7 Lead (PB) - Free Solders For High Reliability and High-Performance Applications

191
Lead (Pb)-Free Solders for High Reliability and

High-Performance Applications
Richard J. Coyle
Nokia Bell Laboratories, Murray Hill, NJ, USA
7.1 Evolution of Commercial Lead (Pb)-Free Solder

Alloys
7.1.1 First Generation Commercial Pb-Free Solders

The European Union RoHS Directive (Restriction on the use of hazardous sub-
stances in electrical and electronic equipment) that drove the industry conversion
to lead (Pb)-free manufacturing was implemented in 2006 [1]. However, before
implementation became required by law or by market pressure, multiple national
and international research projects were formed between 1991 and 2006 in the
United States, the European Union, and Japan to examine Pb-free alternatives
to eutectic tin-lead (63Sn37Pb) solder [2]. Among the most-referenced studies
are the NCMS (US), IDEALS (Europe), DTI summary report (UK), JEITA and
NEDO (Japan), and iNEMI (international) [3–8]. Handwerker et al. [2] provide a
summary of the results from these studies, albeit some of the findings are dated
based on the current understanding of solder alloying effects and advances in
microstructural analysis.
The NCMS project [8] was one of the earliest of these investigations but consid-
ered one of the most influential because it set the industry direction for develop-
ment of Pb-free solder alloys. The NCMS alloy screening methodology considered
toxicology, economics, manufacturing, and reliability. Because the team consisted
primarily of high reliability end users, thermomechanical or thermal fatigue reli-
ability was emphasized. Thermal fatigue requirements always have been a prior-
ity for the products of many high reliability end users [9]. Solder joints age and
degrade during service and eventually fail by the common wear-out mechanism
Lead-free Soldering Process Development and Reliability, First Edition. Edited by Jasbir Bath.
© 2020 John Wiley & Sons, Inc. Published 2020 by John Wiley & Sons, Inc.
192 7 Lead (Pb)-Free Solders for High Reliability and High-Performance Applications
Table 7.1 Composition and melting range of high-Ag, near

eutectic first generation solder alloys.
Alloy Nominal composition (wt%) Melting range (∘ C)

Sn Ag Cu
SAC387 95.5 3.8 0.7 217

SAC405 95.5 4.0 0.5 217–225
SAC305 96.5 3.0 0.5 217–221
SAC396 95.5 3.9 0.6 217–221
SnAg 96.5 3.5 0.0 221
of thermally activated solder fatigue (creep fatigue) [10]. Solder fatigue is the major
wear-out failure mode and major source of failure for surface mount technology
(SMT) components in electronic assemblies [11].
In work following the NCMS study, the high-Ag alloys given consideration were
the ternary eutectic (Sn3.8Ag0.7Cu), hypereutectic (Sn4.0Ag0.5Cu), hypoeutectic
(Sn3.0Ag0.5Cu), hypereutectic NEMI (Sn3.9Ag0.6Cu), and binary eutectic
(Sn3.5Ag) as shown in Table 7.1. This high-Ag family of SnAgCu alloys are known
as SAC alloys and now are referred to as first generation commercial Pb-free alloys.
The implementation of overmolded ball grid array (BGA) technologies coin-
cided approximately with the start of the NCMS study, but these early BGA
packages could not withstand elevated temperature Pb-free soldering, thus were
not included in that study. Later studies, however, confirmed the beneficial effect
of high Ag content on thermal fatigue reliability. Results from thermal cycling
tests of flip chip and BGA packages as a function of Ag content are shown in
Figures 7.1–7.3.
A multi-year program conducted by the IPC Solder Product Value Council
(SPVC) lead eventually to the universal adoption of SAC305 (Sn3.0Ag0.5Cu) in
manufacturing. The comprehensive SPVC program included differential scanning
calorimetry (DSC) melt analysis, solderability testing, visual and X-ray inspection
of solder joint voids, thermal cycling and thermal shock tests for reliability
analysis, and metallurgical analysis of time zero and failed solder joints [15].
7.1.2 Second Generation Commercial Pb-Free Solders

Since the implementation of the European Union RoHS Directive and intro-
duction of those first generation commercial Pb-free alloys, there have been
significant innovations in Pb-free solder alloy formulations. The relatively short
transition period to Pb-free solders has not allowed sufficient time for systematic
7.1 Evolution of Commercial Lead (Pb)-Free Solder Alloys 193
Figure 7.1 Thermal cycling 100

test data from Terashima Thermal Cycle:
et al. showing the direct 233-398 °k (–40/125°C)
relationship between Ag 80 Sn-xAg-0.5Cu
content and thermal fatigue Cu pad
lifetime. The test vehicle was
Failure Rate, F (%)

flip chip (first level)
interconnects [12]. Copyright 60
2003 by Springer Nature.
Used with permission.
40
1Ag
20 2Ag
3Ag
4Ag
0
0 200 400 600 800
Thermal Cycling, N (Cycle)
680 I/O PBGA Thermal Fatigue Reliability

SAC305, SAC105, SAC405, and 63Sn37Pb
0/+100°C temperature cycling, 30 Minute Dwell Times
99
% w/rr
90
P SAC405 outperforms
a 70 SAC305, All Pb free
c 50 alloys outperform the
k SnPb eutectic alloy
a
30 SAC
g
e 405
s
SAC
10
F 105 SAC
63Sn37Pb Eta Beta n/s
a 305
5761 13.2 30/4
i
l 4159 21.6 30/3
e 2709 20.5 20/0
d 6436 19.9 30/1
1
1000 10000
Temperature Cycles
Figure 7.2 Thermal cycling test data from Coyle et al. showing the direct relationship
between Ag content and thermal fatigue lifetime. The test vehicle was a 35 mm body ball
grid array [13].
99
Hard open
90
SAC105
50
Failures (%)
10
5.0 SAC305
Sn-3.5Ag
1.0
1000 10,000
Cycles
Figure 7.3 Thermal cycling test data from Henshall et al. showing the direct
relationship between Ag content and thermal fatigue lifetime. The test vehicle was a ball
grid array [14]. Figure used with permission of IPC International, Inc.
alloy design and comprehensive testing [16]. Consequently, alloy development

has been driven primarily by experience gathered through volume manufac-
turing and increased deployment of a variety of Pb-free products of increasing
complexity. This experience has resulted in an increased number of Pb-free solder
alloy choices beyond the first generation near-eutectic SAC alloys that were
established initially as replacements for eutectic SnPb [17]. Second generation,
lower Ag alloys have been developed and introduced to address the shortcomings
of the first generation near-eutectic SAC alloys, such as poor mechanical shock
performance, higher cost, and copper erosion and plated-through-hole (PTH)
damage during wave soldering and repair and rework operations [9, 17].
Failures due to mechanical loading in high-volume handheld consumer
products provided a significant motivation for development of second generation
solder alloys. Second generation commercial alloys typically are characterized
by lower Ag content and microalloy additions to improve mechanical properties
of the bulk solder or the intermetallic interfaces [17]. Figure 7.4 illustrates the
improved mechanical drop test performance of lower Ag alloys compared to the
7.1 Evolution of Commercial Lead (Pb)-Free Solder Alloys 195
99
90
80
70
60
50
40
30
Percent
20
SAC405
10
5
3 SAC105
LF35
2
1
0.1 1.0 10.0 100.0 1000.0
Minitab Drops to Failure
Figure 7.4 Data of Kim et al. showing cumulative failures versus number of drops to
failure for high silver Sn4.0Ag0.5Cu (SAC405), low silver Sn1.0Ag0.5Cu (SAC105) and
microalloyed low silver Sn1.2Ag0.5Cu+Ni(LF35) [18].
Figure 7.5 Data of Syed 350

1st Failure
# of Drops to Failure
et al. showing improved 300

1st failure drop test Characteristic Life
250
performance of a 0.8 mm
pitch ball grid array (BGA) 200
with lower Ag content and 150
microalloy additions [19]. 100
50
0
SAC305 SAC125Ni SAC101Niln SAC105
Solder Ball Alloys
higher Ag SAC405 [18], and Figure 7.5 shows that the first failure in drop testing
is delayed significantly in alloys with lower Ag content [19].
The advent of second generation commercial Pb-free alloys stimulated research
and development efforts to fill the gap in knowledge associated with thermal
fatigue resistance of first- and second generation alloys. Much of this work has
been sponsored by industrial consortia such as the International Electronics
Manufacturing Initiative (iNEMI), High Density Package User Group (HDPUG),
the Center for Advanced Life Cycle Engineering (CALCE) at the University of
Maryland, Universal Advanced Research in Electronics Assembly and the Center
for Advanced Vehicle and Extreme Environment Electronics (CAVE3 ) at Auburn
University [20–24]. The iNEMI Alloy Alternatives Characterization Program has
been particularly active in publishing results from thermal fatigue studies of these
two initial generations of Pb-free solder alloys [9, 17, 25–38].
7.1.3 Third Generation Commercial Pb-Free Solders

Lead-free solder alloy development is continuing to evolve with the emergence
of third generation commercial alloys designed to provide additional alternatives
as Pb-free manufacturing becomes pervasive, designs continue to evolve in com-
plexity, and operating environments become increasingly more aggressive. The
impetus for development of third generation Pb-free solders has been the dramatic
increase in electronic content in automobiles. Automotive electronic assemblies
must perform in environments characterized by increasing temperatures, ther-
mal and power cycling, vibration, and thermal and mechanical shock. Automotive
electronics no longer can be characterized simply as “under the hood.”
Many automotive control modules, sensors, and components are mounted in
areas that experience high operating temperatures, and rapid thermal and power
cycling, in combination with vibration and shock [39]. Figure 7.6 illustrates
locations of sensors and control modules and the anticipated range of aggressive
thermal exposures [40, 41]. Software and electronics are now considered core
competencies of automotive manufacturing, and this is driving innovation and
an increase in electronic content in automotive. Figure 7.7 offers two perspectives
on the cost of automotive electronics relative to total car cost with Figure 7.7a
showing various areas of innovation driving the increase in electronic content.
The projected growth of electric vehicles shown in Figure 7.8 is expected to
accelerate the growth of electronic content. Furthermore, Figure 7.9 shows the
compound annual growth rate (CAGR) for automotive electronics is twice that
of the consumer electronics market [42]. The prevailing opinion is that common
SAC alloys cannot satisfy the reliability requirements for these applications and
use environments.
7.2 Third Generation Alloy Research and Development

7.2.1 Limitations of Sn-Ag-Cu Solder Alloys
Solder alloys based on the SnAgCu system (SAC) are more resistant to thermal
fatigue than the eutectic SnPb alloy, but they have reliability limitations at
higher operating temperatures [43] such as those illustrated in Figure 7.6. During
solidification of SAC solders, the Ag and Sn react to form networks of Ag3 Sn
precipitates at the primary Sn dendrite boundaries [44, 45]. These intermetallic
precipitates are recognized as the primary strengthening mechanism in SAC
Combustion Chamber < 500 °C
129 °C lgnition
• Pressure Sensors Engine Compartment Surface 150 °C Alternator
Engine Compartment < 150 °C
Close to Engine 120 °C Surface
• Power Train Control
• Motor Control
• Transmission Control Engine Compartment
Remote from Engine 105 °C
Engine - 140 °C
° to
90 °
0
143°
11 Engine Surface
140°C Exhaust
38°
System 587 °C
90°
to
110
°
Exhaust System < 800 °C
• Exhaust Sensors
Exterior - Accessible
Wheel Mounted Components < 300 °C Engine Oil 148 °C
Engine Transmission < 200 °C to Splash, etc. 70 °C
• Engine mounted ECUs • Brake-by-Wire Transmission Oil 148 °C
• Integrated TCUs • Steer-by-Wire
(a) (b) Road Surface 66 °C
• Shift-by-Wire
Figure 7.6 Illustrations of (a) sensor and electronic control module locations and anticipated thermal exposures [40], and (b) an engine
compartment thermal profile [41].
?
60%
Advanced Driver Assistance 50%
Active-Passive Safety
Green Powertrain 50%
50%
Electronic cost as % of total car cost
Radar/Vision
Telematics
AUTOMOTIVE ELECTRONICS
Infotainment
SHARE OF TOTAL COST
40%
Airbag
35% 35%
ABS/ESP
Body Electronics 30% 30%
Multiplexing 30%
22% 20%
20%
Electronic 15%
15%
Fuel injection
10% 10%
10%
3% 4%
5% 1%
2.5% 0%
1950 1960 1970 1980 1990 2000 2005 2010 2030? 1950 1960 1970 1980 1990 2000 2010 2020 2030
(a) ©2020 NXP B.V. (b) Statista: https://www.statista.com
Figure 7.7 (a) Areas of innovation that are driving the increase in electronic content Source: Courtesy/Used with permission of EDN
Network and NXP and, (b) the projected growth of automotive electronics and impact on cost per car. Source: Courtesy/Used with permission
of Statista Inc.
7.2 Third Generation Alloy Research and Development 199
Figure 7.8 Data

projecting the growth of EXPECTED GROWTH OF ELECTRIC VEHICLE SALES
electric vehicle sales.
Source: Courtesy Roskill Hybrid electric
Information Services. 25 Million
Plug in hybrid electric
20 Million Electric vehicle
Commercial
15 Million
10 Million
5 Million
0 Million
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
Roskill Information Services: https://roskill.com
×2 Aut
omotive industry
$2.3T
CAGR + 2.7%
electronics indus
nsu
mer try 96M units - ASP $24,000
Co
$1.2T
%
CAGR + 4.4
its
3,046M un
ASP $374
Automotive electronics $142B CAGR + 7%

96M units - ASP $1,614
$3.7B
Automotive packaging CAGR + 11%

27.4B units
Yole Development: http://www.yole.fr
Figure 7.9 The need for advanced technologies is driving increased automotive
electronic content with a compound annual growth rate (CAGR) of 7%, which is twice the
CAGR for the consumer electronics market [42].
solders [12, 33, 44–46]. During thermal or power cycling and extended high
temperature exposure, the Ag3 Sn precipitates coarsen and become less effective
in inhibiting dislocation movement and slowing damage accumulation. Even-
tually, this leads to recrystallization of Sn grains in areas of stress concentration
during thermal aging or thermal cycling [47–50]. This pattern of microstructural
evolution is characteristic of the thermal fatigue failure process in these Sn-based
10 μm
10 μm
SAC305 as solidified SAC305 after thermal cycling
Figure 7.10 Scanning electron micrographs illustrating Ag3 Sn intermetallic precipitate

coarsening that precedes recrystallization and fatigue crack propagation during thermal
cycling of SAC305 [39].
Pb-free alloys and was described originally in detail by Dunford et al. [51].
Figure 7.10 shows scanning electron micrographs illustrating coarsening of the
Ag3 Sn precipitates in SAC305 solder caused by thermal cycling. Figure 7.11 illus-
trates the combined effects of strain and temperature on precipitate coarsening
and fatigue damage during thermal cycling.
Figure 7.11 A scanning

electron micrograph
showing the accelerated
intermetallic precipitate
coarsening in the
strain-localized region of
a SAC305 BGA sample
after 0/100 ∘ C thermal
cycling. The combination
of strain (𝜀) and
temperature in this region
promotes recrystallization
and fatigue crack
propagation. In the
absence of higher strain,
coarsening is much slower
as evidenced by the
smaller particles and
higher particle density in
the region adjacent to
(below) the crack [35].
Figure 7.12 Data from the Characteristic Lifetime, SAC305

iNEMI 3rd Generation Alloy test 8000
program demonstrating the 7000
Thermal Cycles
dramatic reduction in 6000
characteristic lifetime of a 5000
SAC305 BGA as the thermal 4000
cycling profiles become more 3000
2000
aggressive [52].
1000
0
0/100 °C –40/125 °C –55/125 °C
Thermal Cycling Profile
Figure 7.12 is a bar chart comparing the thermal cycling performance of SAC305
BGA tested with 0/100 ∘ C (TC1), −40/125 ∘ C (TC3), and −55/125 ∘ C (TC4) ther-
mal profiles as defined by the IPC-9701 attachment reliability guidelines [53]. The
chart illustrates the dramatic reduction of 60–70% in characteristic lifetime of the
SAC305 solder when tested with the more aggressive thermal cycling profiles. The
TC3 and TC4 test profiles are required to qualify products that will be deployed
into automotive and military/defense, and avionic use environments.
While the economic motivation for third generation Pb-free alloy development
clearly is coming from the automotive sector, high reliability electronic challenges
exist in medical, military/defense, and avionics. Thermal and mechanical require-
ments in aerospace/defense applications are similar or more demanding than
automotive, and product lifetimes typically are greater. The aerospace/defense
industry has increasing pressure from its supply chain as they continue in their
effort to maintain traditional SnPb components and manufacturing. Figure 7.13
shows the small market share for electronics in this sector, which provides
minimal leverage in cost or availability of components and hence, the latest
Pb-free Electronics Operating Environment vs. Operational Life-time1

(Humidity, Temperature, Shock, Vibration)
C Medical Satellites D
Harshness of Service Environment
High Equipment Missiles
Emerging
Consequences of Failure
A&D Risk Reduction

Project Focus Aircraft
Cars Challenges
8% <1%
62% Network Servers Industrial 29%

Products
Cell Phones
Major Home
Desktop PCs
Appliances Electronics
Low Market Share
A B
1
1 3 5 10 20 30
IPC PERM Council
Expected Operational Service Life (Years)
Figure 7.13 Aerospace/defense applications are characterized by harsh use

environments and long operational service lifetimes. The aerospace/defense sector has
minimal influence on the global supply chain due to its small market share.
technologies [54]. Because of concerns regarding the impact of technology

limitations on designs, the aerospace/defense sector, in conjunction with the IPC
PERM Council has begun to explore steps needed for Pb-free conversion in that
sector [55].
7.2.2 Emergence of Commercial Third Generation Alloys

7.2.2.1 The Genesis of 3rd Generation Alloy Development
Early in the twenty-first century, as it appeared certain that the European Union
RoHS Directive [1] would proceed to implementation, a task group formed of
forward-looking solder suppliers, end users, and academic researchers to develop
a commercial Pb-free alloy that could meet the performance challenges of higher
temperature automotive applications [56–61] The investigation that ensued was
known initially as the hotEL (hot ELectronics) Project and more commonly as
the Innolot Project. After several years of metallurgical experimentation and
mechanical and thermal testing, an alloy was developed based on the SAC387
(Sn3.8Ag0.7Cu) nominal ternary eutectic composition, with alloying additions of
3.0 wt% bismuth (Bi) and 1.5 wt% antimony (Sb) substituted for Sn, along with
a microalloy addition of 0.15 wt% nickel (Ni). This complex alloy was identified
by the project code name 90iSC and idiomatically as the six-part alloy since its
composition consisted of tin (Sn) plus five alloying elements. The metallurgical
concept of substituting elements such as bismuth (Bi), antimony (Sb), and indium
(In) for Sn had been explored extensively in the original NCMS study [3], but this
alloy development is regarded as the first commercial, third generation Pb-free
alloy [60], with a European patent issued [62].
7.2.2.2 An Expanding Class of 3rd Generation Alloys

Third generation high reliability solder alloys may contain significant major alloy
and microalloy additions to promote better high temperature performance. Most
of these alloys are based on the Sn-Ag-Cu (SAC) system, although a few alloys
are based on the Sn-Cu system and do not contain Ag. Harsh environment appli-
cations have various requirements for resistance to damage from high strain rate
shock and vibration loading, in addition to the requirement for superior resistance
to thermal fatigue damage [63]. The need for higher overall reliability performance
has driven the development of an expanding class or group of commercial third
generation, high reliability solder alloys [39, 52].
This new group of solder alloys is so different than conventional solder alloys
that it can be considered an emergent class of specialty or special-purpose
materials. These alloys are proprietary formulations and typically trademarked,
patent-protected, or held as a trade secret. A list of the names and compositions
of many of these commercial alloys is shown in Table 7.2. A notable metallurgical
Table 7.2 A list of the trade names, alloy developers, and chemical
compositions of third generation, commercial high reliability solder alloys.
Trade names and nominal composition (wt%) of high reliability solder alloys
Alloy Developer Sn Ag Cu Bi Sb In Other
405Y Inventec 95.5 4.0 0.5 0.05 Ni; Zn
Cyclomax (SAC-Q) Accurus 92.8 3.4 0.5 3.3
Ecalloy Accurus 97.3 0.7 2.0 0.05 Ni
HT1 Heraeus 95.0 2.5 0.5 2.0 Nd
Indalloy 272 Indium 90.0 3.8 1.2 1.5 3.5
Indalloy 277 Indium S9.0 3.8 0.7 0.5 3.5 2.5
Indalloy 279 Indium 89.3 3.8 0.9 5.5 0.5
Innolot Heraeus 91.3 3.8 0.7 3.0 1.5 0.12 Ni
LF-C2 Nihon 92.5 3.5 1.0 3.0
M794 Senju 89.7 3.4 0.7 3.2 3.0 Ni
M758 Senju 93.2 3.0 0.8 3.0 Ni
MaxRel plus Alpha 91.9 4.0 0.6 3.5
PS48BRa) Harima Bal. 3.2 0.5 4.0 3.5 Ni, Co
REL22a) AIM Bal. 3.0 0.7 3.0 0.6 0.05Ni; other
REL61a) AIM Bal. 0.6 0.7 2.0
SB6NX Koki 89.2 3.5 0.8 0.5 6.0
SN100CV Nihon 97.8 0.7 1.5 0.05Ni
SN100CW1 Nihon 95.8 0.7 1.5 2.0
Violet Indium 91.25 2.25 0.5 6.0
Viromet 347 Asahi 88.4 4.1 0.5 7.0
Viromet 349 Asahi 91.4 4.1 0.5 4.0
a) Nominal values; actual composition proprietary.
parallel in alloy development is found with the class of special-purpose materials

that includes specialty stainless steels and heat-resistant alloys (superalloys) [64].
7.2.3 Metallurgical Considerations

The addition of Ag strengthens Sn and improves the creep resistance of the SAC
solder by precipitation hardening as illustrated in Figure 7.14. The addition of
other alloying elements can improve the creep resistance of the solder by means
of two other well-known metallurgical strengthening mechanisms, solid solution
hardening and dispersion hardening. The introduction of solute atoms into solid
Small Solute atom Large Solute atom Figure 7.14 A simple schematic
illustrating lattice distortion due
to substitutional solute atoms.
β-Sn lattice
Solid Solution Dispersion Figure 7.15 A simple schematic

Strengthening Strengthening comparing solid solution (left) and
dispersion strengthening (right).
solution of a solvent-atom lattice invariably produces an alloy that is stronger than

the pure metal [65]. Figures 7.14 and 7.15 show a simplified schematic illustration
of substitutional solid solution strengthening. Substitutional or interstitial solute
atoms strain the lattice and dislocation movement or deformation is inhibited
by interaction between dislocations and solute atoms incorporated into the β-Sn
lattice.
Even when solute atoms precipitate from solution during thermal excursions in
service, the solder alloy may strengthen due to subsequent dispersion hardening.
Dispersion strengthening occurs when insoluble particles are finely dispersed
in a metal matrix. Typical dispersion-strengthened alloys employ an insoluble,
incoherent second phase that is thermally stable over a large temperature
range (Figure 7.15) [66]. For Sn-based SAC solder alloys, the strength could be
derived from a combination of increased solid solution strengthening at higher
temperatures due to increased solubility, and dispersion strengthening that would
supplement the solid solution effect at lower temperatures where solubility has
decreased.
The commercial development of the Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni (Innolot)
alloy provides evidence that substitutional solid solution strengthening can
improve resistance to creep and fatigue at higher temperatures in Sn-based,
Pb-free solders. Figure 7.16 shows that this alloy outperforms a SnAgCu (SAC)
alloy in a combined thermal cycling and shear test [61]. The current working
Better Reliability Than SAC Alloy

SnAgCu 1206 SnAgCu 0805 SnAgCu 0603 SnAgCu 0402
130 Innolot 1206 Innolot 0805 Innolot 0603 Innolot 0402
120
110
100
90
Shear force [N]
80
70
60
50
40
30
20
10
0
0 50 100 250 500 1000
Source: HOTEL project
Number of Cycles
–40°C/+165°C, 30 minute dwells
Figure 7.16 Data of Miric showing Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni (Innolot) outperforms

SnAgCu (SAC) in a combination thermal cycling/shear test [61]. The test vehicles were
four different size chip resistors.
hypothesis is that solid solution and dispersion strengthening not only can sup-
plement the Ag3 Sn precipitate hardening found in SAC solders, but continue to be
effective once precipitate coarsening reduces the effectiveness of the intermetallic
Ag3 Sn precipitates [67].
The elements proposed most commonly for improving high temperature proper-
ties in third generation high reliability solders are bismuth (Bi) and antimony (Sb).
The element indium (In) is used to a lesser extent, at least partly due to its high cost.
The elements Bi and In, when used as major alloying elements (not as microalloy-
ing), also reduce the melting point of most solder alloy formulations, while the
addition of Sb tends to increase the melting point [61, 68]. These modified SAC
alloys have off-eutectic compositions, are characterized by non-equilibrium solid-
ification, and often have significant melting or so-called pasty ranges [2, 3, 69, 70].
Although many third generation Pb-free solders have been commercialized, the
concept of using major element alloying to improve mechanical properties or to
alter melting behavior is not novel. Formulations incorporating Bi, Sb, and In into
basic Sn-Ag or Sn-Ag-Cu eutectics were studied by the NCMS consortium of indus-
trial partners in 1997 [3] and the properties were documented by NIST and the
Colorado School of Mines beginning in 2002 [69]. The NCMS study was consid-
ered a comprehensive study at the time, but the thermal fatigue aspect of the work
ultimately was limited because the study predated the widespread introduction of
area array technology. A more recent, general discussion of the effects of alloying
on solidification, melting behavior, and properties can be found in reference [2]. A
review of the effects of alloying Sb, In, and Bi with Sn and Sn-based solders, mostly
based on observations from the respective binary phase diagrams, is provided in
the following sections.
7.2.3.1 Antimony (Sb) Additions to Tin (Sn)

The binary Sn-Sb phase diagram in Figure 7.17 shows solubility of Sb in Sn of
approximately 0.5 wt% at room temperature to 1.5 wt% at 125 ∘ C [71, 72]. Thus,
WEIGHT PER CENT ANTIMONY

10 20 30 40 50 60 70 80 90
650
630.5°
600
550
500
450
425° (Sb)
TEMPERATURE,°C
49.4 (50) 64.4 (65)

400
β
350
325° 47.4 (48)
42.4
58.4 (59) 320°
21.6
57.9
(22) (43)
300 (58.5)
57.4 (58)
8.8
250 (9) 246° 41.4
10.3 (42)
232° (10.5)
(Sn) βʹ
200
150
100
0 10 20 30 40 50 60 70 80 90 100
Sn ATOMIC PER CENT ANTIMONY Sb
Figure 7.17 The Sn-Sb binary phase diagram.

some contribution might be expected from solid solution strengthening due to Sb

dissolved in Sn-based Pb-free solders.
Alloying with Sb may improve performance through other strengthening mech-
anisms. Studies by Li et al. [73, 74] and Belyakov et al. [75] show that Sb slows the
growth rate of Cu6 Sn5 intermetallic compound (IMC) layers at attachment inter-
faces. Fast interfacial IMC growth on Cu surfaces tends to produce irregular and
non-uniform IMC layers. This can lead to reduced mechanical reliability by induc-
ing fractures at IMC interfaces or through the IMC in drop/shock loading [76].
Figure 7.17 also shows that Sb has the potential to form multiple different inter-
mediate phases or IMCs with Sn (Sb2 Sn3 , SbSn, Sb4 Sn3, Sb5 Sn4 , and SbSn2 ) in the
bulk solder [71]. Lu et al. [77] and El-Daly et al. [78] identified SbSn intermedi-
ate phase precipitates <5 μm in size and distributed throughout the Sn dendrites.
Beyer et al. showed that Sn5Sb and Sn8Sb alloys have increased shear strength and
ductility compared to conventional SAC solders and maintain their shear strength
with good ductility after isothermal aging [79]. El-Daly et al. suggested that alloy-
ing Sn with Sb can improve creep performance and tensile strength [80]. In this
case, the SbSn precipitates form within the Sn dendrites, unlike the well-known
SAC Ag3 Sn mechanism, where the precipitates form at the Sn dendrite boundaries.
Presumably, the SbSn precipitates work to resist recrystallization by strengthening
the Sn dendrites [65].
7.2.3.2 Indium (In) Additions to Tin (Sn)

The binary In-Sn phase diagram is shown in Figure 7.18. While there is some dis-
agreement over the solid solubility of In in Sn, a reasonable estimate is ∼7 wt%
at room temperature and as much as 12 wt% at 125 ∘ C [71]. Because of its range
of solubility in Sn, In has been explored as a solid solution strengthening agent
in Sn-based, Pb-free solders [68, 81]. The equilibrium diagram shows that indium
forms two intermediate phases (β and γ) of variable composition with Sn [71], but
does not appear to form any true stoichiometric compounds with Sn.
Results from multiple solder alloy studies indicate that In additions can improve
drop and shock resistance by slowing the growth of interfacial IMC layers. Yu et al.
report improved drop [82] and thermal shock [83] performance by adding as little
as 0.4% indium, and Amagai et al. report improved drop performance at or below
0.5% indium [84]. Hodúlová et al. report that indium slows growth of Cu3 Sn and
that the hybrid IMC phase Cu6 (Sn, In)5 forms [85]. Sharif et al. also observed the
formation of Cu6 (Sn, In)5 as well as formation of (Cu, Ni)3 (Sn, In)4 on Ni substrates
with Sn-Ag-Cu solder alloyed with In [86]. Those IMCs were found in the bulk as
well as the soldered interfaces. In these hybrid IMCs, In substitutes for Sn which
fundamentally is different than the common modified IMCs (Cu, Ni)6 Sn5 or the
(Ni, Cu)3 Sn4 where Cu and Ni exchange.
WEIGHT PER CENT TIN

10 20 30 40 50 60 70 80 90
300
150
275
REF. 2
125 124°
250
β 48.5 116°
100 γ 232°
225
75
30 40 50 60 70 80 90 100
200 AT.-% Sn
TEMPERATURE,°C
175
155°
150
126° (Sn)
33.2
125 26.4
117°
(In) 42.2 47.2
(48)
100
β
75
50
γ
25
0 10 20 30 40 50 60 70 80 90 100
In ATOMIC PER CENT TIN Sn
Figure 7.18 The In-Sn binary phase diagram.
Other reactions can occur when indium is added to SAC-based solders, and this
complicates the ability to understand the effect of indium content on solder joint
reliability. In a study by Chantaramanee et al. additions of 0.5% In and Sb in combi-
nation with indium was found to promote formation of Ag3 (Sn, In) and SbSn [87].
They reported that small precipitates reduced the Sn dendrite size by 28%, but they
were unable to determine the relative influence of indium versus Sb on this reac-
tion. With alloys containing indium of the order of 10%, Sopoušek et al. found that
some of the Ag3 Sn transforms to Ag2 (Sn,In) and Ag2 Sn [88]. These observations
are consistent with the Ag-In binary phase diagram that shows Ag3 In, Ag2 In, and
AgIn2 [71]. Wang et al. reported that an addition of 1% indium to Sn-Ag-Cu sol-
der resulted in larger (coarser) Ag3 Sn precipitates [89]. This is a very interesting
observation, since larger or coarser Ag3 Sn precipitate at time zero could shorten
the solder joint lifetime in thermal cycling. In principle there is a large solid solu-
bility of indium in Sn, but the effective indium content in a SAC-based solder may
be diminished by interactions with other elements to form multiple phases.
It is noteworthy that many of the studies were conducted using laboratory bulk
solder samples with microstructures that may be atypical of microelectronic solder
joints. Some studies include more than one significant alloy addition [87], which
makes it difficult to isolate effects due to individual alloying elements. The work
by Wada et al. [68, 81], while it includes tensile testing with relatively large, bulk
samples, also includes thermal cycling and drop testing with surface mount com-
ponents. Their microstructural analysis included X-ray diffraction and they found
InSn4 , In4 Ag9 , Ag3 (Sn,In), and possibly αSn in addition to βSn. Wada et al. con-
cluded that the optimum ductility and reliability was achieved with an indium
content of 6 wt%.
7.2.3.3 Bismuth (Bi) Additions to Tin (Sn)

The binary Sn-Bi phase diagram is shown in Figure 7.19. The solubility of Bi in
Sn is approximately 1.5 wt% at room temperature and increases to almost 7 wt% at
100 ∘ C, and as much as 15 wt% at 125 ∘ C [71]. There is virtually no solubility of Sn
in Bi, and no intermediate phases or IMCs found in the Sn-Bi system.
300
271.4
250 Liquid
232
200
Temperature (°C)
150 139
βSn
13.1 43 99.8
100
Bi
50
13
0 αSn
0 10 20 30 40 50 60 70 80 90 100
Sn Atomic Percent Bi Bi
Figure 7.19 The Sn-Bi binary phase diagram.

Multiple studies have shown that Bi improves the mechanical properties of Sn

and SAC solders [43, 61, 67, 90–100]. Vianco [90, 91] and Witkin [94, 97, 98] have
done extensive mechanical testing and microstructural analysis and discuss the
dual strengthening mechanisms of Bi in solid solution and Bi precipitated within
Sn dendrites and at Sn boundaries. Recently, Delhaise et al. [99] reported results
from their study of the effects of thermal preconditioning (aging) on microstruc-
ture and property improvement in an alloy containing 6 wt% Bi (see Table 7.2,
Violet). They suggest that strain from Bi precipitation induces recrystallization
and an increase in the amount of Sn grain boundaries which in turn, are pinned
by the Bi precipitates at those boundaries. These microstructural features work in
conjunction with Bi in solid solution to resist creep deformation.
The results from the fundamental studies by Vianco [90, 91] and Witkin
[94, 97, 98] leave no doubt that Bi additions can have a positive effect on the
physical properties of Sn and Sn-based solder alloys. However, those studies
used cast, bulk alloy samples and it is debatable if those results can be scaled
effectively to smaller, microelectronic solder joints. Nishimura et al. for example,
recommend a maximum Bi content of only 1.5 wt% because of the uncertainty
that the alloying effect will be sustained as the microstructure evolves in response
to the thermal cycling in normal service [96]. Delhaise has shown that the Bi
distribution and microstructure depend on solidification conditions and subse-
quent thermal exposure, which ultimately determine the relative contributions
of Bi to solid solution and dispersion strengthening (Figure 7.20). Furthermore,
it is possible that adding enough Bi to take advantage of the Bi solubility limit
at higher temperatures may have a negative effect because Bi does not always
precipitate homogeneously. Clustering of Bi is known to occur [99] and in the
extreme case, stratification or segregation may induce brittle behavior [101, 102].
7.3 Reliability Testing Third Generation Commercial

Pb-Free Solders
7.3.1 Thermal Fatigue Evaluations

The reliability challenges for third generation Pb-free solder alloys are similar to,
but even more demanding than, those encountered with the earlier proliferation
of second generation alloys [17]. Second generation alloys fundamentally are
Sn-Ag-Cu (SAC) alloys, whereas the third generation alloys while SAC-based, are
modified heavily with additional alloying elements. The potential for multiple
temperature dependent strengthening mechanisms in the newer alloys increases
the complexity of microstructural response to thermal and mechanical service
conditions (see Section 7.2.3).
300
Optimizing Bi Level
Two Phases
250
Single Phase • Sn with Bi in Solid Solution Liquid
• Bi in solid solution in Sn • Bi
200 α α
Microstructural Instability α
T (°C)
150
T (°C)
• At room temperature alloy with α β-Sn

150
C1 composition C1 microstructure (Sn)
100 has two phases.
• If in service temperature rises 100 (Bi)
T1 above T1 the Bi will start to Bi
50 dissolve
• When temperature falls below 50
T1 Bi will start to precipitate β-Sn
0
0 10 20 30
Sn 0
0 10 20 30 40 50 60 70 80 90 100
Sn Mass % Bi Bi
(a) (b)
Figure 7.20 Emphasis on the Sn-rich regions of the Sn-Bi binary phase diagram showing: (a) Factors to consider when optimizing the Bi
level [101], and (b) Schematic microstructures shown for solid solution (upper) and dispersion strengthening (lower) with a 6 wt% Bi alloy
[99].
Second generation alloy development was driven by experience gathered

through volume manufacturing, increased deployment of a variety of Pb-free
products of increasing complexity, and analysis of failures in service [17]. In that
case, there was a clear need to address certain performance deficiencies of first
generation solder alloys. Third generation alloys are being developed to address
emerging challenges in automotive applications or in the case of aerospace/
defense applications, there is a perception that that Sn-based Pb-free alloys will
struggle to perform as well as the traditional SnPb eutectic solder alloy. For
safety-related and mission-critical requirements, there is no luxury of gathering
information from a learning curve during product deployment. Consequently,
there is an urgency among high reliability end users to begin characterizing and
understanding the long-term attachment performance of these new solder alloys
prior to widespread deployment.
Most of the test and evaluation data for third generation alloys has been col-
lected through various combinations of thermal and mechanical tests on solder
attachments or bulk property measurements. Furthermore, much of the perfor-
mance data in the literature specifically is for the Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni
(Innolot) alloy. This alloy was developed for high reliability applications and it
was the first alloy of its type introduced commercially. Testing has been limited
mostly to simple components such as ceramic capacitors or chip resistors. Trodler
et al. [103] compared SAC305 to Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni (Innolot) and
Sn2.5Ag0.5Cu2.0In+Nd (HT1), another highly alloyed solder. Trodler showed
that both high reliability alloys outperformed SAC305 by testing solder joints of
ceramic capacitors with thermal cycling preconditioning followed by shear to
failure [103]. Miric also showed that Innolot outperformed SAC305 (Figure 7.16)
using thermal cycling and shear [61], and Chada et al. [104] demonstrated that
the alloy has higher creep resistance and improved reliability in −40 to +150 ∘ C
thermal cycling compared to Sn3.8Ag0.7Cu (SAC387). The thermal cycling data
reported were for chip resistors, but the test did not include in situ monitoring to
enable development of Weibull statistics [104]. Dudek et al. [105] also provided
additional confirmation of superior performance compared to SAC305 using
electrically functional test modules from production, but they did not have in situ
resistance monitoring of the solder joints. Solder joint cracking was determined
through X-ray computed tomography and metallographic analysis [105]. Barry
reported superior tensile properties of Innolot compared to SAC as part of a high
cycle fatigue study, but was unable to conclude definitively that the highly alloyed
material outperformed eutectic SnPb in high cycle fatigue [106]. Results from the
Barry study suggest that an increase in solder alloy strength does not guarantee
an increase in high cycle fatigue performance.
Su et al. [107] used ball shear as a function of strain rate and mechanical shear
fatigue and Chowdhury et al. [108] used uniaxial tensile tests over the range of
7.3 Reliability Testing Third Generation Commercial Pb-Free Solders 213
room temperature to 200 ∘ C to compare performance of three proprietary solder

alloys: Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni (Innolot), Sn3.4Ag0.5Cu3.3Bi (Cyclomax
or SAC-Q), and Sn0.9Cu2.5Bi (Ecolloy or SAC-R). Their results showed better
fatigue reliability and superior high temperature tensile strength with higher Ag
and Bi contents but results from this type of testing cannot be used to predict
the thermally-driven microstructural evolution that occurs during temperature
or power cycling. Snugovsky et al. [109] and Kosiba et al. [110] used thermal
cycling to screen various alloys with high Bi content. Their testing demonstrated
that adding Bi can improve thermal cycling performance of a basic SAC alloy.
However, they were unable to quantify the effect of Bi content because the test
did not run long enough to develop statistical reliability data.
The thermal cycling study by Hillman and Wilcoxon [111] was likely the first
to use in situ monitoring to evaluate high reliability alloys and report statistical
data for reliability comparisons. Although the study focused on comparisons of
Pb-free and SnPb solders as well as mixed metallurgy joints (Pb-free components
soldered with SnPb solder), the authors had the foresight to include a compari-
son of Sn3.9Ag0.6Cu (SAC396) to Sn3.0Ag0.5Cu3.4Bi (SAC+3.4 wt% Bi). In their
experiments using the −55/125 ∘ C (TC4) thermal cycle defined in IPC-9701 [53],
the SAC+3.4%Bi alloy outperformed SAC396 with a ceramic leadless chip carrier
(CLCC), thin small outline package (TSOP), and thin quad flat pack (TQFP) com-
ponents [111].
Until very recently, the work by Sanders et al. was one of a very few investiga-
tions that used a BGA test vehicle fabricated with third generation high reliability
solder spheres [112]. While the Sanders study included multiple components and
test variables, the relevant findings are from the −40/125 ∘ C (TC3) thermal cycling
comparison of SAC305 and Innolot using a chip array BGA (CABGA208) test vehi-
cle. In this test, the high reliability alloy outperformed SAC305 by roughly a factor
of two based on Weibull two-parameter characteristic lifetimes.
7.3.2 iNEMI/HDPUG Third Generation Alloy Pb-Free Thermal Fatigue

Project
Because there are very few published reports of thermal cycling data for third
generation alloys, a thorough treatment and an understanding of thermal fatigue
performance is a major knowledge gap for these new alloys. To address this
knowledge gap, the high reliability end users and solder suppliers from the
iNEMI Alloy Alternatives Characterization Project [9, 17] agreed to extend
the project to encompass thermal fatigue evaluations of third generation high
reliability solder alloys [39]. The Third Generation Alloy Evaluation project was
launched jointly by iNEMI and the HDPUG, and includes active participation
from both the CALCE (the Center for Advanced Life Cycle Engineering) [22],
Table 7.3 Nominal solder compositions and estimated melting ranges for the
high reliability solder alloys included in the iNEMI/HDPUG Third Generation Alloy
study [39].
Alloy Nominal composition (wt%) Melting range (∘ C)

Sn Ag Cu Bi Sb In Other
SAC305 96.5 3.0 0.5 217–221
Innolot 91.3 3.5 0.7 3.0 1.5 0.12 Ni 206–218
HT 95.0 2.5 0.5 2.0 Nd 206–218
MaxRel plus 91.9 4.0 0.6 3.5 212–220
M794 89.7 3.4 0.7 3.2 3.0 Ni 210–221
M758 93.2 3.0 0.8 3.0 Ni 205–215
SB6NX 89.2 3.5 0.8 0.5 6.0 202–206
Violet 91.25 2.25 0.5 6.0 205–215
Indalloy 272 90.0 3.8 1.2 1.5 3.5 216–226
Indalloy 277 89.0 3.8 0.7 0.5 3.5 2.5 214–223
Indalloy 279 89.3 3.8 0.9 5.5 0.5 221–228
LF-C2 92.5 3.5 1.0 3.0 208–213
SN100CV 97.8 0.7 1.5 0.05Ni 221–225
405Y 95.5 4.0 0.5 0.05 Ni; Zn 217–221
and the AREA (Universal Advanced Research in Electronic Assembly) consortia

[23]. These consortia collectively are supported by members from high reliability
markets including telecom, automotive, avionics, and aerospace/defense end
users, as well as multiple solder suppliers, and contract electronic manufacturers.
The alloys included currently in this investigation are shown in Table 7.3. The
test matrix contains SAC305 as the performance baseline alloy. The table shows
that Bi is the alloying element used most frequently, which is consistent with the
attention given to Bi in the Pb-free alloy literature [43, 61, 67, 90–96].
This project uses the basic experimental approach developed in the original
iNEMI Alloy Alternatives study of second generation solders to enable develop-
ment of thermal fatigue data for third generation Pb-free solders. The investigation
utilizes the components and printed circuit board (PCB) developed as the test
vehicle for the earlier study [17]. Figure 7.21 shows a populated test board and the
two daisy-chained BGAs, a 192 I/O chip array BGA (192CABGA), and an 84 I/O
thin core chip array (84CTBGA) [113–115]. The test boards for each alloy were
built using the same alloy composition for the BGA spheres and the solder paste
[39]. The thermal fatigue test plan incorporates several thermal cycling profiles
used by high reliability end users in telecom, aerospace/defense and avionics, and
Figure 7.21 A fully

populated, daisy-chained
test vehicle and BGA
components used in the
Third Generation Alloy
project for developing
thermal fatigue data.
Table 7.4 Thermal cycling profiles used in the iNEMI Third

Generation Alloy project: telecommunications (TC1), automotive and
consumer (TC3), and avionics and aerospace/defense (TC4).
Thermal Minimum Maximum Temp. range

cycle temp. (∘ C) temp. (∘ C) 𝚫T (∘ C) Dwell time (min)
TC1 0 100 100 10

TC3 −40 125 165 10
TC4 −55 125 180 10
automotive markets as shown in Table 7.4. The experimental matrix includes 165
populated test boards and nearly 5000 components for thermal cycling tests and
time zero microstructural characterization.
The initial results from this thermal cycling study have been reported [113–115].
One of the comparisons made in these papers demonstrates the effect of adding
high amounts of three different individual alloying elements, bismuth (Bi), anti-
mony (Sb), and indium (In), on thermal cycling performance. The alloys include
Violet (6 wt% Bi), SB6NX (6 wt% In), and Indalloy 279 (5.5 wt% Sb) from Table 7.3.
Summaries of the thermal cycling data are shown in Table 7.5 and Figure 7.22.
For the 192CABGA component, the high reliability alloys outperformed SAC305
in all thermal cycles with two notable exceptions. The performance of Violet was
no better than SAC305 in 0/100 and −40/125 ∘ C and the performance of Violet was
substantially worse than SAC305 in −55/125 ∘ C. The thermal cycling results for
the 84CTBGA component were significantly different as illustrated in Figure 7.22.
While all high reliability alloys outperformed SAC305 in all thermal cycles, the
performance of Violet nominally was best in the group, which is in sharp contrast
to the performance of that alloy when tested with the 192CABGA package.
The alloy containing 5.5 wt% Sb, performed consistently well with both compo-
nents and all thermal cycling profiles. The alloy containing 6 wt% In, performed
better than the alloy containing 6 wt% Bi with the larger 192CABGA component
but that trend was reversed with the smaller 84CTBGA. However, those differences
are not always substantial and may not be statistically significant because compar-
isons using these Weibull statistics have limitations due to variations in Weibull
slopes (β) across the data sets.
Metallographic failure analysis of thermally cycled 192CABGA packages
showed the failure mode to be predominantly fatigue in the bulk solder as shown
in Figure 7.23. However, interfacial or mixed mode fracture was detected in the
Violet (6 wt% Bi) and Indalloy 279 (5.5 wt% Sb) alloys as shown in Figure 7.24.
No interfacial cracking was reported for the SB6NX alloy (6 wt% In). Because
there was no evidence of interfacial cracking at time zero in these alloys, it is
assumed that the cracking initiated during thermal cycling. Interfacial cracking
was detected more often in samples from the aggressive −55/125 ∘ C thermal
cycling profile and in the Violet alloy, and some interfacial cracking was detected
in Violet samples tested with the 0/100 ∘ C profile. Considering the range of
elemental additions in these alloys, no clear correlation has emerged between
composition and interfacial fracture.
While it is not obvious if the interfacial cracking impacts the Weibull statistics for
either alloy, the propensity for interfacial fracture and possible early failures could
represent a reliability risk. More detailed failure analysis and Weibull statistics are
provided in the publications [113–115].
Metallographic failure analysis of thermally cycled 84CTBGA packages con-
firmed fatigue cracking in the bulk solder. Figure 7.25 shows backscattered
scanning electron images of examples of crack propagation in the four alloys.
None of the 84CTBGA samples that were analyzed from the three thermal cycles
contained the type of clear interfacial separation found in the 192CABGA images
shown in Figures 7.23 and 7.24. The absence of interfacial cracking could account
Table 7.5 A summary of accelerated temperature cycling failure statistics for SAC305 and three high reliability solder alloys [113].
192CTBGA Thermal Cycling 84CTBGA Thermal Cycling
8000 25000
6917
20447
7000
6183
20000
5607
Characteristic Lifetime
Characteristic Lifetime
5442
6000
14997
5000
15000
4000
9852
2751
10000
7286
3000
2042
5580
1765
1692
5034
1557
2000
1290
3584
3444
1123
3151
2758
5000
2269
834
1946
1000
0 0
0/100C –40/125C –55/125C 0/100C –40/125C –55/125C
ATC Profile ATC Profile
SAC 305 Violet SB6NX Indalloy 279 SAC 305 Violet SB6NX Indalloy 279
Figure 7.22 Bar charts comparing the characteristic lifetimes (N63) of the 192CABGA and 84CTBGA packages with SAC305, Violet, SB6NX,
and Indalloy 279. Data are presented for three different thermal cycling profiles: 0/100, −40/125, and −55/125 ∘ C thermal cycling
profiles [113, 115].
SB6NX Violet Indalloy 279
Figure 7.23 Solder fatigue failures in the 192CABGA component with the SB6NX (6 wt%
In), Violet (6 wt% Bi), and Indalloy 279 (5.5 wt% Sb) alloys [113].
for the distinctly better performance of the Violet alloy with the smaller 84CTBGA
package (Figure 7.22).
7.3.3 Microstructure and Reliability of Third Generation Alloys

After more than a decade of experience, there is a relatively good understand-
ing of the microstructure of SAC305 and its evolution during thermal cycling.
The as-solidified SAC305 microstructure shown in Figure 7.10 consists of primary
Sn dendrites with secondary Ag3 Sn precipitates (shown as the lighter phase) at
the cell or dendrite boundaries. The large amount of undercooling required typi-
cally to nucleate the β–Sn phase tends to suppress the equilibrium ternary eutectic
structure [116] even in near-eutectic alloys such as SAC305 [117]. The mecha-
nism of thermal fatigue in SAC solders is well-documented and is illustrated in
Figures 7.10 and 7.11.
High reliability solders typically contain significant additions of Ag (Table 7.3),
so coarsening of Ag3 Sn precipitates and recrystallization is expected to occur.
However, the potential for solid solution strengthening and formation of addi-
tional phases that can influence performance complicates the understanding and
analysis of these alloys. Figure 7.26 shows the baseline (before thermal cycling)
microstructures of three high reliability alloys to illustrate the effect of In, Bi, and
Sb on microstructure.
The alloy containing 6 wt% In also contains 3.5 wt% Ag and its basic microstruc-
ture is similar to SAC305. There are primary Sn cells with secondary Ag3 Sn pre-
cipitates at the Sn dendrite boundaries. It is reasonable to assume that most of
the In has dissolved into the β–Sn, but the morphology of some precipitates in
the higher magnification image could indicate the presence of Ag3 (Sn, In) [87] or
In4 Ag9 [81]. A more sophisticated analytical method such as electron backscatter
diffraction (EBSD) is required to identify these phases.
The microstructure of the alloy containing 6 wt% Bi has less visible Ag3 Sn pre-
cipitates due to its lower Ag content (2.25 wt%) and because the image is domi-
nated by numerous bright white Bi precipitates. Some Bi undoubtedly remains in
Nominal Composition (wt. %) Melting Nominal Composition (wt. %) Melting
Alloy Alloy
Sn Ag Cu Bi Sb In other Range, °C Sn Ag Cu Bi Sb In other Range, °C
Violet 91.25 2.25 0.5 6.0 205–215 Indalloy 279 89.3 3.8 0.9 5.5 0.5 221–228
fatigue interfacial
–55/125 °C –55/125°C
0/100 °C
Figure 7.24 Interfacial or mixed mode fracture in the 192CABGA component with the Violet (6 wt% Bi) and Indalloy 279 (5.5 wt% Sb) alloys
[113].
10 μm SAC305 10 μm SB6NX
Violet
10 μm 10 μm Indalloy 279
Figure 7.25 Backscattered electron micrographs showing examples of representative

thermal fatigue damage in the 84CTBGA component with the SAC305, SB6NX (6 wt% In),
Violet (6 wt% Bi), and Indalloy 279 (5.5 wt% Sb) alloys [115].
solid solution. Researchers have studied alloys with similar Bi content and have
explored the stability of Bi in solution with Sn after thermal preconditioning [99],
precipitation and morphology of Bi at room temperature [118], and effect on creep
rate [119]. More work is needed to understand the reaction of Bi to various thermal
cycling profiles and to understand the interaction between strain and temperature
on reliability.
The alloy containing 5.5 wt% Sb has 3.8 wt% Ag, which is a higher Ag content
than the alloys containing high Bi or In content and this may explain its higher
density of Ag3 Sn precipitates. There is limited solubility of Sb in Sn, but no obvi-
ous evidence of precipitation of Sn-Sb intermetallic phases. The morphologies
and contrast (density) differences of the precipitates in the alloy prior to thermal
cycling suggest some precipitates are not simply Ag3 Sn. Based on reports from the
literature, these precipitates most likely are SbSn [77, 78, 80], but the precipitates
are extremely small and additional analysis is needed to confirm the composition.
Fundamental microstructural analysis of as-solidified microelectronic-sized sol-
der joints is needed to assess solubility of alloying elements and phase identifica-
tion before any in depth analysis can be attempted on failed solder joints from
Nominal Composition (wt. %) Melting Nominal Composition (wt. %) Melting Nominal Composition (wt. %) Melting
Alloy Sn Ag Cu Bi Sb In other Range, °C
SB6NX 89.2 3.5 0.8 0.5 6.0 202–206 Violet 91.25 2.25 0.5 6.0 205–215 Indalloy 279 89.3 3.8 0.9 5.5 0.5 221–228
10 μm 10 μm 10 μm
SB6NX 5 μm Violet Indalloy 279 5 μm

5 μm
Figure 7.26 Backscattered electron micrographs of the baseline microstructures of alloys containing 6 wt% In (SB6NX), 6 wt% Bi (Violet),
and 5.5 wt% Sb (Indalloy 279) [113].
7.4 Reliability Gaps and Suggestions for Additional Work 223
thermal cycling tests. However, for the microstructural analysis of these highly
alloyed, third generation solders, there are clear limitations with conventional ana-
lytical methods such as optical microscopy and scanning electron microscopy with
energy dispersive X-ray analysis (SEM-EDX).
7.4 Reliability Gaps and Suggestions for Additional

Work
New high reliability solders continue to be developed and introduced, but a fun-
damental understanding of alloy behavior and characterization of solder attach-
ment reliability is lagging. The iNEMI/HDPUG Third Generation Alloy project
was launched to begin the process of generating thermal fatigue data for high relia-
bility solders and establishing an industry state of knowledge for performance and
its relationship to alloy metallurgy and microstructure. After almost two years of
thermal cycle testing some interesting findings are emerging, but the results cur-
rently are limited to only one of the BGA components and a fraction of the solder
alloys in the test matrix [113, 114]. This is the first large-scale project in the indus-
try to address thermal fatigue of these alloys, and projects of this scope and detail
are needed to close knowledge gaps. Even though the early results are incom-
plete, preliminary data analysis has uncovered additional gaps and suggestions
are being made for further investigations. The following sections discuss reliabil-
ity gaps for third generation Pb-free solders and suggest additional investigations
to close those gaps.
7.4.1 Root Cause of Interfacial Fractures

The observation of non-solder fatigue failures shown in Figure 7.24 requires
further destructive analysis to identify failure modes. Beyond complicating
interpretation of the thermal cycling test data, non-fatigue failure modes could
introduce a reliability risk in practice. Interfacial cracking was not anticipated,
and it adds another significant parameter to the evaluation of attachment relia-
bility. It is important to develop a better understanding of the interfacial cracking
phenomenon in terms of root cause and the impact on reliability. Interfacial
accumulation of Bi can occur and in the extreme case, may induce brittle behavior
by the process known as bismuth stratification [101, 102]. However, Bi cannot
be the sole cause of the interfacial cracking because some interfacial cracking
was detected in an alloy containing no Bi (Figure 7.24). Therefore, a careful
microscopic analysis should be performed at soldered interfaces to identify
segregation of Bi or other species that could initiate cracking.
7.4.2 Effect of Component Attributes on Thermal Fatigue

The iNEMI/HDPUG project includes numerous alloys, several thermal cycling
profiles, and large sample sizes, but the investigation is restricted to two relatively
small BGA components. Solder fatigue is driven by strain as well as temperature
and is dependent strongly on the coefficient of thermal expansion (CTE) mismatch
(difference) between the component and the PCB as well as the distance from neu-
tral point (DNP) [120]. The BGA packages in the iNEMI/HDPUG study tend to
minimize the DNP effect, and their relatively large die to package ratios (DPRs)
result in substantial CTE mismatch [121]. The interfacial cracking observed with
the 192CABGA package may be more a consequence of its very large DPR than a
specific alloying effect.
The small volume of the solder balls used with these packages is expected
to result in more Sn undercooling, which is known to influence solidification,
microstructure, and potentially reliability [116, 122]. It is not safe to assume that
the thermal fatigue, interfacial cracking behavior, and overall performance of
alloys from this BGA study will translate or scale to other packages or compo-
nents. Thus, future studies should include BGA packages with larger DNP, larger
solder balls, and various DPR, and wafer-level chip scale packages (WLCSPs)
with much smaller solder balls. Future studies should also include other surface
mount components in common use such as resistors, quad flat no-leads (QFN),
and connectors, which would also provide opportunities to evaluate performance
of high reliability alloy solder pastes.
7.4.3 Effect of Surface Finish on Thermal Fatigue

Interfacial cracking has been reported at the PCB side of Pb-free and SnPb sol-
der joints. Typically, this has been attributed to a surface finish anomaly such
as black pad or micro-voiding [123–126]. It is not known if similar phenomena
occur during the testing of components assembled with highly alloyed SAC-based
solders. The PCB surface finish in the iNEMI/HDPUG study is predominantly
organic solderability preservative (OSP), which results in soldering onto a copper
(Cu) substrate. A single thermal cycling test cell in the experiment used electro-
less nickel/immersion gold (ENIG), which results in soldering onto a nickel (Ni)
substrate. This provides a limited view of possible finish effects within this study,
and there are multiple other PCB surface finishes utilized in practice [127–129].
To characterize any possible interaction between final finishes and highly alloyed
solders, future investigations should consider incorporating PCB finish as a major
test variable.
Interfacial cracking was induced at the package/solder interface during thermal
cycling of some iNEMI/HDPUG samples. The BGA packages in the study used a
traditional electrolytic nickel/gold finish. Some current package designs use other
finishes such as OSP to avoid joint embrittlement at the interface [130, 131]. Sim-
ilarly, an OSP BGA pad finish might mitigate the interfacial fracture reported for
several high reliability alloys shown in Figure 7.24 [113]. The addition of gold (Au)
dissolved from the surface finish was enough to alter Sn grain morphology and
improve reliability in very small volume solder joints [29, 132]. Package surface
finish should be a variable in future thermal cycling experiments but acquiring
daisy-chained packages with other finishes may be costly, so this is likely to have
lower priority for thermal cycling studies.
7.4.4 Thermomechanical Test Parameters and Test Outcomes

7.4.4.1 Thermal Cycling Dwell Time
The thermal cycling profiles used in the iNEMI/HDPUG study are typical of those
used across the industry, and have relatively short hot and cold dwell times of
10–15 minutes. The effect of longer dwell times in thermal cycling of SAC sol-
ders was explored in considerable detail by many researchers [13, 30, 41, 42, 47,
133–141]. Although these studies demonstrated that longer dwell times reduce
reliability of SAC solders, shorter dwells eventually were accepted for SAC-type
solders where Ag3 Sn precipitate coarsening is the only significant solid-state reac-
tion during thermal cycling. The findings from dwell time studies of SAC solders
will not be directly applicable to high reliability solders due to the overlaid effects
of solid solution and dispersion strengthening mechanisms, in some cases from
multiple alloying elements (described in Section 7.2.3). The microstructure and
performance of these high reliability alloys, particularly alloys containing Bi, can
be very sensitive to thermal exposure [75, 99, 117, 119]. Future evaluations should
consider incorporating test profiles with extended dwell times to develop data to
enhance the understanding of alloy behavior in service environments. A suggested
starting point, based on the results for SAC alloys from the literature, would be
60 minutes at the upper and lower temperature extremes.
7.4.4.2 Preconditioning (Isothermal Aging)

The effect of preconditioning, often referred to as aging, is another phenomenon
that has been studied exhaustively for SAC alloys [10, 26, 35, 42, 112, 142–160].
In most of these studies, preconditioning exposure lead to a modest decrease in
characteristic thermal fatigue life, which is attributed to changes in the Ag3 Sn pre-
cipitate distribution. Coarsening of this microstructural hardening phase reduces
the resistance to creep deformation during the thermal cycle. While the magnitude
of this effect typically is only moderate in SAC alloys, Delhaise has shown a sig-
nificant response to preconditioning in SAC alloyed with Bi [99, 161]. This could
present an additional reliability risk for products subject to isothermal exposure
in addition to thermal cycling for applications with long-term storage exposure

[142]. Similar to Ag3 Sn precipitate coarsening, intermediate phase precipitates of
In and Sb would also be expected to coarsen and lose their strengthening effect
with aging. Investigations of aging behavior will be needed to develop a complete
understanding of the reliability of third generation solder alloys, but interpreta-
tion of the results is expected to be more difficult due to the complex metallurgy
and multiple strengthening mechanisms.
7.4.4.3 Thermal Cycling of Mixed Metallurgy BGA Assemblies

The iNEMI/HDPUG test matrix uses matching solder paste and ball composi-
tions to evaluate the situation of a so-called pure high reliability alloy. Today, the
vast majority of BGA packages are fabricated with high-Ag solder balls, generally
SAC305. If a high reliability alloy solder paste was proposed for implementation,
the alloy compositions of the solder paste and all the BGA components would
not be matched. This is known as mixed metallurgy or mixed alloy assembly and
is illustrated schematically in Figure 7.27 [43]. Evaluations of assembly quality
and thermal fatigue attachment reliability of the mixed assemblies are required to
demonstrate alloy compatibility.
7.4.4.4 Thermal Shock or Aggressive Thermal Cycling

Historically, the thermal cycling parameters in IPC-9701 have defined the accepted
practice for evaluating solder attachments [53]. For applications in harsh envi-
ronments, a more aggressive cycling profile such as −40/150 ∘ C may be specified
[103]. Thermal shock, which typically is not specified for long-term board-level
attachment evaluations, may be required for automotive or defense applications.
Thermal shock is characterized by hot and cold ramp rates that are greater than
20 ∘ C min−1 . This contrasts with the ramp rates in thermal cycling that are less
Figure 7.27 Illustration of

mixed metallurgy BGA
assembly with SAC305 BGA
ball and a different solder
paste [43].
SAC305
New Solder
than 20 ∘ C min−1 and are more typically 10 ∘ C min−1 . It also is necessary to explore
the effect of power cycling on attachment reliability [162], but power cycling test
vehicle design is challenging and the experiments are costly and difficult to exe-
cute. An industry workshop on next generation solder alloys highlighted the need
for aggressive testing to evaluate new alloys [163].
A learning curve is expected with this new class of alloys because there is lim-
ited experience using aggressive thermal cycling and thermal shock profiles for
long-term testing. One of the challenges of using aggressive testing profiles is the
potential for introducing multiple failure modes. Figure 7.28 shows multiple dam-
age mechanisms in a BGA sample from a −40/150 ∘ C thermal cycling test. Solder
attachment reliability cannot be characterized and comparisons of alloy perfor-
mance cannot be made when multiple damage mechanism are acting. Test vehicle
design and laminate material selection must be explored to eliminate such a sig-
nificant noise factor from the experiments.
7.4.5 Reliability Under Mechanical Loading: Drop/Shock,

and Vibration
The primary performance focus for the evolving class of high reliability solder
alloys has been to promote better high temperature performance. To achieve this,
solder developers have modified basic Sn-Ag-Cu or Sn-Cu alloys with significant
major and micro alloying additions as illustrated in Table 7.3. In addition to the
requirement for superior resistance to thermal fatigue damage, targeted appli-
cations for high reliability alloys in automotive, avionics, and defense also have
requirements for resistance to damage from high strain rate or repetitive mechan-
ical loading. The electronics industry, including its major consortia, does not have
an active, systematic program in place to address the performance of these high
reliability solders under conditions of mechanical shock and vibration. This is a
major gap in the overall reliability risk assessment for third generation, high reli-
ability solder alloys [163].
The development of shock and vibration data to complement acceptable ther-
mal fatigue performance is necessary for implementation of high reliability solders
in automotive, avionics, and defense applications. However, the development of
shock and vibration data is a challenging task for a variety of reasons. IPC-9701
provides detailed guidance for thermal cycling test methods and requirements for
surface mount solder attachments [53]. IPC-9701 has been in use since 2002 and
is recognized and used universally by device and solder suppliers, OEMs (origi-
nal equipment manufacturers), OSATs (outsourced assembly and tests) and con-
tract electronic manufacturers. Documents with a similar legacy do not exist for
drop/shock or vibration test methods and requirements for Pb-free board-level sol-
der attachments.
Figure 7.28 Multiple

fracture modes in a
single BGA sample after
approximately 2000
thermal cycles using a
−40/150 ∘ C profile.
Cracking is detected in
the die-attach and
package substrate
(above) and in the solder
(center), and pad
cratering is detected in
PCB (below) Source: Data
courtesy of the
iNEMI/HDPUG Third
100 μm Generation Alloy project.
100 μm
The test vehicle design from the latest JEDEC standard for board-level drop test
[164, 165] is being used in a consortium study evaluating low temperature sol-
ders [166], but the JEDEC drop method and test vehicle have not yet achieved the
widespread usage or acceptance that IPC-9701 has for thermal cycling.
There is even less clarity with vibration testing standards. A task force is eval-
uating revisions to the current JEDEC test specification (JESD22-B103-B), which
defines test parameters but does not define a test vehicle [167]. Because vibration
depends strongly on the characteristics of the PCB or test vehicle design, some
companies including avionics and defense contractors have designed test vehicles
representative of their component mix and product [168, 169]. Automotive OEMs
perform vibration testing for qualification, product verification, and research and
development, and often require testing by their component manufacturers. Auto-
motive testing tends to be very application-specific and generally utilizes actual
product boards not general test vehicles. To summarize, vibration testing protocols
often are performed at a system level and target application performance require-
ments and specific use conditions, whereas IPC-9701 testing protocols are per-
formed specifically to evaluate solder attachments under various conditions of
thermal cycling.
In the case of aggressive thermal cycling, the greatest challenge for testing and
ranking relative performance of solder alloys is to generate failures within the bulk
solder (Figure 7.28). This may likewise be the challenge for drop and vibration
testing. Multiple fracture modes in BGA assemblies were identified early in the
transition to Pb-free manufacturing. This provided the incentive to develop and
introduce second generation, commercial Pb-free solders as described in Section
7.1.2. Figure 7.29 shows the different failure or fracture modes that can arise from
mechanical stresses [170]. Figure 7.29 is taken from a paper published in 2005,
which demonstrates the early awareness of these phenomena with solders con-
taining at least 3 wt% Ag. Clearly this should be considered a threat for mechanical
testing of third generation alloys, since Figure 7.28 shows that several of these frac-
ture modes have been detected in aggressive thermal cycling tests of these alloys.
The review of the literature in previous sections leaves no doubt that certain alloy
additions in third generation solders result in higher tensile, shear, and mechan-
ical fatigue strength than SAC alloys. This increased solder strength resists defor-
mation and transfers loads to other areas of the interconnect structure, thereby
causing failure or fracture outside the bulk solder. This compromises test results,
restricts the ability to rank alloy performance, and could have serious implications
in practice. Because of their higher strength, third generation solders may not meet
certain application requirements for mechanical loading despite showing superior
thermal cycling performance. The industry faced a similar situation with high
Ag content first generation SAC alloys that could not meet drop requirements.
Different Failure Modes

Package Substrate A
LEGEND B
A Package Pad Lift/Crater

C
B Pkg Metal/IMC Interface Fracture
Solder Ball
C Pkg IMC/Solder Interface Fracture D
D Bulk Solder Fracture
E PWB IMC/Solder Interface Fracture E
F PWB MetaI/IMC Interface Fracture F
G PWB Pad lift/Cratering
G
PCB
Figure 7.29 An illustration of multiple fracture modes in a BGA sample caused typically
by mechanical loading [170].
Eventually they were replaced by second generation alloys that had lower over-
all strength and thermal fatigue resistance, but resisted catastrophic interconnect
damage. Third generation alloys, unlike their second generation counterparts, will
be required to provide satisfactory reliability in thermal cycling as well as mechan-
ical loading. Ultimately, testing will be necessary in combined environments of
thermal and mechanical exposure that will bring a substantial level of added com-
plexity to the process of testing and evaluation. Therefore, it is critical to begin eval-
uations of the response of third generation solder alloys to drop/shock and vibra-
tion as soon as possible. It also is critical to partner with high-performance end
users to develop effective, representative, and realistic test protocols and require-
ments.
7.4.6 Solder Alloy Microstructure and Reliability

A review of the literature including recent work by Belyakov et al. [171] indicates
that multiple phases can appear in Sn-based alloys when significant amount of Bi,
Sb, and In are present. The initial results from the iNEMI/HDPUG project indicate
phases other than the common Ag3 Sn and Cu6 Sn5 are present based on the density
differences of precipitates in the SEM backscattered images of these alloys. Some
of these precipitates are too small to be identified by the SEM Energy Dispersive
X-ray Spectroscopy (EDS or EDX) used in that investigation but those precipitates
may be critical to performance. Electron backscatter diffraction (EBSD), electron
probe microanalysis (EPMA), and Wavelength Dispersive X-ray Spectroscopy
(WDS or WDX) can be used to identify these phases. Belyakov et al. investigated
the effects of alloying additions to the Sn-Cu system, and showed that Bi is an
effective solid solution strengthening agent, moderate additions of Sb improved

ductility and modified the IMC layer but did not alter the βSn microstructure [75].
The root cause of interfacial cracking in some alloys during thermal cycling is a
major concern. It is not obvious that interfacial cracking impacts thermal cycling
Weibull statistics, but the propensity for interfacial fracture and possible early fail-
ures could represent a reliability risk [113–115]. Detailed microstructural analysis
of the bulk solder and IMC layers will be required to gain a better understanding
of thermal fatigue and mechanical reliability and failure modes of these new
solders.
7.4.7 Summary of Suggestions for Additional Investigation

Based on the gap analysis for Pb-free, third generation solder alloys described in
the previous sections, the topics for additional detailed exploration are as follows:
● Investigate the root cause of interfacial fracture. If third generation Pb-free alloys
are susceptible in practice to interfacial solder joint fractures, the consequences
could be catastrophic. Investigate the root cause of interfacial fractures in ther-
mal cycling, with an emphasis on the influence of alloy composition and thermal
cycling profile. Quantify the effect of interfacial fractures on attachment relia-
bility.
● Expand the available data for thermal cycling. Thermal cycling studies of third
generation Pb-free alloys are very limited in scope. Develop thermal cycling
data for additional BGA components of different sizes, different package con-
structions, and different attachment pad surface finishes. Develop data for other
surface mount components in common use such as resistors, QFN, and connec-
tors.
● Assess the effect of PCB and component surface finish on reliability. Determine if
the surface finish on the PCB side or on the component side of the solder joint
affects reliability by altering microstructure through microalloying or by altering
the strength of the intermetallic layer and the strength of the soldered interface.
● Expand thermal fatigue test protocols. The current thermal cycling data are lim-
ited to short dwell times that are not representative of the targeted use envi-
ronments. Longer dwell times should be incorporated into thermal cycling test
plans to enable a better determination of alloy stability and reliability. Thermal
shock and power cycling tests may be required in addition to thermal cycling.
The ability of component and board test vehicles to survive harsh thermome-
chanical testing must be resolved to differentiate alloy performance.
● Assess the effectiveness of thermal preconditioning (aging) on thermal fatigue life.
Harsh use conditions include operating temperatures that could reach or exceed
100 ∘ C. Isothermal preconditioning should be explored to provide additional
acceleration to thermal cycling. Experiments are needed to establish the rela-

tionship between alloy composition and the effects of preconditioning time and
temperature on microstructural evolution and subsequently on thermal cycling
performance.
● Mixed metallurgy or mixed alloy assembly. Evaluations of assembly quality and
thermal fatigue attachment reliability of the mixed assemblies are required to
demonstrate compatibility of highly alloyed, 3rd generation solders with current
Pb-free materials and assembly processes.
● Drop/shock and vibration testing. Applications for high reliability alloys in auto-
motive, avionics, and defense also have requirements for resistance to damage
from high strain rate or repetitive mechanical loading. Data must be developed
for mechanical reliability in parallel with the thermomechanical fatigue relia-
bility data for the class of high reliability Pb-free solder alloys.
● Relationship between microstructure and the thermal fatigue performance. The
strengthening effects associated with various alloying additions to Sn-Ag-Cu
and Sn-Cu systems are recognized. However, the identification of specific
phases, the dissolution of the alloying elements in the β-tin (Sn) matrix, and
the relationship between microstructure, microstructural evolution, and, ulti-
mately, the thermal fatigue performance have not been determined. Advanced
analytical techniques including EBSD, EPMA, and WDS (wavelength dispersive
X-ray spectroscopy) are needed to characterize phase formation in the bulk
composition, and structure of intermetallic layers, and dissolution of alloying
elements in the β-tin (Sn).
7.5 Conclusions
The drivers, benefits and concerns associated with the development and even-
tual deployment of third generation, high reliability Pb-free solders has been dis-
cussed. These alloys have the potential to solve reliability issues arising from the
need to transition from SnPb to Pb-free manufacturing or to meet performance
requirements in aggressive use environments defined by automotive, avionics, and
defense applications. However, if not managed properly, implementation of these
alloys could introduce new reliability risks. Based on a review of the literature and
input from many active researchers in the field, the following conclusions can be
made at this time.
● There is a long and expanding list of commercialized high reliability alloys based
on the Sn-Ag-Cu and Sn-Cu systems. These alloys are characterized by singular,
major alloying additions of bismuth (Bi), antimony (Sb), or indium (In) or com-
binations of these major alloying elements. Some alloys also contain microalloy
7.5 Conclusions 233
additions of a variety of elements to improve performance. The wide range of

solder alloys and combinations of elemental additions adds to the complexity
of test programs and to the subsequent risk analyses. Although these are classi-
fied as commercial alloys, volume implementations at the time of this writing
are scarce and solder suppliers continue to modify compositions in response to
emerging test data. This adds uncertainty and risk within the supply chain.
● All these new alloys appear to outperform the near-eutectic SAC305 alloy in
thermal cycling. However, early results from the iNEMI/HDPUG project reveal
much remains to be learned about their thermal cycling behavior. Early ther-
mal cycling test results along with other test data from the literature indicate Bi
and In are effective solid solution strengthening agents. Bi may also strengthen
by dispersion hardening. Additions of Sb can improve performance in thermal
cycling, probably through precipitation hardening. However, alloys containing
major additions of Bi or Sb exhibited some interfacial cracking during thermal
cycling. In contrast, an alloy with high In content showed slightly lower thermal
cycling performance but no evidence of interfacial cracking. The observation of
interfacial cracking raises concerns about alloy performance under high strain
rate mechanical loading.
● Other areas related to thermal cycling reliability that need further attention
include effects due to thermal cycling profile and dwell time, isothermal
preconditioning, power cycling, surface finish, and mixed metallurgy assembly.
Test programs should begin to address these issues to enable a complete analysis
and understanding of the performance behavior of this class of alloys.
● A better understanding of the relationships between composition, microstruc-
ture and alloy performance and failure modes during thermal cycling are
needed to assess the roles and effectiveness of the various alloying elements.
This includes developing a better understanding of microstructural stability
and evolution in the bulk and at soldered interfaces.
● The performance requirements in targeted high reliability applications address a
variety of demanding thermal and mechanical use environments. A comprehen-
sive evaluation of the performance of third generation, high reliability Pb-free
solders has not been done for thermal shock, mechanical drop/shock, or vibra-
tion. In fact, test protocols for mechanical testing of these solders are yet to be
developed, which is a significant gap.
● Investigations of existing and new third generation, high reliability alloys are
expected to accelerate as more high reliability end users become engaged. This
work is taking place at solder suppliers, OEMs, contract manufactures, and uni-
versities. Industrial consortia and working groups are expected to play a major
role because of the vast scope, significant knowledge gaps, and breadth and
depth of expertise needed to address the issues and execute these programs.
Acknowledgments
Many people contributed to the development and assessment of the information
in this chapter. The author wants to thank his current and former colleagues at
Nokia Bell Labs who helped to develop and analyze test data over several years:
Charmaine Johnson, Richard Popowich, Pete Read, and Debbie Fleming. The
author also is indebted to the iNEMI/HDPUG Third Generation, Pb-Free Alloy
Team. A large core group of dedicated collaborators has supported investigations
of alternative alloys vigorously starting with second generation alloys in 2009
to the present. Special thanks go to Rich Parker, current Project Co-Chair and
to the two past Project Chairs, Dr. Greg Henshall formerly of Hewlett-Packard
and Elizabeth Benedetto of HP Inc. The author thanks the staffs at iNEMI and
HDPUG, especially Grace O’Malley, Marc Benowitz, and Marshall Andrews for
their help in establishing and nurturing the collaborative agreement and Larry
Marcanti, and Robert Smith for their continued support in coordinating the work
between the cooperating consortia. The author also thanks his colleagues Dave
Hillman, Keith Howell, Keith Sweatman, Tetsuro Nishimura, Chris Gourlay,
Sergei Belyakov, Babak Arfaei, Joe Smetana, Jean-Paul Clech, and Andre Delhaise
for numerous insights and technical guidance. Special thanks go to Melissa
Young, Nokia Bell Labs Reference Librarian, for her expert support of literature
reviews for the project.
References
1 Annex to Directive 2002/95/EC. (2006) Restriction on the use of hazardous

substances (RoHS) in electrical and electronic equipment. Official Journal of
the European Union, 14.10.2006, L283/48-49, October 12, 2006.
2 Handwerker, C.A., de Kluizenaar, E.E., Suganuma, K., and Gayle, F.W.
(2004). Major international lead (Pb)-free solder studies. In: Handbook of
Lead-Free Solder Technology for Microelectronic Assemblies (eds. K.J. Puttlitz
and K.A. Stalter), 665–728. New York: Marcel Dekker.
3 Lead-Free Solder Project Final Report. (1997) NCMS Report 0401RE96,
Section 2.4: Properties Assessment and Alloy Down Selection. Ann Arbor,
MI: National Center for Manufacturing Sciences.
4 Marconi Materials Technology (1999) Improved Design Life and Environmen-
tally Aware Manufacturing of Electronics Assemblies by Lead-Free Soldering:
‘IDEALS’. Contract number BRPR-CT96-0140, 30 June.
5 Richards, B., Levogner, C.L., Hunt, C.P. et al. (1999) Lead-Free Soldering –An
Analysis of the Current Status of Lead-Free Soldering. London: Department
of Trade and Industry.
References 235
6 Lead-Free Soldering Roadmap Committee, Technical Standardization Com-

mittee on Electronics Assembly Technology (2002) JEITA Lead-free Roadmap
2002 for Commercialization of Lead-free Solder. JEITA.
7 JEIDA (2002) NEDO Research and Development on Lead-Free Soldering.
Report No. 00-ki-17. JEIDA.
8 Bradley, E., Handwerker, C.A., Bath, J. et al. (eds.) (2007). Lead-Free Electron-
ics: iNEMI Projects Lead to Successful Manufacturing. Hoboken, NJ: Wiley.
9 Henshall, G., Sweatman, K., Howell, K. et al. (2009). iNEMI lead-free alloy
alternatives project report: thermal fatigue experiments and alloy test require-
ments. In: Proceedings of SMTAI, 317–324. San Diego CA.
10 Smetana, J., Coyle, R., Read, P. et al. (2011). Variations in thermal cycling
response of Pb-free solder due to isothermal preconditioning. In: Proceedings
of SMTAI, Fort Worth, TX, 641–654.
11 Engelmaier, W. (1989). Surface mount solder joint long-term reliability:
design, testing, prediction. Soldering and Surface Mount Technology 1 (1):
14–22.
12 Terashima, S., Kariya, Y., Hosoi, M., and Tanaka, M. (2003). Effect of silver
content on thermal fatigue life of Sn-xAg-0.5Cu flip-chip interconnects.
Journal of Electronic Materials 32 (12): 1527–1533.
13 Coyle, R., McCormick, H., Osenbach, J. et al. (2011). Pb-free alloy silver con-
tent and thermal fatigue reliability of a large plastic ball grid array (PBGA)
package. Journal of Surface Mount Technology 24 (1): 27–33.
14 Henshall, G., Bath, J., Sethuraman, S. et al. Comparison of thermal
fatigue performance of SAC105 (Sn-1.0Ag-0.5Cu), Sn-3.5Ag, and SAC305
(Sn-3.0Ag-0.5Cu) BGA components with SAC305 solder paste. Proceedings
IPC APEX, S05-03, 2009.
15 IPC Solder Products Value Council (2005) Round Robin Testing and Analysis
of Lead-Free Solder Pastes with Alloys of Tin, Silver and Copper, Final
Report. IPC International.
16 Henshall, G. (2011). Lead-free alloys for BGA/CSP components. In: Lead-Free
Solder Process Development (eds. G. Henshall, J. Bath and C.A. Handwerker),
95–124. Hoboken, NJ: Wiley.
17 Henshall, G., Miremadi, J., Parker, R. et al. (2012). iNEMI Pb-free alloy char-
acterization project report: Part I – program goals, experimental structure,
alloy characterization, and test protocols for accelerated temperature cycling.
In: Proceedings of SMTAI, 335–347. Orlando, FL.
18 Kim, D., Suh, D., Millard, T. et al. (2007). Evaluation of high compliant low
Ag solder alloys on OSP as a drop solution for the 2nd level Pb-free intercon-
nection. In: Proceedings 57th Electronic Component and Technology Conference
Reno, NV, 1614–1619.
19 Syed, A., Kim, T.-S., Cha, S.-W. et al. (2007). Effect of Pb free alloy com-
position on drop/impact reliability of 0.4, 0.5, and 0.8 mm pitch Chip scale
packages with NiAu pad finish. In: Proceedings 57th Electronic Component
and Technology Conference, Reno, NV (May 28 – June 1), 951–956.
20 International Electronics Manufacturing Initiative (iNEMI). www.inemi.org
(accessed 4 February 2020).
21 HDP User Group, International, Inc. (HDPUG). https://hdpug.org (accessed 4
February 2020).
22 The Center for Advanced Life Cycle Engineering (CALCE), University of
Maryland. https://calce.umd.edu (accessed 4 February 2020).
23 Universal Advanced Research in Electronics Assembly. http://www.uic.com/
solutions/apl/area-consortium (accessed 4 February 2020).
24 The Center for Advanced Vehicle and Extreme Environment Electronics
(CAVE3 ), Auburn University. http://cave.auburn.edu (accessed 4 February
2020).
25 Coyle, R., Sweatman, K., and Arfaei, B. (2015). Thermal fatigue evaluation of
Pb-free solder joints: results, lessons learned, and future trends. Journal of the
Minerals, Metals & Materials Society 67 (10): 2394–2415.
26 Coyle, R., Parker, R., Smetana, J. et al. (2015). iNEMI Pb-free alloy char-
acterization project report: PART IX – summary of the effect of isothermal
preconditioning on thermal fatigue life. In: Proceedings of SMTAI, Chicago,
IL, 743–755.
27 Coyle, R., Parker, R., Benedetto, E. et al. (2014). iNEMI Pb-free alloy charac-
terization project report: PART VIII – thermal fatigue results for high-Ag
alloys at extended dwell times. In: Proceedings of SMTAI, Chicago, IL,
547–560.
28 Sweatman, K., Coyle, R., Parker, R. et al. (2014). iNEMI Pb-free alloy charac-
terization project report: PART VII – thermal fatigue results for low-Ag alloys
at extended dwell times. In: Proceedings of SMTAI, Chicago, IL, 561–574.
29 Coyle, R., Parker, R., Arfaei, B. et al. (2014). The effect of nickel microalloy-
ing on thermal fatigue reliability and microstructure of SAC105 and SAC205
solders. In: Proceedings of Electronic Components Technology Conference,
425–440. Orlando, FL: IEEE.
30 Coyle, R., Parker, R., Osterman, M. et al. (2013). iNEMI Pb-free alloy charac-
terization project report: part V – the effect of dwell time on thermal fatigue
reliability. In: Proceedings of SMTAI, Ft. Worth, TX, 470–489.
31 Coyle, R., Parker, R., Arfaei, B. et al. (2013). iNEMI Pb-free alloy charac-
terization project report: Part VI – the effect of component surface finish
and solder paste composition on thermal fatigue of SN100C solder balls. In:
Proceedings of SMTAI, Ft. Worth, TX, 490–414.
References 237
32 George, E., Osterman, M., Pecht, M. et al. (2013) Thermal cycling reliability
of alternative low-silver tin-based solders. Proceedings of IMAPS 2013, 46th
International Symposium on Microelectronics, Orlando, FL (October 2013).
33 Parker, R., Coyle, R., Henshall, G. et al. (2012). iNEMI Pb-free alloy charac-
terization project report: Part II – thermal fatigue results for two common
temperature cycles. In: Proceedings of SMTAI, Orlando, FL, 348–358.
34 Sweatman, K., Howell, K., Coyle, R. et al. (2012). iNEMI Pb-free alloy charac-
terization project report: Part III – thermal fatigue results for low-Ag alloys.
In: Proceedings of SMTAI, Orlando, FL, 359–375.
35 Coyle, R., Parker, R., Henshall, G. et al. (2012). iNEMI Pb-free alloy charac-
terization project report: Part IV – effect of isothermal preconditioning on
thermal fatigue life. In: Proceedings of SMTAI, Orlando, FL, 376–389.
36 Henshall, G., Healey, R., Pandher, R.S. et al. (2009) Addressing the opportu-
nities and risks of Pb-free solder alloy alternatives. Proceedings of Microelec-
tronics and Packaging Conference, 1–11 June, Rimini, Italy.
37 Henshall, G., Healy, R., Pander, R.S. et al. (2008). iNEMI Pb-free alloy
alternatives project report: state of the industry. Journal of Surface Mount
Technology 21 (4): 11–23.
38 Henshall, G., Healy, R., Pander, R.S. et al. (2008). iNEMI Pb-free alloy
alternatives project report: state of the industry. In: Proceedings of SMTAI,
Orlando, FL, 109–122.
39 Coyle, R., Parker, R., Howell, K. et al. (2016). A collaborative industrial
consortia program for characterizing thermal fatigue reliability of third gener-
ation Pb-free alloys. In: Proceedings of SMTAI, Rosemont, IL, 188–196.
40 Thompson, R. (2003) Proceedings of the SMTA/CAVE Workshop Harsh Envi-
ronment Electronics, Dearborn, MI (24–25 June, 2003).
41 Fairchild, M.R., Snyder, R.B., Berlin, C.W., and Sarma, D.H.R. (2002) Emerg-
ing substrate technologies for harsh-environment automotive electronics
applications. SAE Technical Paper Series 2002-01-1052.
42 Jolivet, E. and Dhond, P. (2019) Challenges in automotive packaging tech-
nologies. Chip Scale Review 15–16 May – June 2019.
43 Snugovsky, P., Bagheri, S., Romansky, M. et al. (2012). New generation of
Pb-free solder alloys: possible solution to solve current issues with main
stream Pb-free soldering. Journal of Surface Mount Technology 25 (3): 42–52.
44 Coyle, R., Read, P., McCormick, H. et al. (2011). The influence of alloy com-
position and temperature cycling dwell time on the reliability of a quad flat
no lead (QFN) package. Journal of Surface Mount Technology 25 (1): 28–34.
45 Coyle, R., Osenbach, J., Collins, M. et al. (2011). Phenomenological study
of the effect of microstructural evolution on the thermal fatigue resistance
of Pb-free solder joints. IEEE Transactions on Components, Packaging, and
Manufacturing Technology 1 (10): 1583–1593.
46 Henshall, G., Bath, J., Sethuraman, S. et al. (2009) Comparison of thermal

fatigue performance of SAC105 (Sn-1.0Ag-0.5Cu), Sn-3.5Ag, and SAC305
(Sn-3.0Ag-0.5Cu) BGA components with SAC305 solder paste. Proceedings
IPC APEX, S05-03.
47 Zhang, Y., Cai, Z., Suhling, J.C. et al. (2008). The effects of aging temper-
ature on SAC solder joint material behavior and reliability. In: Proceedings
of Electronic Components and Technology Conference, Lake Buena Vista, FL,
99–112.
48 Bieler, T., Jiang, H., Lehman, L. et al. (2008). Influence of Sn grain size and
orientation on the thermomechanical response and reliability of Pb-free
solder joints. IEEE Transactions on Components, Packaging, and Manufactur-
ing Technology 31 (2): 370–381.
49 Darveaux, R., Reichman, C., Berry, C. et al. (2008). Effect of joint size
and pad metallization on solder mechanical properties. In: Proceedings of
Electronic Components and Technology Conference, Lake Buena Vista, FL,
113–122.
50 Kang, S.K., Lauro, P., Shih, D.-Y. et al. (2004). Evaluation of thermal fatigue
life and failure mechanisms of Sn-Ag-Cu solder joints with reduced Ag con-
tents. In: Proceedings of Electronic Components and Technology Conference,
Las Vegas, NV, 661–667.
51 Dunford, S., Canumalla, S., and Viswanadham, P. (2004). Intermetallic mor-
phology and damage evolution under thermomechanical fatigue of lead
(Pb)-free solder interconnections. In: Proceedings of Electronic Components
Technology Conference, Las Vegas, NV, 726–736.
52 Coyle, R., Hillman, D., Johnson, C. et al. (2018) Alloy composition and ther-
mal fatigue reliability of high reliability Pb-free solder alloys. Proceedings of
SMTAI, Rosemont, IL (October 2018).
53 IPC-9701A (2006) Performance test methods and qualification requirements for
surface mount solder attachments. IPC International.
54 IPC Pb-Free Reliability Risk Management (PERM) Council, www.ipc.org/
ContentPage.aspx?pageid=PERM-Council (accessed 4 February 2020).
55 Rafanelli, A.J. (2017) Addressing the challenges of Pb-free technology in high
performance (aerospace/defense) products. SMTA New England Expo and
Technical Forum, Worcester, MA (16 November).
56 Nowottnick, M., Scheel, W., and Wittke, K. (eds.) (2005). Innovative Pro-
duction Processes for High-temperature Electronics in Automotive Electronics
Systems: Construction and Connection Technology, vol. 2. Germany: M. Detert
Publishing.
57 Albrecht, J. (2008) Final presentation BMBF project LIVE; Acceptance cri-
teria of thermally highly stressed miniaturized solder joints. Siemens AG,
Corporate Technology, MM6, Berlin, Germany, 17 September.
References 239
58 Trodler, J. (2009) Summary Innolot – Project March. 2000 to February 2004.

W.C. Heraeus GmbH; CMD-AM-AT, Hanau, Germany, January 2009 (com-
plete final report confidential).
59 Steen, H. and Toleno, B. (2009) Development of a lead-free alloy for
high-reliability, high temperature applications. SMT, January.
60 Albrecht, H.-J., Frühauf, P., and Wilke, K. (2009). Pb-free alloy alternatives:
reliability investigation. In: Proceedings SMTAI, San Diego, CA, 308–316.
61 Miric, A.-Z. (2010). New developments in high-temperature,
high-performance lead-free solder alloys. Journal of Surface Mount Technology
23 (4): 24–29.
62 Albrecht, H.-J., Bartl, K.H.G., Kruppa, W. et al. (2017) Soldering material
based on Sn Ag and Cu. EP1617968B1, European Patent Office, 1 March.
63 Steller, A., Zimmermann, A., Eisenberg, S. et al. (2009). Reliability testing
and damage analysis of lead-free solder joints: new assessment criteria for
laboratory methods. SAE International Journal of Materials and Manufactur-
ing 2 (1): 502–510.
64 ASM International Handbook Committee (ed.) (1990). Specialty steels and
heat – resistance alloys. In: Metals Handbook 10 (ed. ASM International
Handbook Committee), 822–1006. Materials Park, OH: ASM International.
65 Dieter, G.E. (1961). Plastic deformation of polycrystalline aggregates, solid
solution hardening. In: Mechanical Metallurgy, 128. McGraw-Hill.
66 Haasen, P. (1996). Physical Metallurgy, 3e, 375–378. Cambridge University
Press.
67 Delhaise, A., Snugovsky, L., Perovic, D. et al. (2014). Microstructure and
hardness of Bi-containing solder alloys after solidification and ageing. Journal
of Surface Mount Technology 27 (3): 22–27.
68 Wada, T., Tsuchiya, S., Joshi, S. et al. (2017) Improving thermal
cycle reliability and mechanical drop impact resistance of a lead-free
tin-silver-bismuth-indium solder alloy with minor doping of copper additive.
Proceedings of IPC APEX, San Diego, CA (14–16 February).
69 Siewert, T., Liu, S., Smith, D.R., and Madeni, J.C. (2002) Database for solder
properties with emphasis on new lead-free solders: properties of lead-free
solders release 4.0. NIST and Colorado School of Mines, February, 2002.
70 Handwerker, C.A., Kattner, U., Moon, K. et al. (2007). Alloy selection. In:
Lead-Free Electronics (eds. E. Bradley, C.A. Handwerker, J. Bath, et al.), 9–46.
Piscataway, NJ: IEEE Press.
71 Hansen, M. (1958). Constitution of Binary Alloys, 2e, 1175–1177. McGraw-Hill.
72 Elliot, R.P. (1965). Constitution of Binary Alloys, First Supplement, 802.
McGraw-Hill.
73 Li, G.Y., Chen, B.L., and Tey, J.N. (2004). Reaction of Sn-3.5Ag-0.7Cu-xSb sol-
der with Cu metallization during reflow soldering. IEEE Transactions on Elec-
tronics Packaging Manufacturing 27 (1): 77–85.
74 Li, G.Y., Bi, X.D., Chen, Q., and Shi, X.Q. (2011). Influence of dopant on
growth of intermetallic layers in Sn-Ag-Cu solder joints. Journal of Electronic
Materials 40 (2): 165–175.
75 Belyakov, S.A., Nishimura, T., Akaiwa, T. et al. (2019) Role of Bi, Sb, and In
in microstructural formation and properties of Sn-0.7Cu-0.05Ni-X BGA inter-
connections. IEEE International Conference on Electronic Packaging (ICEP),
Niigata, Japan (17–20 April).
76 Tegehall, P.-E. (2006) Review of the impact of intermetallic layers on the
brittleness of tin-lead and lead-free solder joints, Section 3, Impact of inter-
metallic compounds on the risk for brittle fractures. IVF Project Report 06/07.
IVF Industrial Research and Development Corporation.
77 Lu, S., Zheng, Z., Chen, J., and Luo, F. (2010). Microstructure and solder-
ability of Sn-3.5Ag-0.5Cu-xBi-ySb solders. In: Proceedings 11th International
Conference on Electronic Packaging Technology and High Density Packaging,
ICEPT-HDP, 410–412.
78 El-Daly, A.A., Swilem, Y., and Hammad, A.E. (2008). Influences of Ag and
au additions on structure and tensile strength of Sn-5Sb lead free solder alloy.
Journal of Materials Science and Technology 24 (6): 921–925.
79 Beyer, H., Sivasubramaniam, V., Hajas, D. et al. (2014). Reliability improve-
ment of large area soldering connections by antimony containing lead-free
solder. In: PCIM Europe Conference Proceedings, 1069–1076.
80 El-Daly, A.A., Swilem, Y., and Hammad, A.E. (2009). Creep properties of
Sn-Sb based lead-free solder alloys. Journal of Alloys and Compounds 471:
98–104.
81 Wada, T., Mori, K., Joshi, S., and Garcia, R. (2016). Superior thermal cycling
reliability of Pb-free solder alloy by addition of indium and bismuth for harsh
environments. In: Proceedings of SMTAI, Rosemont, IL, 210–215.
82 Yu, A.-M., Jang, J.-W., Lee, J.-H. et al. (2014). Microstructure and drop/shock
reliability of Sn-Ag-Cu-In solder joints. International Journal of Materials and
Structural Integrity 8 (1–3): 42–52.
83 Yu, A.-M., Jang, J.-W., Lee, J.-H. et al. (2012). Tensile properties and thermal
shock reliability of Sn-Ag-Cu solder joint with indium addition. Journal of
Nanoscience and Nanotechnology 12 (4): 3655–3657.
84 Amagai, M., Toyoda, Y., Ohnishi, T., and Akita, S. (2004). High drop test
reliability: lead-free solders. In: Proceedings 54th Electronic Components and
Technology Conference, 1304–1309.
References 241
85 Hodúlová, E., Palcut, M., Lechovič, E. et al. (2011). Kinetics of intermetallic

phase formation at the interface of Sn-Ag-Cu-X (X = Bi, In) solders with Cu
substrate. Journal of Alloys and Compounds 509 (25): 7052–7059.
86 Sharif, A. and Chan, Y.C. (2006). Liquid and solid state interfacial reactions
of Sn-Ag-Cu and Sn-In-Ag-Cu solders with Ni-P under bump metallization.
Thin Solid Films 504 (1–2): 431–435.
87 Chantaramanee, S., Sungkhaphaitoon, P., and Plookphol, T. (2017). Influence
of indium and antimony additions on mechanical properties and microstruc-
ture of Sn-3.0Ag-0.5Cu lead free solder alloys. Solid State Phenomena 266:
196–200.
88 Sopoušek, J., Palcut, M., Hodúlová, E., and Janovec, J. (2010). Thermal anal-
ysis of the Sn-Ag-Cu-In solder alloy. Journal of Electronic Materials 39 (3):
312–317.
89 Wang, J., Yin, M., Lai, Z., and Li, X. (2011). Wettability and microstructure of
Sn-Ag-Cu-In solder. Hanjie Xuebao/Transactions of the China Welding Institu-
tion 32 (11): 69–72.
90 Vianco, P.T. and Rejent, J.A. (1999). Properties of ternary Sn-Ag-Bi solder
alloys: Part I – thermal properties and microstructural analysis. Journal of
Electronic Materials 28 (10): 1127–1137.
91 Vianco, P.T. and Rejent, J.A. (1999). Properties of ternary Sn-Ag-Bi solder
alloys: Part I – wettability and mechanical properties analyses. Journal of
Electronic Materials 28 (10): 1138–1143.
92 Zhao, J., Qi, L., and Wang, X.-M. (2004). Influence of Bi on microstructures
evolution and mechanical properties in Sn-Ag-Cu lead-free solder. Journal of
Alloys and Compounds 375 (1–2): 196–201.
93 Hillman, D., Pearson, T., and Wilcoxon, R. (2010). NASA DOD −55 ∘ C to
+125 ∘ C thermal cycle test results. In: Proceedings of SMTAI, Orlando, FL,
512–518.
94 Witkin, D. (2013) Mechanical properties of Bi-containing Pb-free solders. Pro-
ceedings IPC APEX, S11-01, San Diego, CA (February 2013).
95 Juarez, J.M. Jr., Snugovsky, P., Kosiba, E. et al. (2015). Manufacturability and
reliability screening of lower melting point Pb-free alloys containing bismuth.
Journal of Microelectronics and Electronic Packaging 12 (1): 1–28.
96 Nishimura, T., Sweatman, K., Kita, A., and Sawada, S. (2015). A new method
of increasing the reliability of lead-free solder. In: Proceedings of SMTAI,
Rosemont, IL, 736–742.
97 Witkin, D. (2012). Creep behavior of Bi-containing lead-free solder alloys.
Journal of Electronic Materials 41 (2): 190–203.
98 Witkin, D.B. (2012). Influence of microstructure on quasi-static and dynamic
mechanical properties of bismuth-containing lead-free solder alloys. Materials
Science and Engineering A 532: 212–220.
99 Delhaise, A.M., Snugovsky, P., Matijevic, I. et al. (2018). Thermal pre-

conditioning, microstructure restoration and property improvement in
Bi-containing solder alloys. Journal of Surface Mount Technology 31 (1):
33–42.
100 Sweatman, K. (2017) Nihon Superior, private communication, November
2017.
101 Raeder, C.H., Felton, L.E., Knott, D.B. et al. (1993). Microstructural evolution
and mechanical properties of Sn-Bi based solders. In: Proceedings of Inter-
national Electronics Manufacturing Technology Symposium, Santa Clara, CA,
119–127.
102 Coyle, R., Aspandiar, R., Osterman, M. et al. (2016). Thermal cycle reliability
of a low silver ball grid array assembled with tin bismuth solder paste. In:
Proceedings of SMTAI, Rosemont, IL, 72–83.
103 Trodler, J., Dudek, R., and Röllig, M. (2016). Risk for ceramic component
cracking dependent on solder alloy and thermo-mechanical stress. In: Pro-
ceedings of SMTAI, Rosemont, IL, 197–203.
104 Chada, S., Currie, M., Toleno, B. et al. (2011). Developing a Pb-free solder
through micro-alloying. In: Proceedings of SMTAI, Fort Worth, TX, 724–730.
105 Dudek, R., Hildebrandt, M., Doering, R. et al. (2014) Solder fatigue accelera-
tion prediction and testing results for different thermal test- and field cycling
environments. Proceedings 5th Electronics System-integration Conference
(ESTC), Helsinki, Finland (16–18 September).
106 Barry, N. (2008) Lead-free solders for high-reliability applications: high-cycle
fatigue studies. Ph.D. thesis. Department of Metallurgy and Materials, School
of Engineering, The University of Birmingham.
107 Su, S., Jian, M., Akkara, F.J. et al. (2018) Fatigue and shear properties of high
reliable solder joints for harsh applications. Proceedings of SMTAI, Rosemont,
IL (October 2018).
108 Chowdhury, M.R., Ahmed, S., Fahim, A. et al. (2016). Mechanical charac-
terization of doped SAC solder materials at high temperature. In: 15th IEEE
Intersociety Conference on Thermal and Thermomechanical Phenomena in
Electronic Systems (ITherm), Las Vegas, NV, 1202–1208.
109 Snugovsky, P., Kosiba, E., Kennedy, J. et al. (2013). Manufacturability and
reliability screening of lower melting point Pb-free alloys containing Bi. In:
Proceedings, IPC APEX, San Diego, CA, 171–208.
110 Kosiba, E., Bagheri, S., Snugovsky, P., and Perovic, D. (2013) Microstructure
and reliability of low Ag Bi-containing solder alloys. International Conference
on Soldering and Reliability (ICSR), Toronto, ON.
111 Hillman, D. and Wilcoxon, R. (2006) JCAA/JG-PP No-lead solder project: −55
∘ C to +125 ∘ C thermal cycle testing final report. Rockwell Collins Advanced
Manufacturing Technology Group, Contract: GST 0504BM3419, May 28, 2006.
References 243
112 Sanders, T., Thirugnanasambandam, S., Evans, J. et al. (2015). Component

level reliability for high temperature power computing with SAC305 and
alternative high reliability solders. In: Proceedings of SMTA International,
Rosemont, IL, 144–150.
113 Coyle, R., Hillman, D., Parker, R. et al. (2018) The effect of bismuth, anti-
mony, or indium on the thermal fatigue of high reliability Pb-free solder
alloys. Proceedings of SMTAI, Rosemont, IL (October).
114 Coyle, R., Hillman, D., Johnson, C. et al. (2018) Alloy composition and ther-
mal fatigue of high reliability Pb-free solder alloys. Proceedings of SMTAI,
Rosemont, IL, (October).
115 Coyle, R., Johnson, C., Hillman, D. et al. (2019). Thermal cycling reliability
and failure mode of two ball grid array packages with high reliability Pb-free
solder alloys. In: Proceedings of SMTAI, Rosemont, IL, 439–456.
116 Kinyanjui, R., Lehman, L.P., Zavalij, L., and Cotts, E. (2005). Effect of sample
size on the solidification temperature and microstructure of SnAgCu near
eutectic alloys. Journal of Materials Research 20 (11): 2914–2918.
117 Yin, L., Wentlent, L., Yang, L.L. et al. (2011). Recrystallization and precipitate
coarsening in Pb-free solder joints during thermomechanical fatigue. Journal
of Electronic Materials 41 (2): 241–252.
118 Belyakov, S.A., Xian, J., Zeng, G. et al. (2019). Precipitation and coarsening
of bismuth plates in Sn-Ag-Cu–Bi and Sn-Cu-Ni–Bi solder joints. Journal of
Materials Science: Materials in Electronics 30: 378–390.
119 Sweatman, K., Akaiwa, T., and Nishimura, T. (2018) Effect of creep rate
of alloying additions to Ni-stabilised Sn-Cu eutectic solder. Proceedings of
SMTAI, Rosemont, IL, (October).
120 Engelmaier, W. (1990). The use environments of electronic assemblies and
their impact on surface mount solder attachment reliability. IEEE Transac-
tions on Components, Hybrids, and Manufacturing Technology 13 (4): 903–908.
121 Syed, A., Panczak, T., Darveaux, R. et al. (1999). Solder joint reliability of
chip array BGA. Journal of Surface Mount Technology 12 (2): 1–7.
122 Arfaei, B., Wentlent, L., Joshi, S. et al. (2012). Improving the thermomechan-
ical behavior of lead free solder joints by controlling the microstructure. In:
Proceedings of ITHERM, 392–398.
123 Zeng, K., Stierman, R., Abbott, D., and Murtuza, M. (2006). Root cause of
black pad failure of solder joints with electroless nickel/immersion gold
plating. In: Proceedings of Thermal and Thermomechanical Phenomena in
Electronics Systems, IEEE ITHERM ’06, 1111–1119.
124 Ejim, T.I., Hollesen, D.B., Holliday, A. et al. (1997). Proceedings of the 21st
IEMT, Austin, TX, 25–31.
125 Mei, Z., Callery, P., Fisher, D. et al. (1997). Proceedings Pacific Rim/ASME
Inter. Intersociety Electronic, and Photonic Packaging Conference, Advances in
Electronic Packaging, 1543–1550.
126 Cullen, D.P. (2006) Characterization, reproduction, and resolution of solder
joint microvoiding. Proceedings IPC APEX, S26-2-1.
127 Thein, G., Geiger, D., and Kurwa, M. (2014) Study of various PCBA surface
finishes. Proceedings IPC APEX, S13-03 (March 25–27 2014).
128 The PCB Magazine (2015) Featured content surface finishes. February. http://
iconnect007.uberflip.com/i/457216-pcb-feb2015
129 Shen, C., Hai, Z., Zhao, C. et al. (2017). Packaging reliability effect of ENIG
and ENEPIG surface finishes in board level thermal test under long-term
aging and cycling. Materials https://www.mdpi.com/1996-1944/10/5/451.
130 Chang, D., Bai, F., Wang, Y.P., and Hsiao, C.S. (2004). The study of OSP
as reliable surface finish of BGA substrate. In: Proceedings 6th Electronics
Packaging Technology Conference (EPTC), Singapore, 149–153.
131 Arfaei, B., Anselm, M., Mutuku, F., and Cotts, E. (2014). Effect of PCB sur-
face finish on Sn grain morphology and thermal fatigue performance of
lead-free solder joints. In: Proceedings of SMTAI, Rosemont, IL, 406–414.
132 Arfaei, B., Mutuku, F., Sweatman, K. et al. (2014). Dependence of solder joint
reliability on solder volume, composition and printed circuit board surface
finish. In: 64th Electronic Component and Technology Conference, Orlando,
FL, 655–665.
133 George, E., Osterman, M., Pecht, M., and Coyle, R. (2012) Effects of extended
dwell time on thermal fatigue life of ceramic chip resistors. Proceedings of
IMAPS 2012 45th International Symposium on Microelectronics, San Diego,
CA (September).
134 Coyle, R., Reid, M., Ryan, C. et al. (2009). The influence of the Pb-free solder
alloy composition and processing parameters on thermal fatigue performance
of a ceramic chip resistor. In: Proceedings of Electronic Components Technol-
ogy Conference, Piscataway, NJ, 423–430. IEEE.
135 Manock, J., Coyle, R., Vaccaro, B. et al. (2008). Effect of temperature cycling
parameters on the solder joint reliability of a Pb-free PBGA package. Journal
of Surface Mount Technology 21 (3): 36.
136 Osterman, M., Dasgupta, A., and Han, B. (2006). A strain range based model
for life assessment of Pb-free SAC solder interconnects. In: Proceedings of
Electronic Components and Technology Conference, Piscataway, NJ, 884. IEEE.
137 Bath, J., Sethuraman, S., Zhou, X. et al. (2005) Reliability evaluations of
lead-free SnAgCu PBGA676 components using tin-lead and lead-free SnAgCu
solder paste. Proceedings of SMTAI, Edina, MN, 891.
References 245
138 Clech, J.-P. (2005). Acceleration factors and thermal cycling test efficiency
for lead-free Sn-Ag-Cu assemblies. In: Proceedings of SMTA International,
Chicago, IL, 902–917.
139 Fan, X., Raiser, G., and Vasudevan, V. (2005). Effects of dwell time and ramp
rate on lead-free solder joints in FCBGA packages. In: Electronic Components
and Technology Conference, 901–906.
140 Lee, J. and Subramanian, K. (2003). Effect of dwell times on thermomechani-
cal fatigue behavior of Sn-Ag–based solder joints. Journal of Electronic Materi-
als 32 (6): 523–530.
141 Bartelo, J., Cain, S., Caletka, D. et al. (2001) Thermomechanical fatigue
behavior of selected Pb-free solders. Proceedings of IPC APEX 2001, LF2-2,
Bannockburn, IL.
142 Raj, A., Sridhar, S., Gordon, S. et al. (2018) Long term isothermal aging of
BGA packages using doped lead free solder alloys. Proceedings of SMTAI,
Rosemont, IL (October).
143 Thirugnanasambandam, S., Sanders, T., Evans, J. et al. (2014). Component
level reliability for high temperature power computing with SAC305 and
alternative high reliability solders. In: Proceedings of SMTAI, 262–270.
144 Zhang, Y., Cai, Z., Suhling, J.C. et al. (2009). The effects of SAC alloy com-
position on aging resistance and reliability. In: Proceedings of Electronic
Components and Technology Conference, San Diego, CA, 370–389.
145 Ma, H., Suhling, J.C., Zhang, Y. et al. (2007). The influence of elevated tem-
perature aging on reliability of lead free solder joints. In: Proceedings of
Electronic Components Technology Conference 2007, Reno, NV, 653–668.
146 Ma, H., Suhling, J.C., Zhang, Y. et al. (2006). Reliability of the aging lead free
solder joint. In: Proceedings of Electronic Components Technology Conference,
San Diego, CA, 849–864.
147 Snugovsky, L., Perovic, D., and Rutter, J. (2005). Experiments on the aging of
Sn-Ag-Cu solder alloys. Powder Metallurgy 48: 193–198.
148 Hasnine, M., Mustafa, M., and Suhling, J.C. (2013). Characterization of aging
effects in lead free solder joints using nanoindentation. In: Proceedings of
Electronic Components Technology Conference, 166–178.
149 Chavali, S., Singh, Y., Kumar, P. et al. (2011). Aging aware constitutive
models for SnAgCu solder alloys. In: Proceedings of Electronic Components
Technology Conference, 701–705.
150 Mysore, K., Chan, D., Bhate, D. et al. (2008). Aging-informed behavior of
Sn3.8Ag0.7Cu solder alloys. In: Proceedings of ITHERM, 870–875.
151 Venkatadri, V., Yin, L., Xing, Y. et al. (2009). Accelerating the effects of
aging on the reliability of lead free solder joints in a quantitative fashion. In:
Proceedings of Electronic Components Technology Conference, 398–405.
152 Lee, T.-K., Ma, H., Liu, K.-C., and Xue, A.J. (2010). Impact of isothermal
aging on long-term reliability of fine-pitch ball grid array packages with
Sn-Ag-Cu solder interconnects: surface finish effects. Journal of Electronic
Materials 39 (12): 2564–2573.
153 Wilcox, J., Coyle, R., Lehman, L., and Smetana, J. (2014). Effect of isothermal
preconditioning on thermal fatigue life and microstructure of a SAC305 BGA.
In: Proceedings of SMTAI, Chicago, IL, 122–133.
154 Dompierre, B., Aubin, V., Charkaluk, E. et al. (2001). Cyclic mechanical
behaviour of Sn3.0Ag0.5Cu alloy under high temperature isothermal ageing.
Materials Science and Engineering A 528: 4812–4818.
155 Zhang, Y., Cai, Z., Suhling, J., et al. (2009) Aging effects in SAC solder joints.
Proceedings of the SEM International Congress and Exposition on Experimental
and Applied Mechanics, Albuquerque, NM (June).
156 Bhate, D., Chan, D., Subbarayan, G. et al. (2007). Constitutive behavior of
Sn3.8Ag0.7Cu and Sn1.0Ag0.5Cu alloys at creep and low strain rate regimes.
IEEE Transactions on Components, Packaging, and Manufacturing Technology
31 (3): 621–633.
157 Xiao, Q., Nguyen, L., and Armstrong, W. (2004). Aging and creep behavior
of Sn3.9Ag0.6Cu solder alloy. In: Proceedings of Electronic Components
Technology Conference, Las Vegas, NV, 1325–1332.
158 Xiao, Q., Bailey, H., and Armstrong, W. (2004). Aging effects on microstruc-
ture and tensile property of Sn3.9Ag0.6Cu solder alloy. Journal of Electronic
Packaging 126: 208–212.
159 Mysore, K., Chan, D., Bhate, D. et al. (2008). Aging-informed behavior of
Sn3.8Ag0.7Cu solder alloys. In: Proceedings Thermal and Thermomechanical
Phenomena in Electronic Systems (ITHERM), Florida, FL, 870–875.
160 Kim, D.H., Lee, T.-K., Kim, S.H. et al. (2008). Study on dynamic shock perfor-
mance of SAC305 solder joint after different aging conditions. In: Proceedings
of SMTAI, Orlando, FL, 182–186.
161 Delhaise, A.M., Brillantes, M., Tan, I. et al. (2018) Restoration of microstruc-
ture and mechanical properties of lead-free bismuth containing solder joints
after accelerated reliability testing using a thermal treatment. Proceedings
SMTAI, Rosemont, IL (October).
162 Lutz, J., Hermann, T., Feller, M. et al. (2011) Power cycling induced failure
mechanism in the viewpoint of rough temperature environment. 5th Inter-
national Conference on Integrated Power Electronics Systems, Nuremberg,
Germany (11–13 March).
163 iNEMI Next generation Solder Materials Workshop. Norman Armendariz
(2019) Defense industry circuit card assembly lead (Pb)-free transition
challenges, Jie Geng, Hongwen Zhang, and Ning-Cheng Lee, “Die attach
soldering,” Richard Coyle, “Recommendations for improvements in reliability
References 247
testing for next generation lead free solder alloys,” and Dock Brown, “Next
generation solder technology”. IPC APEX EXPO 2019 (31 January). www
.inemi.org/solder-workshop-apex-2019-get-presentations (accessed 4 February
2020).
164 JESD22-B111A (2016) Board level drop test method of components for hand-
held electronic products. Arlington, VA: JEDEC.
165 Thukral, V., Zaal, J.J.M., Roucou, R. et al. (2018). Understanding the impact
of PCB changes in the latest published JEDEC board level drop test method.
In: Proceedings of the 2018 IEEE 68th Electronic Components and Technology
Conference, 756–763.
166 Fu, H., Radhakrishnan, J., Ribas, M. et al. (2018) iNEMI project on process
development of Bi-Sn-based low temperature solder pastes – Part IV: com-
prehensive mechanical shock tests on POP components having mixed BGA
BiSn-SAC solder joints. Proceedings of SMTAI, Rosemont, IL (October 2018).
167 JESD22-B103-B (2010) Vibration, variable frequency test method. Arlington,
VA: JEDEC.
168 Wong, S.F., Malatkar, P., Rick, C. et al. (2007). Vibration testing and analy-
sis of ball grid array package solder joints. In: Proceedings of 57th Electronic
Components Technology Conference, Reno, NV, 373–380.
169 Woodrow, T. (2010) NASA-DoD lead-free electronics project: vibration test.
Boeing Electronics Materials and Process Report – 603 Rev. A, 18 November.
170 Mukadam, M., Long, G., Butler, P., and Vasudevan, V. (2005). Impact of
cracking beneath solder pads In printed circuit boards on reliability of ball
grid array packages. In: Proceedings SMTAI, Rosemont, IL, 324–329.
171 Belyakov, S.A., Arfaei, B., Johnson, C. et al. (2019). Phase formation and
solid solubility in high reliability Pb-free solders containing Bi, Sb, or In. In:
Proceedings of SMTAI, Rosemont, IL, 492–508.

7 Lead (PB) - Free Solders For High Reliability and High-Performance Applications

Uploaded by

Copyright:

Available Formats

7 Lead (PB) - Free Solders For High Reliability and High-Performance Applications

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

7 Lead (PB) - Free Solders For High Reliability and High-Performance Applications

Uploaded by

Copyright:

Available Formats

191

Lead (Pb)-Free Solders for High Reliability and

7.1 Evolution of Commercial Lead (Pb)-Free Solder

7.1.1 First Generation Commercial Pb-Free Solders

Table 7.1 Composition and melting range of high-Ag, near

Alloy Nominal composition (wt%) Melting range (∘ C)

SAC387 95.5 3.8 0.7 217

7.1.2 Second Generation Commercial Pb-Free Solders

Figure 7.1 Thermal cycling 100

Failure Rate, F (%)

680 I/O PBGA Thermal Fatigue Reliability

alloy design and comprehensive testing [16]. Consequently, alloy development

Figure 7.5 Data of Syed 350

et al. showing improved 300

7.1.3 Third Generation Commercial Pb-Free Solders

7.2 Third Generation Alloy Research and Development

Figure 7.8 Data

Automotive electronics $142B CAGR + 7%

Automotive packaging CAGR + 11%

SAC305 as solidified SAC305 after thermal cycling

Figure 7.10 Scanning electron micrographs illustrating Ag3 Sn intermetallic precipitate

Figure 7.11 A scanning

Figure 7.12 Data from the Characteristic Lifetime, SAC305

Pb-free Electronics Operating Environment vs. Operational Life-time1

High Equipment Missiles

A&D Risk Reduction

62% Network Servers Industrial 29%

Figure 7.13 Aerospace/defense applications are characterized by harsh use

technologies [54]. Because of concerns regarding the impact of technology

7.2.2 Emergence of Commercial Third Generation Alloys

7.2.2.2 An Expanding Class of 3rd Generation Alloys

parallel in alloy development is found with the class of special-purpose materials

7.2.3 Metallurgical Considerations

Solid Solution Dispersion Figure 7.15 A simple schematic

solution of a solvent-atom lattice invariably produces an alloy that is stronger than

Better Reliability Than SAC Alloy

Figure 7.16 Data of Miric showing Sn3.8Ag0.7Cu3.0Bi1.5Sb0.15Ni (Innolot) outperforms

7.2.3.1 Antimony (Sb) Additions to Tin (Sn)

WEIGHT PER CENT ANTIMONY

49.4 (50) 64.4 (65)

Figure 7.17 The Sn-Sb binary phase diagram.

some contribution might be expected from solid solution strengthening due to Sb

7.2.3.2 Indium (In) Additions to Tin (Sn)

WEIGHT PER CENT TIN

Figure 7.18 The In-Sn binary phase diagram.

7.2.3.3 Bismuth (Bi) Additions to Tin (Sn)

Figure 7.19 The Sn-Bi binary phase diagram.

Multiple studies have shown that Bi improves the mechanical properties of Sn

7.3 Reliability Testing Third Generation Commercial

7.3.1 Thermal Fatigue Evaluations

• At room temperature alloy with α β-Sn

Second generation alloy development was driven by experience gathered

room temperature to 200 ∘ C to compare performance of three proprietary solder

7.3.2 iNEMI/HDPUG Third Generation Alloy Pb-Free Thermal Fatigue

Alloy Nominal composition (wt%) Melting range (∘ C)

and the AREA (Universal Advanced Research in Electronic Assembly) consortia

Figure 7.21 A fully

Table 7.4 Thermal cycling proﬁles used in the iNEMI Third

Thermal Minimum Maximum Temp. range