Contents lists available at ScienceDirect

Microelectronics Reliability

journal homepage: www.elsevier.com/locate/microrel

A comparative analysis of microstructural features, tensile properties andwettability of hypoperitectic and peritectic Sn-Sb solder alloys

Marcelino Diasa, Thiago A. Costab, Bismarck L. Silvac, José E. Spinellid, Noé Cheunga,⁎,Amauri Garciaa

a Department of Manufacturing and Materials Engineering, University of Campinas, UNICAMP, 13083–860 Campinas, SP, Brazilb Federal Institute of Education, Science and Technology of Pará, IFPA, 66093-020 Belém, PA, Brazilc Department of Materials Engineering, Federal University of Rio Grande do Norte, UFRN, 59078-970 Natal, RN, Brazild Department of Materials Engineering, Federal University of São Carlos - UFSCar, 13565-905 São Carlos, SP, Brazil.

A R T I C L E I N F O

Keywords:Solder alloysSn-Sb alloysMicrostructureTensile propertiesWettability

A B S T R A C T

Sn-Sb alloys are among the current alternatives for the development of alloys for high-temperature lead-freesolders. The Sn-Sb alloys having 5.5 wt.% Sb or less are known to have good mechanical properties, and despitethe quite low liquidus temperature have been considered adequate in the development of solder joints. Theincrease in the Sb content up to the limit of solubility in Sn at about 10 wt.% is supposed to be detrimental to themechanical properties due to the extensive formation of an intermetallic compound. Investigations on the in-terrelation of microstructure of this alloy and the corresponding mechanical properties are fundamental to anappropriate evaluation of its application in solder joints. The present investigation analyses the relationshipbetween microstructural features of the peritectic Sn-10 wt.% Sb alloy, solidified under a wide range of coolingrates, and the resulting mechanical properties. A cellular β-Sn matrix, typified by cellular spacings that decreasewith the increase in the solidification cooling rate, and Sn3Sb2 particles are shown to characterize the alloymicrostructure. The ultimate tensile strength is higher as compared with the corresponding values of the hy-poperitectic Sn-5.5 wt.% Sb solder alloy, however the elongation is shown to decrease. A comparison with Bi-Agalloys, considered good high temperature solders alternatives, has shown that the tensile properties of the Sn-10 wt.% Sb alloy, including elongation, are significantly higher. Wettability tests have been carried out and theexperimental results, according to reports from the literature, are associated with good wettability.

1. Introduction

In the light of pressures to remove lead from electronics around theworld, the lead-free manufacturing has demanded studies on newsolder alloys to replace the traditional tin‑lead alloys, which were themainstay of the electronics industry in the last century. Not only yieldproduction but also field reliability has been the challenging issuesfaced by various candidate alloys that hardly compete with lead-basedalloys in terms of cost and technical benefits, such as low processingtemperatures associated with appropriate mechanical properties. Eachlead-free alloy has advantages and disadvantages i.e., none alloy isrecognized to fit all demands, and in this sense each different alloycomposition is suitable for niche applications [1–5]. A particular pro-cess, known as step soldering, involves solder alloys related to hightemperature electronic devices (from 270 °C to 350 °C), in which mul-tiple solders of successively lower melting points are used in con-secutive joints, in order to preserve the integrity of earlier soldered

joints [6–8]. This kind of technique employs mostly alloys far fromeutectic compositions, but with higher melting temperatures, and ag-gregating as consequence, more complex transformations such as theperitectic one. The literature is scarce on information related to solderalloys comprising a peritectic reaction, denoting a gap to be explored interms of characterization of microstructure and determination ofproperties.

The peritectic reaction is an invariant transformation in which aliquid phase (L) reacts with a solid phase (α) on cooling, giving rise to asecond solid phase (β), i.e., L + α > β. The solid (α) separate out from(L) at the peritectic temperature and must dissolve in L, and β mustfreeze out of L. Because the high cooling rates used in industrial pro-cessing, the completion of the equilibrium peritectic reaction is rarelyobserved, since β surrounds α and the peritectic reaction is stifled sinceL cannot reach α [9]. The literature reports works on peritectic alloysunder fast cooling rates such as: ferrous [10–12] and non-ferrous alloys[13–16]. Regarding the soldering process, high cooling rates are

https://doi.org/10.1016/j.microrel.2017.12.029Received 14 October 2017; Received in revised form 22 November 2017; Accepted 15 December 2017

⁎ Corresponding author.E-mail address: cheung@fem.unicamp.br (N. Cheung).

Microelectronics Reliability 81 (2018) 150–158

0026-2714/ © 2017 Elsevier Ltd. All rights reserved.

T

inherent during solidification inducing the formation of metastablephases. The Sn-Cu [17] and Sn-Sb [18] alloys systems are practicallythe only binary tin-based alloy systems used in high temperature sol-dering that undergo peritectic reactions. There exists a broad literaturerelated to alloys of the Sn-Cu system, although more concentrated inlow temperature applications, because of its importance as base ofternary alloys such as Sn-Cu-Ni [19,20] and Sn-Ag-Cu alloys [21,22].On the other hand, the same does not occur in similar extent to alloys ofthe Sn-Sb system, which deserve to be investigated in greater depth.

A peritectic reaction, L + Sn3Sb2 > Sn (β), occurs at the tin-richpart of the Sn-Sb binary phase diagram (243 °C), and from the peritecticcomposition the liquidus temperature (TL) increases with increasing Sbcontent [7,23]. It is worth noting the existence of conflicting inter-pretation in the literature concerning the composition of the first solidphase to separate out from the liquid at the liquidus temperature (i.e.

the intermetallic compound - IMC). The occurrence of either Sn3Sb2 orSnSb IMCs is reported in the literature. Some works affirm that Sn3Sb2decomposes into SnSb and β below 242 °C [24,25]. The most promisingSn-Sb alloys for soldering are the Sn-5wt.%Sb, which has a near peri-tectic composition, and the Sn-10 wt.% Sb alloy, the peritectic com-position. Kim and collaborators [6] have obtained samples of Sn-5 wt.%Sb and Sn-10wt.%Sb alloys solidified in air and performed character-ization of the resulting phases using X-ray diffraction and energy-dis-persive X-ray analyses. For both alloys, a β-Sn matrix and β-SbSn wereshown to constitute the microstructure. Recently, Yilmaz and colla-borators using the same characterization techniques, found the samephases in a Sn-10.2 wt.% Sb alloy directionally solidified under steady-state conditions [26]. The Sn-5.5wt.%Sb alloy was reported to haveboth good mechanical strength and creep properties, which are fun-damental to pin-attachment reliability. Dias et al. [18] established

Fig. 1. Schematic representation of the solidification ex-perimental setup.

Table 1Chemical composition (wt.%) of metals used to prepare the Sn-10 wt.% Sb alloy.

Element Pb Fe Sb Sn Cd Ni Cu Bi Zn Si

Sn 0.0469 0.0081 0.0005 Balance 0.00001 0.0001 0.0047 0.0046 0.0001 –Sb 0.215 0.075 Balance – – 0.034 0.034 – – 0.009

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

151

correlations between the length scale of the microstructure of mono-phasic Sn-2 wt.% Sb and Sn-5.5 wt.% Sb alloys and the correspondingtensile properties, and demonstrated that the size of the cellular matrixhas a significant effect on the tensile properties of the Sn-5.5 wt.% Sb

alloy. With the increase in the alloy Sb content, the extensive formationof the intermetallic phase has been considered to have possible dele-terious effect on the mechanical properties, mainly when the maximumsolubility of Sb in Sn is reached, i.e. about 10 wt.% Sb [1,27,28].

Fig. 2. Sn–Sb phase diagram with the analyzedSn–10 wt.% Sb alloy indicated by the dashed line[7,30].

Fig. 3. Schematics of removal of specimens from the DS castingfor tensile tests.

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

152

Although some of the studies on Sn-Sb alloys existing in the litera-ture emphasize mechanical properties, the literature is still scarce oncorrelations of such properties with microstructure features obtainedunder transient solidification conditions, which are also typical of sol-dering processes. In the present work, the Sn-10wt.%Sb alloy is direc-tionally solidified (DS) under transient cooling conditions, encom-passing a wide range of solidification cooling rates, including thosetypical of soldering practice. The solidification cooling rate will bedetermined and the phases composing the microstructure will be sub-jected to deep characterization by X-ray diffraction (XRD), ScanningElectron Microscopy/Wavelength-Dispersive Spectroscopy (WDS) andElectron Backscatter Diffraction (EBSD) analyses. Experimental lawscorrelating the representative length scale of the microstructure alongthe length of the DS casting with both the cooling rate and the corre-sponding tensile properties (ultimate-σu and yield-σy-tensile strengthsand elongation-to-fracture - δ) are envisaged. A comparative analysis ofthe present results with those of a Sn-5.5 wt.% Sb alloy will also beundertaken, which includes the examination of the wettability of bothalloys.

2. Experimental procedure

2.1. Solidification experiment

A water-cooled directional solidification setup was used, in whichtransient solidification takes place vertically upwards from a water-cooled bottom made of low carbon steel, as shown schematically inFig. 1. Some of the present authors and collaborators have described thecasting assembly, mold details and the thermocouples arrangementalong the length of the casting in previous studies [15,29]. The alloywas melted in situ and the lateral electric heaters had their powercontrolled in order to achieve a desired melt superheat. In the case ofthe present experiments when a superheat of 10% above the liquidustemperature was reached, the electric heaters were disconnected and atthe same time the water flow was initiated, thus permitting the onset ofsolidification. The compositions of metals used to prepare the Sn-10 wt.% Sb alloy are shown in Table 1. The amount of Pb in Sn, is in ac-cordance with ASTM B339 “Standard specification for pig tin”, for TinGrade “A”, which prescribes a lead content< 0.05 wt.%. The dashedline in the phase diagram of Fig. 2 indicates the Sn-10 wt.% Sb alloy.The liquidus temperature (TL = 268 °C) has been determined by slowlycooling the molten alloy in a refractory crucible having a thermocoupleplaced in its center. Such arrangement permitted the cooling curveunder conditions close to equilibrium, and consequently the change inslope corresponding to the liquidus temperature to be determined.

The solidification setup imposes a wide range of different coolingrates permitting, as a result, a significant variation in microstructurefeatures in a single casting experiment. The cooling curves, recorded byeach thermocouple positioned along the length of the directionallysolidified (DS) castings, have been used to determine the solidificationcooling rates.

2.2. Microstructural characterization

Selected transverse (perpendicular to the growth direction) andlongitudinal samples extracted at different positions along the length ofthe DS casting were polished and etched (with a solution 2% HCl, 3%NHO3 and 95% alcohol) for metallography. The same solution was usedto reveal the macrostructure of the DS casting. The etching process wasperformed by carefully immersing the samples in the etchant for only afew seconds, followed by cleaning the samples surface with runningwater and ethanol. Optical microscopy was performed using anOlympus Inverted Metallurgical Microscope (model 41GX) and thecellular spacing (λc) was measured by using the linear interceptmethod [26,31] from the optical images of the samples: about 40measurements at each selected section. The same samples were used forthe X-ray diffraction (XRD) analysis. A Panalytical X'pert PRO dif-fractometer was employed to obtain the XRD patterns under the 2θrange from 25° to 95° using Cu Kα radiation with a wavelength of0.15406 nm. Samples, for checking local alloy composition, were ex-tracted along different positions from the cooled surface of the DScastings and measured by an X-ray fluorescence spectrometer (XRF),Shimadzu EDX-720 model, to estimate local average concentrationsthrough an area of 100 mm2 probe.

In order to characterize the intermetallic compounds (IMCs)through Wavelength-Dispersive Spectroscopy (WDS) and ElectronBackscatter Diffraction (EBSD) techniques, the surface of the sampleswas carefully polished with colloidal silica using an automatic polishingmachine (Vibromet 2). A Scanning Electron Microscope (SEM) ZEISS-EVO-MA15 equipped with WDS and EBSD, both from OxfordInstrumental, and the Inca and Nordlys softwares were used, respec-tively, for the analyses.

2.3. Tensile tests

Transverse specimens were extracted from different positions along

Fig. 4. (a) Cooling curves along the length of the DS casting (TL is the liquidus tem-perature); (b) Cooling rate as a function of position in the casting.

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

153

the length of the DS casting, as shown schematically in Fig. 3, andprepared for tensile testing according to specifications of the ASTMStandard E 8M and tested in a MTS 810 machine at a strain rate ofabout 3 × 10−3 s−1. Since the solidification cooling rate is high atregions close to the bottom of the DS casting, decreasing progressivelytoward the top of the casting, the scale of the microstructure is affectedaccordingly, i.e., smaller values are associated with regions close to thecooled bottom increasing toward the top of the casting. With a view toexamining the effect of the scale of the microstructure on the tensileproperties, the tensile specimens were extracted from transverse sec-tions, as shown in Fig. 3. The approach is similar to the representationof a smooth continuous curve (e.g., represented by the trend of

experimentally measured cellular spacings along the length of thecastings) by a discrete variation in steps. That is, each step is related toa mean λc value that is the same all along each tensile specimen.

In order to ensure the reproducibility of results, three specimenswere tested for each selected position, so that average tensile propertieswere determined: yield tensile strength (σy), ultimate tensile strength(σu) and elongation (δ).

2.4. Wettability tests

For the wetting tests, specimens consisting of 4.0 mm-height × 4.0 mm-diameter cylindrical bars were extracted from thecentral part of the DS castings. The samples were dried properly andfinally coated by adequate flux (RMA – Rosin Mildly Activated) fortesting. The specimens were individually placed on the tester, heated ata rate of 10 °C/min up to temperatures of 20% above the correspondingliquidus temperature of the alloy. The contact angles (θ) were measuredin a Goniometer Krüss DSHAT HTM Reetz GmbH model from theaverage of θR and θL (R—right and L—left) values provided by acomputational method (tangent-2) and three tests were carried out foreach couple solder alloy/carbon steel substrate. The contact angleshave been measured continuously as far as the form of the moltendroplet changed.

3. Results and discussion

3.1. Solidification cooling rate, growth rate and microstructure features

The cooling curves acquired during directional solidification of theSn-10 wt.% Sb alloy are shown in Fig. 4a. They represent the thermalevolution at five different positions along the length of the DS castingmeasured by thermocouples placed along the centerline of the casting.

P= 10 mm

=7.4 K/s

λC=19.9 µm

P= 4 mm

=24.2 K/s

λC=12.6 µm

Fig. 5. Macrostructure of the DS Sn-10wt.%Sb alloy casting andtypical microstructures along the length of the casting (P = 4 andP = 10 mm from the cooled bottom).

0 2 4 6 8 10 12 140

5

10

75

80

85

90

95

100

Sb Sn

Sn, S

b (w

t. %

)

Position (mm)

Fig. 6. Sn and Sb profiles along the length of the DS casting.

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

154

These temperature-time data allowed the solidification cooling rate (Ṫ)to be determined. Ṫ has been computed by the time-derivative of thecooling curve (dT/dt) right after the passage of the liquidus isotherm byeach of the thermocouples, as shown in Fig. 4b as a function of position(P) along the length of the casting.

It can be seen in Fig. 5 that the DS casting has a typical columnarmacrostructure, with the grains aligned along the heat flow path, up toabout 25 mm from the cooled bottom. The corresponding optical mi-crostructures shows a cellular β-Sn matrix, characterized by cellularspacings (λC), which increase with the increase in Ṫ, as shown in Fig. 5for P = 4 and P = 10 mm.

With a view to permitting the occurrence of eventual macro-segregation profiles along the length of the DS castings to be analyzed,the experimental Sn and Sb concentrations have been determined, asshown in Fig. 6. The experimental points are quite close to the alloynominal composition, thus demonstrating that the vertical directionalsolidification has not induced macrosegregation profiles.

The evolution of λC as a function of Ṫ, during the directional soli-dification of the Sn-10 wt.% Sb alloy, has been experimentally de-termined and is plotted in Fig. 7. An experimental growth law in theform of a power function relating λC to Ṫ has been derived and is alsoshown in Fig. 7. For comparison purposes the corresponding growthlaw determined for the Sn-5.5 wt.% Sb solder alloy in a previous study[18], has also been included in Fig. 7. It can be seen that the increase in

the alloy solute content gives rise to increase in the cellular spacing (themultiplier of the experimental equation increases with the increase inthe alloy Sb content), while preserving a same exponent for both ex-perimental equations.

Since there is some controversy in the literature about the compo-sition of the primary intermetallic phase (IMC), with reports of occur-rence of either Sn-Sb [6,32] or Sn3Sb2 [7] IMCs, a deeper micro-structural analysis has been carried out by using differentcharacterization techniques such as: X-ray diffractometry (XRD), elec-tron microscopy (SEM)/Wavelength-dispersive spectroscopy (WDS)and Electron backscatter diffraction (EBSD).

Fig. 8 shows the XRD diffraction patterns (CuKα radiation i.e.,λ = 1.5418 Å), of samples extracted from positions along the DScasting associated with quite different solidification cooling rates(Ṫ= 7.3; 14 and 23.5 K/s). The XRD patterns were compared withcrystallographic data from the Inorganic Crystal Structure Databa-se–ICSD [33] and the JCPDS - Joint Committee on Powder DiffractionStandards file number 04-16-8009. It can be seen that despite the sig-nificant differences in Ṫ, the Sn3Sb2 IMC has been identified in any alloysample examined. Furthermore, selected SEM microstructures and WDSanalyses have also been employed to determine the phase compositionof the IMC particles disseminated throughout the microstructure, asshown by the SEM image of Fig. 9a. Table 2 shows the elemental SEM-WDS results of the IMC particles: 43.3 wt.% Sb and 56.7 wt.% Sn. Chenet al. [7] reported that the Sn3Sb2 IMC was found to have 43.2 wt.% Sband 56.2 wt.% Sn, which are values quite close to those experimentallydetermined in the present study.

According to Okamoto [30] the crystal structure of the Sn3Sb2 IMCis still unknown. Lidin et al. [34] emphasize the enigmatic nature ofsuch compound and report that some evidences indicate a possiblecubic structure, however, they pointed out that even one of the ele-ments forming such compound, i.e. Sb, has in fact a rhombohedraldistorted cubic lattice. The EBSD micrograph of Fig. 9b, and BackscatterKikuchi patterns (Fig. 9c) obtained through this technique, have shownthe presence of the Sn3Sb2 IMC that, associated with the X-ray dif-fraction data of Fig. 8 (with high intensity angles of 2θ= 28.9°and41.6°), give evidences that the Sn3Sb2 IMC has a rhombohedral crystalphase: Fm-3m Wyckoff space group and lattice parameters:a = 0.434 nm, b = 0.434 nm, c = 0.523 nm [33] The β-Sn phase hasbeen identified by XRD patterns in Fig. 8, with high intensity angles of2θ = 30.5°, 31.9°and 44.7°. Also, the detection of this crystalline phaseallows establishing a correspondence of these results with those ob-served in the EBSD patterns of Fig. 9b, since the β-Sn phase has a tet-ragonal crystal phase, R-3m Wyckoff space group and lattice para-meters: a = 0.583 nm, b = 0.583 nm, c = 0.318 nm [34–36].

3.2. Tensile properties and wettability

The length scale of the cellular microstructure was shown to affecthardness and tensile properties of a number of solder alloys. Refinedmicrostructures having lower cellular spacings, which are associatedwith higher cooling rates during solidification, have been related tohigher hardness for Zn-(10–40 wt.%)Sn alloys [37] and higher tensilestrength for Sn-0.7wt.%Cu-xNi (0–0.1wt.%Ni) [38]; Sn-2.0 and 5.5 wt.% Sb alloys [18]. This is generally attributed to the more homogeneousdistribution of the reinforcing phase throughout the microstructuretypified by cells that are more refined. The morphology and the re-presentative length scale of the microstructure of a Sn-based solder wasalso shown to affect other final properties, such as the electrochemicalcorrosion resistance [39].

In the present study, along the range of experimental cellular spa-cings examined, the tensile properties, that is, the ultimate tensile stress(σu), the yield stress (σy: 0.2% proof stress) and the elongation (δ) wereshown not to be affected by the size of the cells. The reinforcing phase,that is, the Sn3Sb2 IMC particles (shown in Fig. 9) seem to be moreeffective in the blockage of dislocations making the role of the scale of

Fig. 7. Experimental evolution of the cellular spacing as a function of cooling rate:Comparative analysis between Sn-10 wt.% Sb and Sn-5.5 wt.% Sb alloys.

Fig. 8. XRD patterns related to three different samples extracted from the DS Sn-10 wt.%Sb alloy casting from positions associated with different solidification cooling rates.

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

155

the cellular β-Sn matrix not significant. This is reflected by the constantvalues of ultimate and yield strengths and elongation vs. the cellularspacing, shown in Fig. 10. However, a previous study with the mono-phasic Sn-5.5wt.%Sb alloy [18] has shown that the increase in thecellular spacing provokes decrease in tensile strength and increase inelongation, and these results have been inserted in Fig. 10 for com-parison purposes. These results refer to cooling rates higher than 1.2 K/

s, i.e., despite having a monophasic alloy composition, the non-equili-brium solidification conditions induced the presence of lower Sn3Sb2volume fraction. Under these microstructural conditions associatedwith the Sn-5.5wt.%Sb alloy, that is, lower IMC fraction and cellular β-Sn matrix, the more refined cells have played a reinforcement role, thatis, with a more complex cellular network, the cell boundaries con-tributed to block more efficiently the motion of dislocations, thus in-creasing the tensile properties with the decrease in the cellular spacing,as shown in Fig. 10a and b. Such increase in strength is associated withthe way the refined microstructure makes the dislocation pile-up moreefficient during deformation. Smaller cellular spacings are related tobetter distribution of the harder Sn3Sb2 phase throughout the alloymicrostructure.

It can be seen in Fig. 10 that both σu and σy values of the Sn-10wt.%Sb alloy are quite higher when compared with the corresponding valuesof the Sn-5.5wt.%Sb alloy (from 18% up to 51% and 10% up to 47% for

Fig. 9. (a) SEM image highlighting the distribution of IMCsthroughout the Sn-10 wt.% Sb alloy microstructure; (b)EBSD micrograph: Sn3Sb2 (black) and β-Sn (gray); (c)Typical Backscatter Kikuchi Pattern of the Sn3Sb2 IMC.

Table 2Composition of the IMCs analyzed by WDS.

Element wt.% Sigma

Sb 43.322 0.162Sn 56.678 0.162Total 100.000

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

156

σu and σy, respectively). This is explained by the higher volume fractionof the hard Sn3Sb2 particles that are present in the microstructure of theSn-10 wt.% Sb alloy. In contrast, the elongation decreased from 0.4 upto 2.5 times. The mean experimental values of tensile properties ob-tained for the Sn-10 wt.% Sb alloy are: σu = 53 MPa, σy = 44 MPa andδ= 16%. Suganuma et al. [2] evaluated various characteristics of high-temperature lead-free alternative solders and concluded that Zn-Sn andBi-Ag alloys are among the best choices. Comparing the present resultswith those reported in the study of Suganuma et al. [2], it can be seenthat σu and δ of the Sn-10wt.%Sb alloy are lower than the corre-sponding values of the Zn-Sn alloys, while σy values are similar. How-ever, when compared with the values reported by the Bi-Ag alloys, thetensile properties of the Sn-10 wt.% Sb alloy are significantly higher.

Fig. 11 depicts the progression of contact angles along the wett-ability tests of Sn-5.5 wt.% Sb and Sn-10 wt.% Sb alloys samples on acarbon steel substrate. The initial contact conditions are associated withhigher contact angles, i.e., when the molten drop was not able tospread. After reaching an energetic equilibrium with the surface of thesubstrate, it can be seen that a quasi-steady-state regime is established.

Fig. 10. (a) Ultimate tensile stress; (b) yield tensile stress and (c) elongation, as a functionof the cellular spacing for the Sn-10wt.%Sb solder alloy.

Fig. 11. Experimental evolution of contact angle during the wetting tests of Sn-5.5 wt.%Sb and Sn-10 wt.% Sb alloys on a carbon steel substrate. The inset images refer to theinitial contact angles.

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

157

The initial contact angles of both alloys tested are quite close, that is,45.1° and 43.7° for the Sn-5.5 wt.% Sb and Sn-10wt.%Sb alloys, re-spectively. The present results are in agreement with some studies re-ported in the literature. Mahidhara et al. [40] reported a contact angleof 43 ± 4° for the Sn-5wt.%Sb alloy, which has been considered ade-quate for solder joints in microelectronics packaging. Plevachuk et al.[41] carried out an experimental study of density, surface tension andcontact angle of Sn-Sb based alloys and concluded that a contact angleof 40° is associated with good wettability. Kim et al. [6] investigated thewettability of Sn-5 wt.% Sb and Sn-10wt.%Sb alloys on Cu and Nisubstrates and reported contact angles of 13° (Cu); 14° (Ni) and 46°(Cu); 28° (Ni), respectively.

4. Conclusions

The following conclusions can be drawn from the present experi-mental investigation:

• Under the range of experimental cooling rates examined (2.5 to25 K/s) the optical microstructures of the Sn-10 wt.% Sb alloy wereshown to be characterized by a cellular β-Sn matrix, with the cel-lular spacing, λC, increasing with the decrease in the solidificationcooling rate, Ṫ. An experimental growth law relating λC to Ṫ hasbeen proposed.

• SEM images have shown IMC particles disseminated throughout themicrostructure. The composition of such IMC has been identified asSn3Sb2 by SEM-WDS, XRD and EBSD analyses.

• The ultimate tensile stress (σu), the yield stress (σy) and the elon-gation (δ) of the Sn-10wt.%Sb alloy were shown not to be affectedby the size of the cells. The mean experimental values were shown tobe: σu = 53 MPa, σy = 44 MPa and δ= 16%. These tensile strengthvalues, are quite higher when compared with the correspondingvalues of the Sn-5.5 wt.% Sb solder alloy, however, the elongationdecreased. As compared to values reported in the literature for high-temperature Bi-Ag solder alloys, the tensile properties of the Sn-10 wt.% Sb alloy are significantly higher.

• The results of wettability tests of Sn-5.5 wt.% Sb and Sn-10 wt.% Sballoy samples indicated contact angles of 45.1° and 43.7° for the Sn-5.5 wt.% Sb and Sn-10 wt.% Sb alloys, respectively, that, accordingto reports in the literature, are associated with good wettability.

• The melting point of 268° C associated with the good mechanicalproperties and wettability determined in the present study, giveindications that the peritectic Sn-10 wt.% Sb alloy can also be acompetitive high-temperature lead-free solder alternative.

Acknowledgements

The authors acknowledge the financial support provided byFAPESP-São Paulo Research Foundation, Brazil (grants 2016/18186-1;2013/09267-0; 2017/15158-0) and CNPq - The Brazilian ResearchCouncil. The authors would like to thank the Brazilian NanotechnologyNational Laboratory – LNNano for the use of its facilities.

References

[1] K.J. Puttlitz, K.A. Stalter, Handbook of Lead-Free Solder Technology for MicroelectronicAssemblies, CRC Press, New York, 2004.

[2] K. Suganuma, S. Kim, K. Kim, High-temperature lead-free solders: properties and possi-bilities, JOM-J. Met. 61 (2009) 64–71.

[3] E. Çadırlı, U. Böyük, S. Engin, H. Kaya, N. Maraşlı, M. Arı, Investigation of microhardnessand thermo-electrical properties in the Sn–Cu hypereutectic alloy, J. Mater. Sci. Mater.Electron. 21 (2010) 468–474.

[4] E. Çadırlı, H. Kaya, The effect of composition on microhardness and determination ofelectrical and thermal properties in the Sn-Cu alloys, J. Mater. Sci. Mater. Electron 22(2011) 1378–1386.

[5] T. Lee, T.R. Bieler, C. Kim, H. Ma, Fundamentals of Lead-Free Solder InterconnectTechnology: From Microstructures to Reliabilty, Springer, New York, 2015.

[6] J.H. Kim, S.W. Jeong, H.M. Lee, Thermodynamics-aided alloy design and evaluation of Pbfree solders for high-temperature applications, Mater. Trans. 43 (2002) 1873–1878.

[7] S.W. Chen, C.C. Chen, W. Gierlotka, A.R. Zi, P.Y. Chen, H.J. Wu, Phase equilibria of theSn–Sb binary system, J. Electron. Mater. 37 (2008) 992–1002.

[8] A.A. El-Daly, A. Fawzy, A.Z. Mohamad, A.M. El-Taher, Microstructural evolution andtensile properties of Sn–5Sb solder alloy, containing small amount of Ag and Cu, J. AlloysCompd. 509 (2011) 4574–4582.

[9] F.C. Campbell, Phase Diagrams – Understanding the Basics, ASM International, Ohio,2012.

[10] H.W. Kerr, W. Kurz, Solidification of peritectic alloys, Int. Mater. Rev. 41 (1996)129–164.

[11] Y.Z. Chen, F. Liu, G.C. Yang, Y.H. Zhou, Nonequilibrium effects of primary solidificationon peritectic reaction and transformation in undercooled peritectic Fe-Ni alloy, J. Mater.Res. 25 (2010) 1025–1029.

[12] Y.Z. Chen, K. Wang, X.H. Shi, X.Y. Ma, F. Liu, Phase selection and microstructural for-mation of rapidly directionally solidified peritectic Fe-Ni alloys by laser surface remelting,J. Alloys Compd. 585 (2014) 260–266.

[13] G.-D. Zhang, Z.-Y. Li, W. Xu, Y. Li, Laser resolidification of Zn-rich Zn-Ag peritectic alloys,T. Nonferr. Metal. Soc. China 12 (2002) 433–436.

[14] C. Brito, C.A. Siqueira, J.E. Spinelli, A. Garcia, Cellular growth during the transient di-rectional solidification of Zn-rich Zn-Cu monophasic and peritectic alloys, J. Phys. Chem.Solids 73 (2012) 1173–1181.

[15] M. Dias, C. Brito, F. Bertelli, O.L. Rocha, A. Garcia, Interconnection of thermal para-meters, microstructure, macrosegregation and microhardness of unidirectionally solidi-fied Zn-rich Zn–Ag peritectic alloys, Mater. Des. 63 (2014) 848–855.

[16] M.A.P. Castanho, P.R. Goulart, C. Brito, J.E. Spinelli, N. Cheung, A. Garcia, Steady andunsteady state peritectic solidification, Mater. Sci. Technol. 31 (2015) 105–114.

[17] S. Smith, G. Zeng, J. Read, S.D. McDonald, K. Nogita, Peritectic reactions and phasetransformations of Sn-30wt%Cu for high temperature Pb-Free soldering applications,Mater. Sci. Forum 857 (2016) 58–62.

[18] M. Dias, T. Costa, O. Rocha, J.E. Spinelli, N. Cheung, A. Garcia, Interconnection ofthermal parameters, microstructure and mechanical properties in directionally solidifiedSn–Sb lead-free solder alloys, Mater. Charact. 106 (2015) 52–61.

[19] C.H. Lin, S.W. Chen, C.H. Wang, Phase equilibria and solidification properties of Sn-Cu-Nialloys, J. Electron. Mater. 31 (2002) 907–915.

[20] B.L. Silva, N. Cheung, A. Garcia, J.E. Spinelli, Evaluation of solder/substrate thermalconductance and wetting angle of Sn-0.7wt%Cu-(0-0.1wt%Ni) solder alloys, Mater. Lett.142 (2015) 163–167.

[21] S.K. Kang, P. Lauro, D.Y. Shi, D.W. Henderson, J. Bartelo, T. Gosselin, S.R. Cain,C. Goldsmith, K. Puttlitz, T.K. Hwang, W.K. Choi, The microstructure, thermal fatigue,and failure analysis of near-ternary eutectic Sn-Ag-Cu solder joints, Mater T. JIM 45(2004) 695–702.

[22] B. Silva, J.E. Spinelli, A. Garcia, The roles of dendritic spacings and Ag3Sn intermetallicson hardness of the SAC307 solder alloy, Microelectron. Reliab. 54 (2014) 2929–2934.

[23] S.W. Chen, C.H. Wang, S.K. Lin, C.N. Chiu, Phase diagrams of Pb-free solders and theirrelated materials systems, in: K.V. Subramanian (Ed.), Lead-Free Electronic Solders - ASpecial Issue of the Journal of Materials Science: Materials in Electronics, KluwerAcademic Publishers-Plenum Publishers, 2007, pp. 19–37.

[24] W.P. Allen, J.H. Perepezko, Constitution of the tin-antimony system, Scr. Met. Mater. 24(1990) 2215–2220.

[25] W.P. Allen, J.H. Perepezko, Solidification of undercooled Sn-Sb peritectic alloys: part I.Microstructural evolution, Metall. Trans. A. 22 (1991) 753–764.

[26] E. Yılmaz, E. Çadırlı, E. Acer, M. Gündüz, Microstructural evolution and mechanicalproperties in directionally solidified Sn–10.2 Sb peritectic alloy at a constant temperaturegradient, Mater. Res-Ibero-Am J. 19 (2016) 370–378.

[27] A.R. Geranmayeh, R. Mahmudi, Power law identation creep of Sn-5%Sb solder alloy, J.Mater. Sci. 40 (2005) 3361–3366.

[28] J.H. Bae, K. Shin, J.H. Lee, M.Y. Kim, C.W. Yang, Development of high-temperaturesolders: contribution of transmission electron microscopy, Appl. Microsc. 45 (2015)89–94.

[29] M.V. Canté, J.E. Spinelli, N. Cheung, A. Garcia, The correlation between dendritic mi-crostructure and mechanical properties of directionally solidified hypoeutectic Al-Ni al-loys, Met. Mater. Int. 16 (2010) 39–49.

[30] H. Okamoto, Sb-Sn (Antimony-Tin), J. Phase Equilib. Diffus. 33 (2012) 347.[31] O.L. Rocha, C.A. Siqueira, A. Garcia, Cellular spacings in unsteady-state directionally

solidified Sn-Pb alloys, Mater. Sci. Eng. A 361 (2003) 111–118.[32] G.L.F. Powell, G.A. Colligan, A.W. Urquhart, Nucleation in undercooled Sn-Sb alloys,

Metall. Trans. A. 2 (1971) 918–919.[33] Inorganic Crystal Structure Database, Crystallographic Information Framework (CIF)

files, http://www.fiz-karlsruhe.de/icsd.html Accessed 24 August 2017.[34] S. Lidin, J. Christensen, K. Jansson, D. Fredrickson, R. Withers, L. Norén, S. Schmid,

Incommensurate stistaite – order made to order, Inorg. Chem. 48 (2009) 5497–5503.[35] T.N. Kolobyanina, S.S. Kabalkina, L.F. Vereshchagin, M.F. Kachan, V.G. Losev, A high-

pressure unit for X-ray diffractometer DRON-1, High Temp-High Press. 4 (1972)207–2011.

[36] H.E. Swanson, E. Tatge, Standard Xray Diffraction Powder Patterns National Bureau ofStandards (U.S.), Circular, 539 (1953), pp. 1–95.

[37] W.L.R. Santos, C. Brito, F. Bertelli, J.E. Spinelli, A. Garcia, Microstructural developmentof hypoeutectic Zn-(10–40) wt%Sn solder alloys and impacts of interphase spacing andmacrosegregation pattern on hardness, J. Alloys Compd. 647 (2015) 989–996.

[38] B.L. Silva, N. Cheung, A. Garcia, J.E. Spinelli, Sn-0.7wt.%Cu-(xNi) alloys: microstructure-mechanical properties correlations with solder/substrate interfacial heat transfer coeffi-cient, J. Alloys Compd. 632 (2015) 274–285.

[39] W.R. Osório, J.E. Spinelli, C.R.M. Afonso, L.C. Peixoto, A. Garcia, Microstructure, cor-rosion behaviour and microhardness of a directionally solidified Sn-Cu alloy, Electrochim.Acta 56 (2011) 8891–8899.

[40] R.K. Mahidhara, S.M. Sastry, I. Turlik, K.L. Murty, Deformation and fracture behavior ofSn-5%Sb solder, Scr. Met. Mater. 31 (1994) 1145–1150.

[41] Y. Plevachuk, W. Hoyer, I. Kaban, M. Kohler, R. Novakovic, Experimental study of den-sity, surface tension, and contact angle of Sn-Sb based alloys for high temperature sol-dering, J. Mater. Sci. 45 (2010) 2051–2056.

M. Dias et al. Microelectronics Reliability 81 (2018) 150–158

158