Author's personal copy - University of California, Berkeleyaminor.mse.berkeley.edu/Publications_files/109 Acta cryo.pdfmajor constituent of most solders used by industry. Reli-ability

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

http://www.elsevier.com/authorsrights

Author's personal copy

Cryogenic in situ microcompression testing of Sn

A. Lupinacci a, J. Kacher a,b, A. Eilenberg a, A.A. Shapiro c, P. Hosemann d,A.M. Minor a,b,⇑

a Department of Materials Science and Engineering, University of California, Berkeley, CA, USAb National Center for Electron Microscopy, Lawrence Berkeley National Laboratory, Berkeley, CA, USA

c Jet Propulsion Laboratory, Pasadena, CA, USAd Department of Nuclear Engineering, University of California, Berkeley, CA, USA

Received 30 April 2014; received in revised form 11 June 2014; accepted 11 June 2014Available online 16 July 2014

Abstract

Characterizing plasticity mechanisms below the ductile-to-brittle transition temperature is traditionally difficult to accomplish in asystematic fashion. Here, we use a new experimental setup to perform in situ cryogenic mechanical testing of pure Sn micropillars atroom temperature and at �142 �C. Subsequent electron microscopy characterization of the micropillars shows a clear difference inthe deformation mechanisms at room temperature and at cryogenic temperatures. At room temperature, the Sn micropillars deformedthrough dislocation plasticity, while at �142 �C they exhibited both higher strength and deformation twinning. Two different orienta-tions were tested, a symmetric (100) orientation and a non-symmetric (4�51) orientation. The deformation mechanisms were found tobe the same for both orientations.� 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Tin; Solder; DBTT; Small-scale testing; EBSD

1. Introduction

The use of small-scale mechanical testing techniquesprovides a number of advantages for evaluating materialin extreme environments. For example, one advantage isthat small single crystals can be milled from polycrystallinesamples so that similarly oriented samples can be systemat-ically tested over different conditions. A further advantageis that small-scale testing gives us the ability test samplesin situ in an electron microscope in order to observe defor-mation mechanisms directly as well as monitor the qualityof the tests. One environmental extreme of interest foraerospace applications is ultracold temperatures. For spaceapplications in particular, it is important to understand the

mechanical performance of components at the extremetemperature conditions seen in service for an accurateassessment of a component’s performance. For solderalloys used in microelectronics, cryogenic temperaturescan prove especially problematic. At low temperaturesSn-based solders undergo a ductile-to-brittle transition thatleads to brittle cracks, which can result in catastrophic fail-ure of electronic components, assemblies and spacecraftpayloads.

Sn–Pb solder joints are used extensively in packaging ofelectronics. These joints provide both electrical connectionsand mechanical support for the electronic package. Differ-ences in the thermal expansion coefficients of materialsused in electronic packaging cause cyclical strains in thesolder joints upon thermal cycling [1]. The mechanicalproperties of Sn are of particular interest since Sn is themajor constituent of most solders used by industry. Reli-ability of the solder joint depends upon a number of

http://dx.doi.org/10.1016/j.actamat.2014.06.026

1359-6454/� 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Department of Materials Science andEngineering, University of California, Berkeley, CA, USA.

E-mail address: [email protected] (A.M. Minor).

www.elsevier.com/locate/actamat

Available online at www.sciencedirect.com

ScienceDirect

Acta Materialia 78 (2014) 56–64


mechanical and microstructural evolutionary processesthat occur interactively during its service lifetime [2].Approximately 70% of failure in electronic circuitry is sol-der related; therefore understanding the deformation mech-anisms associated with Sn is increasingly important [2]. Atroom temperature Sn has a body-centered tetragonal (bct)structure. Sn exhibits an allotropic phase transformationfrom the bct structure to a diamond cubic structure at13.2 �C, commonly known as tin pest. The transformationis accompanied by a volume increase of 27% [3]. However,this transformation is rather sluggish, and in pure Snrequires nucleation and growth with an incubation periodrequiring months and even years for completion [4,5]. Sincemost space applications are thermally cycled from hot tocold temperature extremes over a relatively short time-frame (hours), the primary structure of interest is the bctstructure of Sn and the corresponding ductile-to-brittletransition that this structure exhibits. The crystal orienta-tion and microstructure of the Sn phase will have signifi-cant effects on the mechanical properties of solder joints[6]. This structure gives rise to several unique deformationmechanisms that change as a function of temperature,strain rate and orientation. At room temperature and rela-tively slow strain rates, Sn deforms by dislocation-mediatedslip. The primary slip systems at room temperature havebeen identified as {1 00) < 001], {1 10) < 001] and

{100) < 01 0] [7,8]. While these are the preferred slip sys-tems, depending on the crystal orientation, Sn can accessup to 10 different slip systems [9]. Conversely, at higherstrain rates, one of the more recognizable deformationmechanisms at room temperature is tin cry [10]. Tin cryis a result of twinning, or shearing of crystallographicplanes in Sn. Chalmers [11] first identified the primarytwinning system at room temperature to be the{301}h103i. Additional studies have also identified{101}h101i as a secondary twinning system [12]. In thelow temperature regime, solder has been primarily charac-terized by Charpy impact testing [13]. Only limited studieshave been performed below room temperature but abovethe ductile-to-brittle transition temperature (DBTT) of�125 �C [14]. While the Charpy test is very useful for iden-tifying the transition from ductile to brittle behavior, itdoes not shed light on the mechanisms that are drivingthese transitions. Previous studies have demonstrated theuse of in situ small-scale cryogenic indentation testing[15]; however, no similar systems have been reported thatcould reach a low enough temperature to characterize theductile-to-brittle transition of Sn. There has yet to be a sys-tematic study that explores the deformation mechanismsbelow the DBTT and at strain rates that do not inducefracture, which is representative of solder joints that arethermally cycled in this regime. The objective of this study

Fig. 1. Overview of the in situ cryogenic testing system. (a) Schematic of the cooling system and the nanoindentation stage. (b) Cooling curve showing thatthe system stabilizes �2000 s after cooling is turned on. (c) Demonstration of the thermal stability of the system by using a thermocouple as a sample. Asmall deviation in the temperature recording can be seen around 25 s corresponding to the point at which contact to the thermocouple was made by thediamond indenter tip. After this initial contact the temperature stabilized almost immediately while the tip remained in contact with the sample.

A. Lupinacci et al. / Acta Materialia 78 (2014) 56–64 57


is to better understand the change in deformation mecha-nisms responsible for the ductile-to-brittle transition inSn. This study discusses the use of a new experimentalapparatus, an in situ scanning electron microscopecryogenic mechanical testing system, and its initial applica-tion to characterize the plasticity of Sn both at roomtemperature and below the DBTT.

2. Experiments and methods

2.1. Specimen fabrication

Microcompression specimens (micropillars) with asquare cross-section of 3 � 3 lm in width were fabricatedfrom a polycrystalline Sn sample using a dual-beamfocused ion beam (FIB) microscope. Prior to micropillarfabrication, the sample was mechanically polished on twoadjacent faces. During polishing the Sn was mounteddirectly next to a thin piece of steel in order to preventdeformation and rounding near the edge of the Sn. Thesamples were planarized using SiC grinding paper withwater as a lubricant and were then polished with 0.3 and0.1 lm alumina polishing solutions, and 0.05 lm colloidalsilica polishing solution.

Prior to ion beam machining, electron backscatter dif-fraction (EBSD) was used to identify grains for micropillarlocations. Two grains were chosen for micropillar fabrica-tion in order to investigate any effect of orientation ondeformation behavior below the DBTT. The first grainhad a (10 0) orientation with respect to the applied loaddirection and the second grain was rotated �52� from thefirst orientation with a (4�5 1) orientation. Fabrication ofthe micropillars was similar to the method demonstratedby Reinhold et al. [16] using a FEI Quanta dual-beam

FIB scanning electron microscope with a Ga+ ion sourceoperated at 30 keV. Micropillars were fabricated close tothe polished edge using an ion current of 1 nA for coarsemilling and 10 pA for a final polish of the pillar faces.The sample taper was controlled by polishing all four sam-ple sides individually with adjusted tilt angles of the orderof 2�. A final polish was done on the top surface of themicropillar by tilting the sample to 90� in order to ensurethat top of the micropillar is flat and uniform. These stepswere taken in order ensure a homogeneous stress distribu-tion during loading. A total of seven micropillars weremanufactured in the (100) grain and four micropillars weremanufactured in the (4�51) grain.

Fig. 2. Summary of the stress–strain behavior of the Sn micropillars fromthe (100) grain, compressed both at room temperature and at �142 �C.The cryogenic micropillars show both higher strength and large strainbursts at yielding.

Fig. 3. SEM micrographs from in situ microcompression testing at room temperature (top row) and �142 �C (bottom row). Top row: (left) image near thestart of the room temperature test, (middle) image towards the end of the room temperature test, (right) image from a different angle after the room temptest. Bottom row: (left) image near the start of the cryogenic test, (middle) image directly after a large deformation event in the cryogenic test, (right) imagefrom a different angle after the cryogenic test.

58 A. Lupinacci et al. / Acta Materialia 78 (2014) 56–64


2.2. Cryogenic micromechanical testing

Micropillar tests were performed on the (100) and (4�51)oriented specimens at both room temperature at �142 �C.Testing was performed in situ in the FIB using a HysitronPI-85 nanoindenter with a flat-tip diamond punch. To per-form cryogenic testing, a custom-built cryogenic coolingsystem manufactured by Hummingbird Scientific was usedin conjunction with the Hysitron PI-85 nanoindenter (seeFig. 1). The cryo system cooled the tip and sample simulta-neously in order to minimize thermal drift. Fig. 1a shows aschematic of the cooling system and the nanoindenter.Fig. 1b and c show temperature data taken by attachinga thermocouple directly to the sample platform to charac-terize the stability of the cooling system. We want toemphasize that accurate sample temperature and goodsample-tip temperature calibration is critical for environ-mental testing such as this. As an example of the stabilityachieved by our system, Fig. 1b is a cooling curve showingthat the system stabilizes �2000 s after cooling is initiated.Fig. 1c demonstrates the thermal stability of the systemwhen the tip contacts the sample (in the case of the calibra-tion study, the thermocouple was the sample). A Type Kthermocouple was used for this study which is rated from�200 to 900 �C. A small deviation in the temperaturerecording can be seen at �25 s corresponding to the pointat which contact with the thermocouple was made by thediamond indenter tip. After this initial contact the temper-ature stabilized almost immediately while the tip remainedin contact with the thermocouple. It is necessary to per-form these tests under vacuum since otherwise ice will formon the specimen from the air humidity.

Cryogenic testing of the Sn micropillars was performedwhen the sample and tip stabilized at �142 �C (±2.5 �C) asmeasured by a thermocouple attached to a pure Sn samplethat was mounted on the sample platform. A flattened con-ical diamond tip with a diameter of 5 lm was used to com-press the micropillars. Sample loading was performed indisplacement control with a strain rate of 10�3 s�1. Theload and displacement data as well as SEM images werecaptured during the test.

2.3. EBSD and transmission EBSD of the micropillars

As mentioned in Section 2.1, EBSD was initially used toidentify grains for micropillar fabrication. After the micro-pillars were compressed, both at room temperature and at�142 �C, EBSD was used to characterize the orientation ofthe micropillars. Following EBSD analysis, a representa-tive room temperature and cryocompressed micropillarwas selected from each grain tested. These selected micro-pillars were lifted out of the bulk sample using an FEIQuanta Dual Beam FIB with Kleindiek nanomanipulators.Micropillars were mounted to a standard transmission elec-tron microscopy (TEM) copper half grid. A platinum layerwas deposited prior to any ion milling to protect the micro-pillar microstructure. Micropillars were thinned to electron

transparency in the FIB with a final polish of 5 keV and44 pA in order to minimize FIB damage.

The technique of transmission EBSD (t-EBSD) [17] wasutilized to characterize the microstructure of the micropil-lars. The primary advantage of this technique is theimproved spatial resolution in comparison to traditionalEBSD, which can be <10 nm [18]. t-EBSD was used tocharacterize the microstructure of both room temperatureand cryocompressed micropillars from each grain usingthe conditions outlined by Kacher et al. [18]. Analysiswas performed using Oxford/HKL data collection softwareat an accelerating voltage of 30 kV. The samples were ana-lyzed at a working distance of 5 mm using a custom-builtstage that positioned the TEM grid normal to the electronbeam. A step size of 50 nm was used for each sample. Thecollected data were used to construct inverse pole figure(IPF) maps. A low level of data interpolation was usedwith a criterion of seven neighbors for extrapolation. Wildspikes and zero solutions were also removed. All maps were

Fig. 4. The seven (001) Sn micropillars corresponding to the mechanicaldata shown in Fig. 2. (a) SEM image prior to compression, and (b) thecorresponding EBSD map (IPF Z) after all seven micropillars have beencompressed. The four micropillars on the left were compressed at roomtemperature, while the three micropillars deformed at �142C are on theright. As can be seen from the EBSD map, the cryogenically testedmicropillars have a different orientation to that of the parent grain belowafter compression.



compared with band contrast maps to verify that the noisereduction did not lead to any spurious results.

2.4. Transmission electron microscopy of micropillars

All of the micropillars that were examined with t-EBSDwere also characterized using transmission electron micros-copy (TEM) in order to verify the t-EBSD results and toanalyze the underlying defect structure. Characterizationof the lift-out samples was performed in a JEOL 3010LaB6 microscope operated at 300 kV.

3. Results

3.1. Micromechanical testing results

Fig. 2 shows the mechanical testing data for the micro-pillars compressed in the (100) orientation both at roomtemperature and at �142 �C, which is well below theDBTT of Sn. Micropillars that were compressed at roomtemperature exhibited slip, as evident in both the stress–strain behavior and slip traces that could be seen on theSEM micrographs of the micropillar surface. The cryocom-pressed micropillars exhibited a significant increase instrength and different deformation behavior consisting ofuncontrolled strain bursts. Fig. 3 shows images from repre-sentative micropillar tests from both a sample strained atroom temperature and one at �142 �C. Unlike the room-temperature micropillars, which remained in constantcontact with the diamond tip throughout the test, thecryocompressed micropillars deformed suddenly and cata-strophically and lost contact with the tip after reachingmaximum strength. This loss of contact is reflected in the

unloading curves of the cryocompressed micropillars. Theobserved difference in the mechanical response betweenthe micropillars tested at room temperature and at�142 �C suggests that a different deformation mechanismis activated above and below the DBTT. Additional char-acterization techniques such as EBSD, t-EBSD and TEMwere employed in order to identify and understand thesedifferent mechanisms.

3.2. Characterization of microcompressed micropillars

EBSD was used to characterize both groups of micropil-lars directly following compression testing. As can be seenin Fig. 4, all of the cryocompressed micropillars showed adifferent orientation than the parent grain aftercompression. Further analysis revealed that all of the cryo-compressed micropillars exhibited twinning. Each cryo-compressed (10 0) micropillar was twinned to a neworientation that was related to the base orientation by a60� rotation about the [100] direction, as can be seen inthe t-EBSD scan shown in Fig. 5. This twinning behavioris consistent with the formation of the {301} and {101}twin in Sn [20,21]. The twinning angles associated withthese two twins are similar, 62.8� and 57.2�. The 60�observed rotation results from misfits between segments,which is consistent with observations of low-angle tiltboundaries in the transmission electron microscope.However, the t-EBSD analysis of the cryocompressedmicropillar, shown in Fig. 5, clearly resolves two twinswithin the micropillar and enabled direct characterizationof each individual twin. The microstructure shown inFig. 5 has two twin orientations. The first twin, labeledas twin 1 in Fig. 5 is consistent with the {301} twin with

Fig. 5. Transmission EBSD maps of a cryogenically compressed micropillar from the (100) orientation showing the presence of two twins (Maps areshown as IPF Z, Y and X.).



a 62.9� rotation about the h0 01i, and second twin isrotated 58.9� from the first twin, also about the h001i,which is consistent with the {101} twin.

TEM analysis was performed on the micropillars inorder to verify the t-EBSD results and look for evidenceof any other underlying deformation mechanisms. Fig. 6shows a TEM micrograph of the cryocompressed micropil-lar. It can be clearly seen from the TEM micrograph thatthe twin boundaries directly correlate to the t-EBSDresults. TEM analysis also revealed the presence of low-angle tilt boundaries. Fig. 7 shows one of the twin bound-aries as indicated in Fig. 6 at higher magnification and withselected-area diffraction patterns. It is interesting to notethat the interiors of both the parent (untwinned) andtwinned grains are relatively free of defects, consistent withthe deformation being accommodated exclusively throughtwinning.

Fig. 8 shows the t-EBSD scan of a micropillarcompressed at room temperature. Here, the orientationremains unchanged from the parent grain, consistent withdislocation plasticity and an absence of twinning at roomtemperature. The observed orientation gradient across thelength of the micropillar in the EBSD map suggests anaccumulation of geometrically necessary dislocations.TEM was also performed on the room-temperature com-pressed micropillar, which confirmed the t-EBSD resultsas shown in Fig. 9. The room-temperature compressed

Fig. 6. TEM micrograph of cryocompressed micropillar 6; the boxedregion corresponds to the boundary shown in Fig. 7.

Fig. 7. (a) TEM micrograph of a twin boundary from the cryogenically compressed micropillar shown in Figs. 5 and 6. Selected-area diffraction pattern ofthe boundary (b), left of the boundary (c), and right of the boundary (d). The zone axes are labeled in (c) and (d).



micropillar did not show any evidence of twinning, ratherthe microstructure was characterized by a high density ofdislocations. Due to the high dislocation density, uponthinning the room-temperature compressed micropillarbent considerably, preventing reliable characterization ofthe Burgers vectors of the dislocations.

3.3. Characterization of micropillars from a (4�51) oriented

grain

In order to verify that the observed twinning behavior atcryogenic temperatures was not related to the highly sym-metric (100) orientation, the experiment was repeated ona (4�51) oriented grain, which is rotated �52� from the(10 0) orientation. Presumably, this orientation can be con-sidered random and would not result in any preferred axisfor twinning. Similar to the first orientation, the (4�51) ori-ented micropillars compressed below the DBTT exhibited asignificantly higher strength and very different deformationbehavior to the micropillars compressed at room tempera-ture. Fig. 10 summarizes the stress–strain behavior formicropillars compressed in the (4�51) orientation. EBSDand t-EBSD of these micropillars after compression wasperformed, and again the cryocompressed micropillarsexhibited twinning and the room temperature micropillarsexhibited dislocation slip. Similarly both twin types wereobserved in the (4�51) orientated grain. As the stress–straindata shows, for the (4�5 1) orientation both the cryocom-pressed micropillars and the room-temperature micropil-lars lost contact with the tip upon yielding anddemonstrated large strain bursts. The reason why the(4�5 1) room temperature case also lost contact with thetip upon loading in this orientation is presumably due tolarge dislocation avalanches along a limited number of sys-tems. Whereas the symmetric (100) orientation has multi-ple dislocation slip systems with similar Schmid factors, the(4�5 1) orientation has a few highly preferred slip systems.

4. Discussion

By examining the deformation behavior of Sn within asingle grain at both ambient conditions and below theDBTT of Sn, the deformation behavior can be examinedas a function of temperature. As the temperature is low-ered, dislocation motion tends to become more limiteddue to the lack of thermal assistance for dislocations to

Fig. 8. Transmission EBSD map (IPF X) of a (001) micropillarcompressed at room temperature showing that the micropillar maintainedits original orientation after compression. Even though the entiremicropillar is nominally the same orientation as it was before compres-sion, a slight color gradient can be seen corresponding to geometricallynecessary dislocations arranged in the micropillar. The boxed regioncorresponds to the region imaged in Fig. 9. (For interpretation of thereferences to colour in this figure legend, the reader is referred to the webversion of this article.)

Fig. 9. TEM micrographs of the deformed room-temperature compressed micropillar shown in Fig. 8. The sample is heavily dislocated after deformation,unlike the cryogenically compressed pillar shown in Fig. 6.



overcome the Peierls barrier. In body-centered cubic (bcc)systems the DBTT is observed at temperatures similar tothose at which plastic deformation by twinning becomesmore important than dislocation slip [22]. Deformationtwins provide high stress concentrations that have been rec-ognized as potentially important for crack nucleation [20].Similar to bcc systems, the bct structure of pure Sn exhibitsdeformation twining below the DBTT. As deformationtwining was observed in both the (100) and (4�51) orienta-tions, this suggests that this mechanism is a direct result oftemperature and not orientation.

5. Summary

In order to design engineering applications that canwithstand extreme environments, we must first understandthe underlying deformation mechanisms that can hindermaterial performance. Prior to this study, deformationmechanisms of pure Sn below the DBTT were not wellcharacterized due to a lack of suitable experimentalmethods. In this study, we have developed a novel in situcryogenic apparatus that has made it possible to systemat-ically study plasticity mechanisms down to �142 �C.Through EBSD and TEM characterization of the twinningbehavior at cryogenic temperatures, we have gained a

greater understanding of the deformation mechanisms thatare active below the DBTT.

Acknowledgments

Part of this research was carried out at the Jet Propul-sion Laboratory, California Institute of Technology, undera contract with the National Aeronautics and SpaceAdministration. A.L. was supported by a NASA GSRPFellowship and also by Boeing, Inc. We would like tothank both Hummingbird Scientific, Inc. and Hysitron,Inc. for help with the design and fabrication of thecryogenic testing apparatus. The TEM analysis wasperformed at the National Center for Electron Microscopyat Lawrence Berkeley National Laboratory, which is sup-ported by the US Department of Energy under ContractNo. DE-AC02-05CH11231.

References

[1] Frear D, Grivas D, Morris JW. J Electron Mater 1989;18:671–80.[2] Telang AU, Bieler TR, Choi S, Subramanian KN. J Mater Res

2002;17:2294–306.[3] Peng W. Microelectron Reliab 2009;49:86–91.[4] Plumbridge W. J Mater Sci: Mater Electron 2007;18:307–18.[5] Maio D, Hunt C. Microelectron Eng 2011;88:117–20.

Fig. 10. (a) Stress–strain data from the (4�51) oriented micropillars tested at room temperature (red) and �142 �C (blue). (b) t-EBSD of a (4�51)micropillar compressed at �142 �C, showing that the entire micropillar has reoriented through twinning. (c) An in situ SEM image before a twinning eventin a micropillar compressed at �142 �C, and (d) the corresponding image after a twinning event. (For interpretation of the references to color in this figurelegend, the reader is referred to the web version of this article.)



[6] Lehman LP, Xing Y, Bieler TR, Cotts EJ. Acta Mater2010;58:3546–56.

[7] Darbandi P, Bieler TR, Pourboghrat F, Lee T. J Electron Mater2013;42:201–14.

[8] Brar NS, Tyson WR. Can J Phys 1972;50:2257–64.[9] Bieler T, Telang A. J Electron Mater 2009;38:2694–701.

[10] Chalmers B. Nature 1932;129:650–1.[11] Chalmers B. Proc Phys Soc 1935;47:733–46.[12] Tu KN, Turnbull D. Acta Metall 1970;18:915–29.[13] Ratchev P, Vandevelde B, Verlinden B, Allaert B, Werkhoven D.

IEEE Trans. Comp. Pack. Tech. 2007;30:416–23.[14] Maruyama S. J Phys Soc Jpn 1960;15:1243–51.

[15] Lee SW, Meza L, Greer J. Appl Phys Lett 2013;103:101906.[16] Reinhold M, Kiener D, Knowlton WB, Dehm G, Mullner P. J Appl

Phys 2009;106. 053906-6.[17] Keller RR, Geiss RH. J Microsc 2012;245:245–51.[18] Suzuki S. JOM 2013;65:1254–63.[19] Kacher J, Elizaga P, House S, Hattar K, Nowell M, Robertson I.

Mater Sci Eng 2013;568:49–60.[20] Christian JW, Mahajan S. Prog Mater Sci 1995;39:1–157.[21] Cahn RW. Philos Mag 1954;3:363–445.[22] Hertzberg RW, Vinci RP, Hertzberg JL. Deformation and fracture

mechanics of engineering materials. 5th ed. New Jersey: Wiley; 2013.


Documents

Author's personal copy - University of California, Berkeleyaminor.mse.berkeley.edu/Publications_files/109 Acta cryo.pdfmajor constituent of most solders used by industry. Reli-ability