Confining effect of oxide film on tin whisker growth

1. Introduction

For 70 plus years, spontaneous Sn whisker growth has been a serious reliability problem in electronics, and thus extensively studied. However, although a host of mechanisms have been proposed, such as dislocation mechanism [[1], [2], [3]], recrystallization mechanism [4], compressive stress mechanism [[5], [6], [7]], electrostatic mechanism [8], interface energy driven mechanism [9,10], the phenomenon of spontaneous Sn whisker growth still has been resisting interpretation. The complexity of this phenomenon makes it much difficult to elucidate the mechanism behind it, and thus it has been a long-pending challenge for related industrial sectors.

Multiple difficulties associated with the research in this field include the random incubation periods, variable growth rates, morphology diversity, numerous influencing factors, and so forth [11]. The generally slow growth rate is one of the key issues that hinders whisker study. Although several methods have been developed to accelerate whisker growth [[12], [13], [14], [15]], cultivating Sn whisker is still a time-consuming and unpredictable process, with possible incubation time as long as a few years [16]. Developing a fast and controllable whisker cultivation method is certainly very beneficial to study whiskering phenomenon. Sn whisker morphology is another focus accompanied with whisker growth study. A number of types of morphology of Sn whisker including straight whisker, kinked whisker and curved whisker, have been reported [15]. Despite of the diverse morphologies, most of reported Sn whiskers have striations along their growth direction, and this feature is considered as one of the evidences by the stress-based mechanism [5]. Unfortunately, investigation of the striated Sn whisker morphology formation has been seldom reported. Worse still, the limited research results are inconsistent. Oxide [5] and hydroxide [17] crack mechanisms ascribed the striated morphology to the crack shape of the surface. Pei et al. [18] considered that the whisker morphology inherited from the shape of the initial whiskering grain, while LeBret et al. [19] claimed that Sn whisker has striated surface only when it nucleates on multiple grains. Also, the restriction of thin oxide film on the surface Sn whisker is considered to account for the filamentary morphology [20], but direct evidence of the shaping effect of the thin oxide film has been absent up to now. Therefore, probing into the origin of the unique morphology feature might make a significant contribution to uncover the mystery of whisker growth mechanism.

In this work, on a platform of Ti₂SnC-Sn with fast-growing Sn whiskers, the effect of the oxide film on whisker morphology is investigated via in-situ scanning electron microscopy (SEM) observation. The origin of the longitudinal striations on Sn whisker and the cause of whisker morphology evolution in vacuum were discussed.

2. Experimental

Ti (99.0%, 250 mesh), Sn (99.5%, 200 mesh), and graphite (99.85%, 500 mesh) powders with molar ratio of 2:1.2:1 were mixed for 24 h, and then heated at 1330 °C for 2 h in argon. After cooling to room temperature, Ti₂SnC with limited amount of free Sn (noted as Ti₂SnC-Sn) was synthesized. The Ti₂SnC-Sn was ball-milled and then cold-pressed into disc-shaped samples. The contact angle between Ti₂SnC and Sn was measured by the sessile drop method in vacuum. The morphology and composition of the Sn whiskers were characterized by scanning electron microscopy (SEM, Sirion 200, FEI), transmission electron microscopy (TEM, Tecnai G2-T20, FEI), and dual beam microscopy (SEM-FIB, Quanta 3D FEG, FEI).

3. Results and discussion

After only 2 days in open air, a great number of whiskers appeared on the disc sample, Fig. 1(a), and the longest whisker was measured ∼1667 μm. Given that the measured is actually the projected length and there is an incubation period for the whisker, the growth rate of this whisker is estimated as at least 10 nm/s, approximately 10³ to 10⁴ times greater than the whisker growth rate reported on Sn platings [21]. After 30 days, the forest of Sn whiskers covers the entire substrate, as shown in the inset of Fig. 1(a). The good predictability and controllability whiskering phenomenon on the platform of Ti₂SnC-Sn has been proved in our previous work, and it should be pointed that the existence of free Sn in Ti₂SnC substrate is necessary for Sn whisker to grow, and high-purity Ti₂SnC without Sn cannot grow whisker [22]. The predictability and ultra-high rate of Sn whisker growth achieved in this work enable the in-situ observation.

View original graphic| Download| PPT slide

Fig. 1.   (a) Sn whiskers grown on a Ti₂SnC-Sn sample after 2 days, and the sample entirely covered by a forest of Sn whiskers after 30 days (inset). (b) Typical striations on a Sn whisker. (c) Bright-field TEM image of a Sn whisker (the inset shows a SAED pattern). (d) HRTEM image of a selected area of a Sn whisker. (e) Oxide film on whisker surfaces. (f) Poor wettability between molten Sn and Ti₂SnC.

All whiskers wear longitudinal striations, Fig. 1(b), which is a typical feature of spontaneously grown Sn whiskers [17,19,23]. However, the understandings of origin of the striations are still controversial [5,18,19]. Fig. 1(c) shows a bright-field TEM image of a whisker. Different areas of the whisker share the same selected area electron diffraction (SAED) pattern, as shown in the inset of Fig. 1(c), meaning the whisker is β-Sn single crystal and the growth directions of this whisker is [010]. Fig. 1(d) shows the high-resolution TEM (HRTEM) image of a selected area of a Sn whisker. The lattice fringes with an interplanar spacing of 0.28 nm are assigned to the (101) plane of β-Sn. The structural perfection is evident from the regular crystal lattice in the HRTEM image, and a thin amorphous oxide film covers the whisker surface, as shown in Fig. 1(d) and (e). The growth direction of this whisker is indexed to be [103]. This is consistent with the conclusion that the growth direction of Sn whiskers is not exclusive [17].

Compressive stress is widely accepted as the fundamental driving force behind Sn whisker growth. The stresses may be intrinsic and/or extrinsic. In this work, stresses introduced during sample preparation and the volume expansion associated with the oxidation of Sn provide origin of stress. In addition, interface energy was also proposed as driving force for metal whisker growth recently [9,10]. Fig. 1(f) reveals the poor wettability between Sn and Ti₂SnC. The contact angle is as high as 150°. Therefore, the high interface energy between Sn and Ti₂SnC also contributes to Sn whisker growth. The interface flow also accounts for the fast mass transport of Sn atoms to feed fast Sn whisker growth [17,24]. Enhanced mass transport could be achieved according to Sn atoms diffusion along the large quantities of interfaces between Ti₂SnC and Sn, realizing the ultra-high growth rate of Sn whisker, facilitating the in-situ investigation.

It is natural to assume that the Sn whisker growth is an energy reduction process for a system. However, it is still not clear why the Sn growth takes the striated whisker morphology, which evidently experiences more surface energy penalty in comparison with other smooth configurations. Elucidating the origin of the striated whisker morphology is critical to restore the whisker growth process. Therefore, in-situ SEM observation of a sample alternately cultivated in air and SEM vacuum chamber was conducted, aiming at growing a whisker with alternate segments with/without oxide film. Luckily enough, this was realized taking the advantage of the controllability and high growth rate of Sn whiskers on Ti₂SnC-Sn.

Fig. 2(a)-(e) show the images of a Sn whisker after being alternately cultivated in air and vacuum for 1, 2, 3, 4 and 5 times, respectively. After 2 days in room ambience, the sample was transferred into the SEM vacuum chamber for the first time, and the projected length of the striated whisker (φ∼1.3μm) was evaluated on the SEM image as ∼23.4 μm. Interestingly, the newly grown segment of the whisker in vacuum has a faceted surface rather than the typically striated morphology, as pointed by arrow A in Fig. 2(a). As can be seen in Fig. 2(b), the faceted segment grows away from the substrate as the sample was taken out of the vacuum and cultivated in air for another 3 days, and another faceted segment (pointed by arrow B) was formed at the bottom of the whisker when it was cultivated in vacuum again. The whisker tip morphology remains unchanged as the whisker grew longer, which confirms that Sn whiskers grow by adding atoms to their bases [25]. More interestingly, the new segment of the whisker formed in air recovers the striated morphology, and it even shares the same cross-sectional contour as the previous segment formed in air, as seen in Fig. 2(f), which is the magnified view of the faceted segment pointed by arrow A in Fig. 2(b). As shown in Fig. 2(b)-(e), the alternate growth of faceted and striated segments happens at every switching of the environment, and the morphology of the faceted segments remain unchanged with continual whisker growth, as shown in Fig. 2(g) and (h).

View original graphic| Download| PPT slide

Fig. 2.   SEM images of a Sn whisker alternately cultivated in air and vacuum for (a) 1 time, (b) 2 times, (c) 3 times, (d) 4 times, and (e) 5 times (faceted segments formed in vacuum were pointed by white arrows). (f)-(h) are the high magnification SEM images of the faceted segments indicated by arrow A in (b), arrow C in (c), and arrow C in (d), respectively.

Electron irradiation is reported to affect Sn whisker growth [17,26]. To clarify whether the faceted segment formed in SEM chamber has relationship with the electron irradiation, another sample prepared in the same way was cultivated in SEM vacuum chamber (electron beam off) for 12 h. This resulted in a faceted segment up to ∼10 μm long (Fig. 3(a)). This long faceted segment could not be a result of the growth in a short SEM observation time of less than 5 min, when the electron beam was on. Thus it is reasonable to conclude that the faceted segment was formed in a vacuum without any other assistance or interference.

View original graphic| Download| PPT slide

Fig. 3.   (a) Sn whisker with a faceted segment length up to ∼10 μm was observed on the sample cultivated in SEM chamber free of electron beam for 12 h. (b) Sn whisker growing on Ti₂SnC-Sn substrate and (c) FIB cross-section of the whisker in (b).

As shown in Fig. 1(d) and (e), oxide film form on the Sn whisker surface in air, while it is absent in the SEM vacuum chamber. Therefore, the presence/absence of oxide film is essential for the striated-/faceted-morphology formation, and this provides direct evidence for the confining effect of the surface oxide film on the striated Sn whisker growth. Furthermore, this also indicates that the surface relaxation for Sn whisker easily happens (even at room temperature) without the confinement of oxide since Sn has a large diffusion coefficient [27,28]. In addition, the high mobility of Sn atoms at room temperature demonstrated in this platform means that spontaneous Sn whisker growth might have a strong relationship with its diffusion behaviour.

To understand why the whiskers evolve into quadrangular morphology in vacuum, morphological predication is performed according to the Bravais-Friedel-Donnay-Harker (BFDH) principle. The lowest-energy plane of β-Sn is calculated to be (200) plane, followed by (101) and (220). The faceted segments in both Fig. 2, Fig. 3(a) match with the configuration bounded by (200) planes and grew along the [001] direction, which is a construction resulted from the minimization of free energy. The frequently observed quadrangular morphology is consistent with the fact that [001] is reported to be one of the most preferential growth direction of Sn whiskers [17,29]. Given that the growth direction of Sn whiskers is not exclusive, other faced morphologies are also possible. In a word, Sn whiskers trend to take low energy configuration as long as they can freely relax their surface (without the restriction of oxide films) and thus result in a faceted morphology.

Densified Ti₂SnC-Sn synthesized by spark plasma sintering was employed to view the whisker root using the focused ion beam (FIB) sectioning. Fig. 3(b) shows a Sn whisker growing on the polished Ti₂SnC-Sn substrate. The tips of the whisker appear to be polished, indicating that the whisker nucleus pre-exists on Sn grain within the substrate. FIB cross-section of this whisker reveals that the whisker is embedded in the substrate and the cross-section of the whisker is consistent with the top-view of the whisker nucleus, Fig. 3(c), which means the striations on the Sn whisker are inherited from the nucleus.

The process of whisker growth and morphology evolution in this work is therefore illustrated in Fig. 4. The pre-existing free Sn within Ti₂SnC-Sn substrate serves as nucleus for Sn whisker, which is shown in Fig. 4(a). Sn whisker starts to grow after a short incubation period. Driven by compressive stress and the interface energy, Sn atoms diffuse fast via interface between Sn and Ti₂SnC. Both in air and vacuum, the whiskers grow by the same mechanism, and the only difference for the two scenarios is the effect of the oxygen in air on the whisker surface morphology evolution. In air, the cross-sectional contour originates from the nucleus will be reserved with the confining effect of instantly formed surface oxide film, and thus result in the formation of striated Sn whisker, as shown in Fig. 4(b). However, as the sample is transferred into vacuum where oxygen is absent, the confining effect of the oxide film disappears. Therefore, the newly grown segment of the whisker will evolve into faceted morphology via surface reconstruction as a result of surface energy reduction, as shown in Fig. 4(c). Given that the striations on whisker are inherited from the whisker nucleus, the whisker will recover the striated morphology once it is cultivated in air again, and the cross-section shape of different striated segments are the same, as shown in Fig. 4(d). The faceted segment grew in vacuum will also be covered by oxide film when the sample is taken out of vacuum.

View original graphic| Download| PPT slide

Fig. 4.   Schematic of the growth process of Sn whisker alternately cultivated in air and vacuum. (a) Sn whisker nucleus in Ti₂SnC substrate. (b) Striated Sn whisker formed in air. (c) Faceted Sn whisker formed in vacuum. (d) Sn whisker recovers the striated morphology once it is cultivated in air again.

4. Conclusions

In summary, rapid Sn whisker growth was realized on ball-milled Ti₂SnC-Sn, which facilitates the in-situ investigation of Sn whisker growth. Based on the compelling evidence captured by in-situ SEM observation, the confining effect of oxide film on Sn whisker morphology evolution is identified. The longitudinal striations on whisker surface are inherited from the whisker nucleus and are reserved by the confinement of the surface oxide film, while the formation of faceted segment in vacuum is achieved by the relaxation and reconstruction of the whisker surface driven by surface energy reduction. The findings in this work answer the mystery of the typical morphology of most spontaneously grown Sn whiskers, and are beneficial to further uncover Sn whisker growth mechanism.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 51731004, 51501038 and 51671054); Zhishan Scholar Program; and the Fundamental Research Funds for the Central Universities.

The authors have declared that no competing interests exist.