAC305 solder joint (Yakymovych et al. 2016, 2018a). Furthermore, the changes in the structure as well as in physical and chemical properties of the liquid SAC solder caused by addition of various metal NPs were also investigated (Yakymovych et al. 2018b, c). It should be noted that the solder is reflowed during the soldering process. Therefore, these studies have given essential information to understand and simulate processes in the SAC nanocomposite solder during soldering; while, the reactive metal NPs are dissolved in the liquid Sn-based SAC matrix.

Other approach to reinforce the commercial SAC solder joints is based on an addition of NPs into the commercial flux. This strategy was firstly exanimated by Haseeb ASMA and his group used a flux doped with various metal NPs, such as Co, Mo, Ni and Ti (Ghosh et al. 2013; Sujan et al. 2014). Based on the obtained results, it was suggested that the morphology and growth of interfacial intermetallic compound layers between the solder and Cu substrate can be successfully controlled even with an addition of up to 0.1 wt% of metal NPs into the flux. We refer such solder/NPs doped flux/substrate joints as “hybrid solder joints”.

In the present study, the nanocomposite flux with 0.5 wt% and 1.0 wt% of various ceramic NPs was placed between SAC305 solder foil and Cu substrate providing the maximal concentration of nanoinclusions within the weakest place of the solder joint, namely at the interface solder/substrate. A comparison analysis of the microstructure of the interfacial IMCs layer and the shear stress measurement has been performed.

Materials and methods

Production of hybrid solder joints

The Sn–3.0Ag–0.5Sn (SAC305) solder alloy was synthesized from silver casting grains (99.9% metallic purity), tin ingot (99.998% metallic purity) and copper rod (99.9% metallic purity, all metals from Alfa Aesar, Karlsruhe, Germany). The samples were prepared by heating up to 1173 K accurately weighed amounts of pure components (within ± 0.1 mg) in evacuated and sealed quartz ampoules, and further aging at this temperature during 2 weeks. A SAC305 foil with a thickness of about 50 μm had been formed using a foil rolling mill.

Various NP-doped fluxes were prepared by mechanical mixing method of the commercial Pb-free solder flux (rosin-based Indium no-clean Tacflux 089HF; Indium Co., Milton Keynes, UK) with nominally 0.5 and 1.0 wt% ceramic nanoparticles. For instance, commercially available Al2O3, SiO2, TiO2, and ZrO2 nanopowders (average particle size 10–15 nm, TECNAN, Spain; Table 1) were used in the present study. Mechanical dispersion of NPs was achieved by manual stirring of nanoparticles in the flux for approximately 15 min at room temperature.

Table 1 Characteristics of the used ceramic nanopowders (TECNAN 2012)
Full size table

The solder joints were produced using 2-mm-thick copper strips as substrates, which were ground and polished with emery paper, soaked in 10% diluted sulfuric acid solution, rinsed in isopropanol. In this case, the sandwich-type Cu/(NP-doped flux)/SAC305/(NP-doped flux)/Cu set-ups were placed in a reflow furnace (LPKF Zelflow RO4) to produce the single shear-lap joints. The joint area was about 3.5 mm·2.0 mm·0.3 mm (Fig. 1).

Fig. 1
figure 1

Preparation of the hybrid solder joints

Full size image

This method has been described in detail elsewhere (Roshanghias et al. 2016). A reflow temperature profile implemented in the reflow soldering process is presented in Fig. 2. The joint samples were polished afterwards to remove the extruded solder and flux residue from the edges.

Fig. 2
figure 2

Reflow temperature profile

Full size image

A part of the produced joints was thermally aged in an Ecocell furnace at 353 K for 137 h and 496 h.

Characterization of hybrid solder joints

The shear strength of the solder joint was estimated using a µ-strain tensile machine (ME 30-1, Fuerstenfeld, Austria). The tests were performed at room temperature using a fixed crosshead speed of 2.0 mm min−1 and a minimum load resolution of 10 mN. The obtained results were proven to be safe after repeated trials at least three times for each sample condition.

The microstructure of the joints and their fracture surface after the performed shear tests was studied by scanning electron microscopy (SEM-EDX; FEI Quanta 200 and JEOL JSM 6600).

The average thickness of the IMC layer at the interface between the solder and the substrate was calculated using the Digimizer software, freely available in the Internet (MedCalc Software bv 2005–2019).

Results and discussion

Morphology and growth kinetics of interfacial IMCs layer

The cross-sectional micrographs were taken for the as-reflowed hybrid SAC305 solder joins to investigate changes in morphology of the Cu–Sn IMCs layer at the interface solder/Cu caused by minor additions of nano-Al2O3, -SiO2, -TiO2, and -ZrO2 to the commercial flux. For instance, the BSE micrographs of the unreinforced SAC305 solder joint and the hybrid SAC305 solder joints with 0.5 wt% Al2O3 NPs and SiO2 NPs are presented in Fig. 3.

Fig. 3
figure 3

BSE micrographs of SAC305 (a) and SAC305 + 0.5 wt% nano-Al2O3 and -SiO2 (b and c, respectively) solder joints in as-reflowed condition (I) and after thermal aging at 453 K for 137 h (II)

Full size image

As seen in Fig. 3a(I), the interfacial IMC layer of the as-reflowed SAC305 solder joint consisted of Cu6Sn5 IMCs; while the typical scallop-type Cu6Sn5 grains at the interface solder/Cu extended into the solder matrix. Addition of ceramic nanoparticles led to substantial changes in the morphology of the IMC layer at the interface (Fig. 3b(I) and c(I)). In contrast to our previous studies dealing with nanocomposite Cu/SAC305/Cu solder joints with metal and ceramic nanosized additions (Yakymovych et al. 2016, 2018a), no interfacial Cu3Sn IMC layer was observed at the investigated solder joints.

The thickness of the interfacial Cu6Sn5 IMC layer of the SAC305 solder joint increased after the thermal aging at 453 K for 137 h with a transition from the discontinuous to a more continuous scallop-type form (Fig. 3a(II)). Furthermore, a Cu3Sn IMC layer is formed between Cu substrate and the interfacial Cu6Sn5 IMC layer. The addition of 0.5 wt% of ceramic nanoparticles led to substantial changes in the morphology of the IMC layer towards a more planar type (Fig. 3b(II) and c(II)).

A line-scan analysis of the unreinforced SAC305/Cu solder joint confirmed the presence of a Cu3Sn layer of about 2.0 µm at the interface zone (Fig. 3IIa); while, the calculated average thickness of the entire interfacial Cu6Sn5/Cu3Sn layer was about 7.1 µm (Table 2).

Table 2 The total average thickness of IMC layers at the interface solder/Cu of the aged solder joints at 453 K for 137 h
Full size table

The changes in average thickness of the interfacial Cu–Sn IMCs layer formed at the Cu/flux/SAC305/flux/Cu by addition of ceramic NPs to the flux are presented in Fig. 4.

Fig. 4
figure 4

Average thickness of the interfacial Cu6Sn5 IMCs layer formed at the as-reflowed Cu/(NP-doped flux)/SAC305/(NP-doped flux)/Cu

Full size image

As seen from Figs. 3 and 4, minor additions of nano-Al2O3 and nano-SiO2 to the flux changed mainly the morphology of the interfacial IMCs layer of the hybrid SAC305 solder joints; while, the thickness of the IMCs layer slightly decreased to 10% for the 1.0 wt% nano-SiO2.

In contrast, additions of 0.5 wt% nano-TiO2 and nano-ZrO2 showed an impact on the thickness of the interfacial Cu6Sn5 layer, but in the opposite way: a decrease of the thickness in case of the nano-TiO2, and an increase in case of nano-ZrO2. Further increase of these ceramic nanoparticles in the flux up to 1 wt% resulted in similar impact, slightly increasing the average thickness of the interfacial Cu6Sn5 IMCs layer of the un-reinforced SAC305 solder joint. Based on the presented results, we concluded that an increase of the amount of nano-sized ceramic inclusions in the flux from 0.5 to 1.0 wt% led to less significant changes in the average thickness of the interfacial IMC layer of the as-reflowed hybrid solder joints. Similar tendency, indicated in the nanocomposite SAC305 solder joints, was explained by agglomeration and segregation of NPs during the reflowed process, which led to the decrease in their surface energy and reduced the amount of surface-active NPs (Wang et al. 2015; Yakymovych et al. 2016).

As seen in Fig. 5, the most pronounced decrease in the growth kinetics of the interfacial Cu–Sn layer for as-reflowed and aged solder joints was obtained by addition of up to 1.0 wt% nano-SiO2 (Fig. 5b), and 0.5 wt% nano-ZrO2 (Fig. 5d). In contrast, the addition of 0.5 wt% nano-TiO2 into the flux increased the growth kinetics of the interfacial Cu–Sn layer also of the aged hybrid solder joints, but is less pronounced by increasing the amount of nano-sized inclusions in the flux or/and the aging time (Fig. 5c). The results for the nano-Al2O3 were inconsistent and show an increase or decrease in the growth kinetics of the interfacial Cu–Sn layer depending on concentration of nanoinclusions in the flux and aging time (Fig. 5a).

Fig. 5
figure 5

Average thickness of the interfacial Cu–Sn IMCs layer formed at the aged Cu/(NP-doped flux)/SAC305/(NP-doped flux)/Cu edged at 453 K [(a)—nano-Al2O3, (b)—nano-SiO2, (c)—nano-TiO2, (d)—nano-ZrO2; black color—unreinforced flux; red color—0.5 wt%, green color—1.0 wt%]

Full size image

The shear stress investigations

The changes in the microstructure of nanocomposite solder joints described above should also result in changes of the mechanical properties. Therefore, it was decided to investigate the shear stress, as one of the main properties of the joint reliability.

Based on the results presented in Fig. 6, it can be concluded that the shear stress of the as-reflowed SAC305 solder joint increased by initial additions of various ceramic nanoparticles, namely by 0.5 wt% of nano-sized inclusions into the flux. The samples with 1.0 wt% of ceramic NPs showed less pronounced enhance of this mechanical property. This behavior is in agreement with literature related to the investigations of the nanocomposite SAC305 solder joint with CeO2 NPs postulated a gradual increase in the tensile strength for samples with up to 0.5 wt% ceramic NPs. At the same time, further additions of nano-CeO2 into the SAC305 solder joint led to deterioration of this mechanical property (Roshanghias et al. 2013). In contrast, the initial addition of nano-TiO2 lead to a decrease in the shear stress with its further increase for the hybrid solder joint with 1.0 wt% TiO2 NPs.

Fig. 6
figure 6

Shear stress of the as-reflowed hybrid SAC305 solder joins

Full size image

The minor nanosized ceramic additions lead to a slight deterioration of this mechanical property after thermal aging of the investigated hybrid solder joints (Fig. 7). It should be noted that an increased amount of ceramic nanoparticles in the SAC305 solder joint can lead to a higher degree of microporosity, which could form fine microscopic cracks, culminating in the failure of nanocomposite SAC solder joints (Tsao 2011).

Fig. 7
figure 7

Shear stress of the hybrid SAC305 solder joins aged at 453 K for 137 h

Full size image

The EDX analysis of the fractured surface was performed after the shear stress tests to establish the presence of nanosized ceramic inclusions in the solder joints. It was a very challenging topic taking into account minor presence of ceramic NPs in the investigated joints; nevertheless, some metal oxides in the pores of the investigated solder joints were determined. For instance, a presence of Al and Ti was found in the hybrid solder joints reinforced by nano-Al2O3 and nano-TiO2, respectively. As seen in Fig. 8, the results of the EDX analysis for the hybrid solder SAC305 solder joint with 1.0 wt% of nano-TiO2 showed the presence of Ti in the sample.

Fig. 8
figure 8

EDX analysis of the fractured surface of the hybrid SAC305 solder joint with 1.0 wt% nano-TiO2

Full size image

Furthermore, a type of fracture can be estimated from the SEM micrographs. Most of our samples showed a ductile type of fracture due to crack propagation within the solder joint similar to Fig. 9a. Occasionally, a more brittle fracture was also observed as a result of failure in the vicinity of the interface as presented (Fig. 9b). The differences in the shear stress of the hybrid solder joints such as the indicated decrease in this property by addition of 0.5 wt% nano-TiO2 into the flux can be explained by formation of the brittle cluster-like regions with a high concentration of ceramic nanoparticles inside the joint (Fig. 9b).

Fig. 9
figure 9

SEM micrograph of the aged hybrid solder joint with 1.0 wt% nano-SiO2 (a) and as-reflowed hybrid solder joint with 0.5 wt% nano-TiO2 (b)

Full size image

Based on the presented results it can be concluded that initial additions of various ceramic NPs into the flux lead to the changes both in morphology and thickness of the interfacial IMCs layer of the hybrid solder joints. Further increase of the nano-sized ceramic impurities in the flux up to 1 wt% made this effect less pronounced. Similar tendency was indicated also for nanocomoposite SAC305 solder joints using similar ceramic NPs (Yakymovych et al. 2016). Furthermore, a performed comparison analysis of the effects of minor additions of nano-sized SiO2, TiO2, and ZrO2 NPs into the SAC305 solder paste showed the most improved shear stress of the nanocomposite solder joint with nano-sized SiO2. Similar to that, the most profound effects on the shear stress of the hybrid SAC305 solder joints were also indicated for the samples with SiO2 nanoinclusions.

Conclusions

The presented results have shown that initial additions of various ceramic NPs into the flux lead to the changes both in morphology and thickness of the interfacial IMCs layer of the hybrid solder joints. Further increase of the nano-sized ceramic impurities in the flux up to 1 wt% made this effect less pronounced. The comparison analysis of the effects of minor additions of nano-sized SiO2, TiO2, and ZrO2 NPs into the SAC305 solder paste showed the most improved shear stress of the nanocomposite solder joint with nano-sized SiO2. Similarly, the most profound effects on the shear stress of the hybrid SAC305 solder joints were also indicated for the samples with SiO2 nanoinclusions.