Originally published in SMTA International Conference Proceedings.

IMPLEMENTATION AND RELIABILITY ISSUES

WITH LEAD-FREE SOLDERS

Christopher Hunt
National Physical Laboratory
Teddington, United Kingdom
[email protected]

ABSTRACT material properties, of which the solubility of copper is an

Manufacturing high reliability lead-free circuit assemblies is important parameter.
challenging, and understanding the pitfalls and knowing
material properties is clearly desirable. This paper will There are numerous interacting factors affecting dissolution
address a wide range of issues, which include material and which include: copper type, alloy composition, fluidity,
processing properties. Examples of these phenomena temperature, flow rate, intermetallic formation, and
include: copper dissolution, tin whisker mitigation, PTH geometry issues. Therefore there is a requirement to assess
reliability, measuring solder joint reliability, conformal the susceptibility of a PCA to dissolution to a specific set of
coating protection, and tin pest. variables. Interaction of variables is important, and
temperature is one of the most complicated terms, effecting
MEASURING COPPER DISSOLUTION parameters such as fluidity, solubility and intermetallic
PCB metallised interconnect generally comprises exposed formation.
copper, and during the assembly process this copper will be
exposed to molten solder. Since the major constituent of any To characterise copper dissolution in assembly production
solder is tin, dissolution of the copper will occur. This is a there is a need to be able to measure the dissolution rate.
significant issue for product reliability, since copper can be Typically this is not easily achieved since inspection of a
eroded from specific sites during certain soldering PCA will only reveal whether copper is present or not, not
operations to the extent that electrical continuity is lost or how thick it is. Hence this project has devised a simple
reliability compromised. approach that can be used to characterise the dissolution rate
for a given soldering process.
The advent of lead-free soldering has resulted in the
introduction of a number of high tin solder alloys, which for Materials and Experimental Set Up
selective soldering applications, typically means solders One of the aims of this study is the development of a simple
with melting temperatures in excess of 220°C. The higher testing method that can be universally employed for
melting points of the lead-free alloys also dictate higher evaluating a number of variables. For this study a typical
processing temperatures, and usually higher contact times, PCB construction as used in industry was utilised, using a
as the components take longer to heat up to the higher specific design of copper pad. This is shown in Figure 1 The
processing temperature. The copper dissolution process is PCB was a 2.5 mm thick FR4, with 10 copper test pads,
temperature and solder alloy dependent. The solubility of further details of this are given in reference[1]. The test pads
copper in the new lead free alloys with tin compositions of can also be accessed from the bottom side through a hole.
at least 95% tin is potentially higher than that of tin-lead The topside shows small dimples where the copper is
solder. Hence, the higher temperature and solubility effects unsupported, and the hole can be clearly seen from the top
can significantly increase the risk of damaging copper due view. Another requirement of the testing method is an
to dissolution. Exposed surface copper can be removed, automated detection of the dissolution time. A contact for
disconnecting the land from the track. the timing probe can easily be inserted into the hole to sense
when the solder penetrates the copper foil and contacts the
The formation of a solder joint during soldering requires a pin. The soldering machine used for the tests was an ACE
reaction between the solder and the metallisation of the Automated Soldering Machine - KISS 102, using a
substrate. This reaction involves a dissolution process, travelling mini–solder fountain. Alloys with composition
which occurs through an intermediate phase of an close to those commercially available were tested here. The
intermetallic that forms at the interface. The intermetallic alloy composition was analysed using inductively coupled
itself is soluble in molten solder, and hence the intermetallic plasma – atomic emission spectroscopy (ICP-AES), and the
substrate interface proceeds into the substrate with time. The results are shown in Table 1.
nature of this intermetallic and its thickness will be
significant in controlling the overall copper dissolution Experimental Results
process. A number of alloys are now available where A flow rate of 1.35 cm3/s was selected for the tests, as it was
additions significantly below 1% affect a number of found to work well with all alloys and corresponded to a
typical solder fountain profile, as suggested by the selective
 Originally published in SMTA International Conference Proceedings.

soldering machine manufacturer. This achieved the A method was developed for comparing the dissolution rate
necessary solder fountain height, without being too of copper with different solder alloys. The method
turbulent. The correct pump speed to apply during the developed here, presents many advantages. First of all, the
dissolution tests could be calculated from the flow detection of the dissolution time was instrumented, hence
rate/pump speed relationship using linear interpolation. The improving repeatability. This method removes the need for
dissolution rate of the alloys tested at 255, 275 and 300 °C cross sectioning to measure the dissolution rate, and hence
are shown in Figure 2. the measurement process is very quick, and potentially real
time within the machine setup process. Ten copper pads can
Discussion be tested automatically in rapid succession, insuring
Controlling the flow rate during the test was seen to be constant conditions and repeatability of the results. The
important step to insure a better comparison between the procedure is also very flexible and can be adapted and used
different solder alloys. The dissolution rate, which was for different types of soldering machines and providing the
found to increase linearly with temperature, varied between same conditions are used, comparables results should be
the different alloys. This demonstrated the role that the obtainable.
composition of the alloys plays in this phenomenon. It was
observed that the SnPb alloy was not the alloy with the These features make the procedure described here an
lowest dissolution rate within the temperature interval used important tool for future testing in the copper dissolution
here (see Figure 2). area. In particular studies that investigate the effect of
different copper types could benefit from the flexibility and
The slope of the dissolution rate curves of Alloys A, B, C, ability to obtain a large number of results in a short period
F, H and SnPb were very similar. The curves for alloys D, E of time.
and SnAg had a higher slope, indicating a higher
dependence of the dissolution rate with temperature. The TIN WHISKER MITIGATION
alloys that presented the highest copper dissolution rates in A tin whisker can cause catastrophic failure of electronics,
the commonly used for soldering temperature range 275-300 yet there are no guaranteed whisker free tin finishes
°C, were alloys E, G and SnAg. Finding SnAg in this group available. The whisker formation process is not fully
was expected, because of the high solubility of Cu in understood, although there must be compressive stresses in
Sn3.5Ag [2]. Alloys E and G are very similar in the tin layer to drive the whisker formation. This is a
composition (see Table 1), so a similar behaviour was to be significant issue since the component industry makes
expected. However, it is not clear why their copper extensive use of tin finishes, and hence there is considerable
dissolution behaviour was poor. The relatively high Cu interest from high reliability users to mitigate against
content should have a positive effect compared to SnAg, but whiskering. The use of conformal coatings has been
this was not observed. The alloys that showed the lowest suggested as a means of controlling whisker growth, either
copper dissolution rates were alloys A, D and H. It is likely by inhibiting the initiation of growth or by preventing
that the good performance of alloys D and H is due to their whisker growths shorting between adjacent conductors [4-
Ni content. In the presence of Ni, complex Ni-based IMCs 8]. Work carried out by Boeing on commercially available
tend to form instead of Cu 6 Sn 5 . The Ni IMC acts as a coatings has shown suppression of tin whiskers when
superior barrier, protecting the underlying copper [3]. compared to the uncoated controls [6,7]. However, all of the
commercial coatings were eventually penetrated by
It was previously observed that the characteristics of the whiskers, indicating that these coatings cannot be depended
intermetallic layer play an important role in the dissolution on as a foolproof mitigation strategy.
rate. Cross sections of the tested samples were made and
observed using an SEM, in order to measure the IMC A test vehicle design that resembles a parallel plate
thickness. In general it was found that the thicker the capacitor has been developed. This test vehicle allows
intermetallic layer, the smaller the copper dissolution rate, whisker growth to be monitored over a large area using
as shown in Figure 3. This is consistent with the electrical detection [9]. The plates are plated with a
intermetallic acting as a barrier for copper diffusion, and chemistry that by controlling the tin thickness will either
explains why alloy A shows the lowest dissolution rate of whisker in a few days when held at room temperature, or
copper. Variations in intermetallic microstructure may also will not whisker in at least 6 months. Following plating,
influence the dissolution process and reflect the range of samples are conformally coated within 48 hrs, and then
results, but further research would be necessary to confirm folded up to form the parallel plate assembly.
this.
Four different styles of test vehicle can be fabricated as
Summary detailed below and shown in Figure 4, to investigate
The various factors that influence the dissolution of copper whisker growth in different areas:
in molten solder were investigated and the important  Type 1 (whisker out) vehicles were designed to monitor
parameters were found to be: temperature, solder for whisker growth from beneath a conformal coating.
composition and flow rate. The plates to be coated were electroplated with thin Sn
(<2µm) to generate whiskers. The uncoated plates were
 Originally published in SMTA International Conference Proceedings.

electroplated with thick Sn (>10µm) to act as a detector constant electrical monitoring (event detecting) and
plate. Interconnect Stress Testing (IST). The test vehicles included
 Type 2 (whisker in) vehicles were designed to monitor patterns to monitor changes in interconnection spacing
for whisker growth through a conformal coating from an (pitch) and also the effect of removing non-functional pads.
external uncoated source. The plates to be coated were The failure modes generated with both techniques were
electroplated with thick Sn (>10µm) to act as a detector similar as were the relative rankings of the effects. Results
plate. The uncoated plates were electroplated with thin from thermal cycling are shown in Figure 6, and this
Sn (<2µm) to generate whiskers. Weibull plot clearly shows there is a difference in the
 Type 3 (whisker in/out) were designed to monitor for performance with and without non-functional pads. The
whisker growth out from under a conformal coating and results showed that the removal of non-functional pads
then back through an adjacent coating. Both plates to be tended to improve reliability for high aspect ratio plated
coated were electroplated with thin Sn (<2µm) to through holes in thicker substrates, although increasing
generate whiskers. interconnection pitch had little effect on failure rate.
 Type 4, designed as control samples. The plates are left
uncoated and electroplated with thin Sn (<2µm) to The results generated by both thermal cycling and IST
generate whiskers. showed extremely good correlation, as presented in
Figure 7. Failures occurred at a slightly lower number of
In the preliminary investigation three coating types were cycles for IST compared to thermal cycling due to the more
investigated: paraxylene, polyurethane and acrylic. The stringent failure criteria used in IST of a 10% resistance
paraxylene coating proved to resist any whisker growth, change. The relative ranking of the level of failures is
even when it had been coated over whiskers that had just identical for both the thermal cycling and IST but the results
started to form. The polyurethane did resist whiskers were obtained in very different timescales. IST has been
growing out from underneath, but whisker were able to shown to give fast comparable results to thermal cycle
penetrate the coating. The acrylic coating was the least testing with constant monitoring, but thermal cycling may
robust, it allowed more whiskers to penetrate sooner and a be more beneficial if a wider range of experimental
whisker did penetrate out of the coating. parameters (E.g. solder joints) are to be tested
simultaneously.
So the technique was able to determine the relative ability of
a conformal coating to mitigate against tin whisker growth. MEASURING SOLDER JOINT RELIABILITY
All the control samples without conformal coatings One of the main performance criteria for solders relates to
developed whiskers of sufficient length (>250µm) to cause the low cycle fatigue performance. Many studies have been
electrical shorts. The control samples all exhibited electrical performed where circuit assemblies have been thermally
shorts within 14 days, some within as few as 3 days, cycled and the solder joint performance characterised. At
allowing for a relatively rapid i.e. less than 12 week NPL we are attempting to measure the material properties of
evaluation of the tin whisker mitigation benefits of the solder joint that takes into account size effects, using an
conformal coatings. Longer term testing would be required approach originally outlined in [12]. This is being achieved
to ensure the coated samples remained whisker free. Future using a sample with a single solder joint, as shown in
publications of this study will consider a larger range of Figure 8. The sample joint height here is 300µm, and hence
coatings. the intermetallics at the interface are important, and the
overall performance is similar to that of a surface mount
PLATED THROUGH HOLE RELIABILITY joint. The sample is mounted in a test machine to measure
Lead-free solders are all high tin alloys with significantly the stress strain cycle, where the sample temperature can
higher melting points compared to earlier tin-lead materials. also be controlled. The solder response to this displacement
Substrate technology has been developed around reinforced is shown in Figure 9, where the creep and relaxation of the
resin materials and increased degradation caused by the force is presented. The energy lost in increasing cycles of
associated higher processing temperatures is a possibility. strange range is also presented, showing the increasing size
This is compounded by the interconnecting structures being of the hysteresis loop, and the damage accumulated in the
brought into closer proximity as a result of increasing solder joint. This approach can be used to determine the
technology advances driven by miniaturisation. An earlier lifetime degradation response of the solder joint. In
study investigated laminate materials and processing issues Figure 10 the lifetime results are presented for three
[10]. The removal of non-functional pads to facilitate signal isothermal tests for a strain range of ±0.05.
routing and improved drilling conditions for high aspect
ratio vias, may also affect substrate reliability. The damage The creep, relaxation and hysteresis behaviour are all
caused to a PTH is shown in Figure 5, where barrel cracks extremely important material parameters for comparing
can be observed. solders. These material constants can also be used in
modelling, and by comparison with thermal cycling
The National Physical Laboratory and PWB Interconnect experiments the prediction of lifetime with alternative joint
Solutions Inc. have undertaken a joint study [11], following characteristics and various thermal cycle parameters,
identical test structures through both thermal cycling, with improved lifetime prediction should be possible.
 Originally published in SMTA International Conference Proceedings.

CONFORMAL COATING PROTECTION interfacial zone was kept molten for 5 seconds to allow the
Coatings are applied to protect electronics from diffusion of the alloying element into the pure tin. This
condensation and adventitious contamination effects. There generated a gradient in composition between the pure Sn
are ranges of basic coating chemistries and these can be and the alloy. Sn-Cu, Sn-Ag, Sn-Zn, Sn-Ni, Sn-Bi, Sn-In
modified to control other parameters, such as rheology. An and Sn-Pb hybrid samples were prepared for this study,
important aspect of coating performance can be evaluated where the second element added was ~0.4mass%. The seven
by measuring the electrical resistance at the interface binary alloys were tested with this procedure and ranked in
between the coating and substrate using an interdigitated terms of their propensity to transform. The change in
comb pattern. If the electrical pattern to be tested is not a electrical resistance as the transformation proceeded is
comb pattern but formed using component terminations then presented in Figure 13. The typical effect of the
coating coverage is also evaluated. Coatings’ sensitivity to transformation on the appearance of a sample before and
different contaminants can be evaluated by applying a after the transformation is presented in Figure 14, and this is
solution of the contaminant, and allowing it to dry. for SnNi, although they all appear similar.
Subsequent damp heat testing will evaluate the ability of the
contaminant to penetrate the coating by measuring a change The acceleration of the incubation/nucleation phase of the
in the underlying resistance. More details are described in transformation is not expected to occur in the field.
[13-15]. An example from such a study presented in However, this method makes it possible to study the α-Sn
Figure 11 reveals the relative performance. crystal growth in different materials in a reasonable
timescale. This study focused on binary alloys; however
Coating coverage around component termination edges is a commercial alloys are also being studied and will be the
real challenge. In Figure 12 the lack of coating coverage topic of future publications.
along the termination edge provides a weak site where
corrosion can initiate in the damp heat test. The This study has confirmed previous qualitative observations
contamination in the presence of moisture and electric bias indicating Bi to be the main element for preventing the
as facilitated the corrosion process to imitate between transformation of tin. Pb, which in previous studies was
terminations across the top of the coating. Hence damp heat believed to be as effective as Bi, did not stop the sample
testing with resistance monitoring under the coating can from transforming, but just decelerated the process. The
utilised to determine the protection capability of a coating. transformation speeds of SnAg and SnCu were similar to
pure Sn (99.99%).
TIN PHASE TRANSFORMATION (TIN PEST)
The β phase of tin is stable only down to 13oC. Below that CONCLUSIONS
temperature the thermodynamically stable phase is α tin, The implementation of the RoHS directive has clearly had a
more commonly known as tin pest. This phase has gained material impact, but in turn the new materials have had a
notoriety because of the catastrophic consequences to the number of processing impacts, namely an increase in the
transforming metal. Interest in this subject increased after processing temperature during assembly. These changes
the transition from tin-lead to lead-free solder, driven by the have been accommodated in high volume consumer
European RoHS legislation. Lead-free alloys contain from markets, but for high reliability and high stress
95 to 99% tin and could potentially be susceptible to the β/α environments the reliability case is not yet proven.
transformation. The transformation takes a long time to Processing issues such as copper dissolution can be solved
occur, months or even years, especially in alloys of Sn. For by machine design modifications and minor elemental
this reason it is particularly difficult to observe. At NPL a modification to the alloys to reduce copper solubility. The
procedure for observing and measuring the transformation PCB substrate is pushed very heard during the multiple
based on electrical measurements has been developed [16]. higher temperature reflow passes. This can impact on the
Furthermore a sample preparation technique that PTH reliability, can cause delamination and increase
considerably accelerates the transformation in alloys is used susceptibility to CAF. Fatigue of solder, both low and high
based on the introduction of a seed. cycle fatigue, are still areas that are being investigated, and
prediction and defining acceleration factors have yet to be
A seeded sample was prepared by immersing 5 ± 0.5 mg of defined. Clearly for safety critical equipment a high level of
CdTe powder into 1gm of molten Sn, giving samples confidence will be required for these acceleration factors.
typically 40mm long and 1.8mm radius. The sample was While tin pest remains an issue more of an academic
solidified in an oxidised steel mould in the shape of a small interest, that alloys of tin can transform if nucleated is of
cylinder, with the inoculated part at one extremity. The concern. The question that needs to be answered is how
second part of this sample consisted of 1gm of a Sn alloy, does nucleation initiate in this system and what are the
for example Sn-Cu. The two parts were placed next to each environmental factors. Tin whiskers remains an ongoing
other in the mould and the two parts were fused together, issue as no plated tin finish is guaranteed as whisker free,
melting the interfacial region, and allowing any dross to hence the interest in mitigation strategies and whether
appear at the surface and be dispersed. Tweezers were used coatings can mitigate against a whisker short. The test
to gently push the samples together while the interfacial approach developed here will facilitate coating
zone was molten. When a joint had been formed the developments for this purpose. Finally symptomatic of an
 Originally published in SMTA International Conference Proceedings.

increase in electronics reliability across the board there is [12] Dusek, M., & Hunt, C.P., (2006), “Test Approach to
increasing interest in applying coatings, and the work Isothermal Fatigue Measurements for Lead-free
presented here shows opportunities for improved evaluation Solders”, DEPC-MPR 048, March 2006
techniques for coatings. [13] Mensah, A, and Hunt, C.P., “The Role of
Permeability and Ion Transport in Conformal
ACKNOWLEDMENTS Coating Protection”, NPL Report DEPC-MPR 032,
This work was funded by the UK government out of their September 2005
National Measurement Office. The author is grateful to the [14] Zou, L., Hunt, C.P., “Protection Performance of
following: Davide Di Maio, Martin Wickham, Ling Zou and Conformal Coatings in Harsh Environments”,
Owen Thomas. DEPC-MPR 054, July 2006
[15] Zou, L., Hunt, C.P., “Test Method for Conformal
REFERENCES Coating Protection Performance of Electronic
[1] Di Maio, D., Hunt, C.P., (2009), “Measurements of Assembly in Harsh Environments”, DEPC-MPR 060,
Copper Dissolution in Lead-Free Alloys”, NPL March 2007
Report, MAT 26 ISSN: 1754-2979, National [16] Di Maio, D., Hunt, C.P., (2008), “Investigation
Physical Laboratory, Teddington, UK. Methods of the β to α Tin Allotropic
[2] Yu, D.Q., Wu, C.M.L., Law, C.M.T., Wang, L. and Transformation”, NPL Report, MAT 21 ISSN: 1754-
Lai, J.K.L., (2005), “Intermetallic compounds growth 2979, National Physical Laboratory, Teddington, UK
between Sn-3.5Ag lead-free solder and copper
substrate by dipping method”, Journal of Alloys and
Compounds Vol. 392.
[3] Izuta, G., Tanabe, T., Suganuma, K., (2007),
“Dissolution of Sn-Ag-Cu System Lead-Free
Solder”, Soldering & Surface Mount Technology,
Vol. 19 No. 2, pp. 4-11.
[4] Kadesch, J.S., Leidecker, H.; Effects of Conformal
Coat on Tin Whisker Growth; Proceedings of the
37th IMAPS Nordic Conference, September 2000
[5] Kadesch, J.S., Brusse J.; The Continuing Dangers of
Tin Whiskers and Attempts to Control Them with
Conformal Coating; NASA EEE Links Newsletter
July 2001
[6] Woodrow,T.; Evaluation of Conformal Coatings as a
Tin Whisker Mitigation Strategy; IPC/JEDEC 8th
International Conference on Pb-Free Electronic
Components and Assemblies, San Jose, CA, April
18-20, 2005
[7] Woodrow,T.; Evaluation of Conformal Coatings as a
Tin Whisker Mitigation Strategy, Part 2; SMTAI,
Sept. 2006
[8] Nishimi, K.; Space Shuttle Program - Tin Whisker
Mitigation; Intl. Symposium on Tin Whiskers, Apr.
2007
[9] Wickham, M., Hunt, C.P., (2008), “Test Method for
Measurement of the Propensity for Conformal
Coatings to Inhibit Tin Whiskering”, NPL Report,
MAT 28 ISSN: 1754-2979, National Physical
Laboratory, Teddington, UK.
[10] Wickham, M., Dusek, M., & Hunt, C.P., (2007),
“Reliability of Electronic Substrates After Processing
at Lead-free Soldering Temperatures”, NPL Report,
MAT 10 ISSN: 1754-2979, National Physical
Laboratory, Teddington, UK.
[11] Hunt, C.P., and Wickham, M., (National Physical
Laboratory), & Birch, W., and Furlong, J., (PWB
Interconnect Solutions Inc., Nepean, Canada), The
Effect of Functional Pads and Pitch on Via
Reliability using Thermal Cycling and Interconnect
Stress Testing, APEX 2009, S17-02
 Originally published in SMTA International Conference Proceedings.

Table 1: Composition of the alloys used for the tests (weight %).
The remaining contribution is from tin
% Ag Bi Cu Co Ni Pb Sb
A 0.050 <0.01 0.626 0.030 <0.01 0.040 0.020
B 0.024 <0.01 0.510 0.081 <0.01 0.005 0.008
C 2.520 <0.01 0.748 <0.01 <0.01 0.030 0.700
D <0.01 <0.01 0.660 <0.01 0.050 0.018 0.017
E 0.338 <0.01 0.594 <0.01 <0.01 0.006 0.011
F 3.700 <0.01 0.730 <0.01 <0.01 0.015 0.017
G 0.301 0.090 0.682 <0.01 <0.01 0.016 0.017
H 0.210 <0.01 0.690 <0.01 0.040 0.030 0.010
SnPb <0.01 <0.01 <0.01 <0.01 <0.01 60.340 0.010
SnAg 4.100 <0.01 <0.01 <0.01 <0.01 0.012 0.018

Figure1: The bottom and top side of the PCB (left and right respectively) coupon used for the tests.
The holes on the bottom side correspond to the unsupported part of the copper pads seen on the top side.
 Originally published in SMTA International Conference Proceedings.

Figure 2: The dissolution rate of the alloys at the temperatures of 255, 275 and 300 °C.
 Originally published in SMTA International Conference Proceedings.

Figure 3: Dissolution rate of copper in the solder alloys and IMC thickness relationship (275 °C).

Cu/ThinSn/CC_ThickSn/Cu Cu/ThickSn/CC_ThinSn/Cu
Type 1: Monitors whisker penetration Type 2: Monitors whisker penetration
out from under coating in through coating
Plate 1: Thin Sn, conformally coated Plate 1: Thick Sn, conformally coated
Plate 2: Thick Sn, no coating Plate 2: Thin Sn, no coating

Cu/ThinSn/CC_CC/ThinSn/Cu
Cu/ThinSn_ThinSn/Cu
Type 3: Monitors whisker penetration
Type 4: Control
out from under coating and in
thorough coating Plate 1: Thin Sn, no coating
Plate 1: Thin Sn, conformally coated Plate 2: Thin Sn, no coating
Plate 2: Thin Sn, conformally coated

Figure 4: Schematics of the four types of test vehicle arrangement for tin whisker mitigation evaluation
 Originally published in SMTA International Conference Proceedings.

Figure 5: Microsection of via with failure in region where non-functional pads have been retained.

Figure 6: Weibull plot of thermal cycle failures for HFFNA2T (non-function pads included)
and HFFMA2T (non-functional pads removed)
 Originally published in SMTA International Conference Proceedings.

1000
Thermal Cycling
900
IST
800

700
Cycles to failure 600

500

400

300

200

100

0
Fine Grid Coarse Grid Pads In Pads Out

Figure 7: Comparison of number of cycles to 50% failures for thermal cycling and IST

Figure 8: Single solder joint fatigue sample

0.09 20

15
0.06
10
Shear stress [MPa]
Shear strain [rad]

0.03
5

0 0

-5
-0.03
-10
-0.06
-15

-0.09 -20
0 2000 4000 6000 8000 10000
Time [s]

Figure 9: (a) Shows the stress response to the imposed strain profile, where the creep and relaxation can be clearly
seen. (b) Shows the increasing hysteresis loop, and the energy loss in the solder joint during the cycle.
 Originally published in SMTA International Conference Proceedings.

Figure 10: Lifetime degradation of SAC305 tested isothermally at 30, 60 and 125ºC, for ±0.05 strain

11
Surfactant
No contaminant
10
Final Log SIR (ohm)

5
y

te

)
ay

lic

e
p

p
(1

(2
ox

in

an

on
di

di
la

ry
pr

ic
at
lic
Ep

e
e
th

lic
ac
) -S

al
ac
co

bl
bl
ry

Si
r

ou
ou
ur

ac
ed
ro
ac
(1

ly
N

uo

) -D

D
lic

as

ed
ed

Po

e-
ra

Fl

t-b

(2
as

on
as
ac

ic
en

-b
-b

lic
l
ra
ed

er
lv
er

Si
ac
So

at
as

at

W
W
-b

d
se
er

a
at

-b
W

er
at
W

Coating
Figure 11: The relative performance of different coatings with a surfactant contaminant
 Originally published in SMTA International Conference Proceedings.

Figure 12: A dendrite forming on top of the coating

Figure 13: Electrical resistance change of Sn alloys during the transformation at -35°C.
 Originally published in SMTA International Conference Proceedings.

Figure 14: Photos of a simple hybrid Sn-Ni sample before and after the transformation.
The second photograph was taken after storing the sample for approximately 1 month at -35°C.