Materials Transactions, Vol. 48, No. 3 (2007) pp. 584 to 593
#2007 The Japan Institute of Metals
EXPRESS REGULAR ARTICLE
Thermal Properties and Phase Stability of Zn-Sn and Zn-In Alloys
as High Temperature Lead-Free Solder
Jae-Ean Lee1; * , Keun-Soo Kim2 , Katsuaki Suganuma2 , Masahiro Inoue2 and Goro Izuta3
1
Department of Adaptive Machine Systems, Osaka University, Suita 565-0871, Japan
Institute of Scientific and Industrial Research, Osaka University, Ibaraki 567-0047, Japan
3
Manufacturing Engineering Center, Mitsubishi Electric Corporation, Amagasaki 661-8661, Japan
2
The potential of newly-designed Zn-xSn (x ¼ 40, 30, and 20 mass%) and Zn-30 mass%In alloys as high temperature lead-free solders was
evaluated, with particular focus on the fundamental thermal properties and phase stability during thermal and humidity exposure. From DSC
results, the melting temperature of Zn-Sn alloys increased with decreasing Sn content, and the final undercooling was about 3 C. The liquid
fraction of the alloys calculated using Scheil’s model is lower than that of the alloys calculated according to the phase diagram by approximately
10 mass% at the eutectic temperature and 250 C. The coefficients of thermal expansion (CTE) of Zn-Sn alloys increased with decreasing Sn
content, i.e. 29:2 106 K1 to 33:2 106 K1 in the temperature range of 50 C to 200 C for Zn-Sn alloys and 31:3 106 K1 in the
temperature range of 50 C to 140 C for Zn-30In alloy. With increasing temperature above eutectic temperature, all alloys began to
deform, indicating the formation of a liquid phase. The thermal deformation of Zn-Sn alloys decreased with increasing Sn content. The
ultimate tensile strength (UTS) and 0.2% proof stress of the as-cast Zn-Sn alloys were almost the same, but the elongation of the as-cast Zn-Sn
alloys decreased with increasing Sn content. After thermal and humidity exposure for 1000 h (85 C/85% Relative Humidity), only the outer
surface of Zn-Sn alloys oxidized. However, Zn-30In alloy rusted quite seriously resulting in Zn oxidation after 1000 h. The UTS and 0.2% proof
stress of Zn-Sn alloy slightly decreased with increasing exposure time. The elongation of Zn-Sn alloys decreased with decreasing Sn content for
100 h exposure. However, the elongation of Zn-Sn alloys showed no further degradation beyond 100 h exposure.
[doi:10.2320/matertrans.48.584]
(Received October 5, 2006; Accepted December 28, 2006; Published February 25, 2007)
Keywords: lead-free solder, zinc-tin alloy, zinc-indium alloy, high temperature solder, thermal and humidity test
1.
Introduction
Advanced active devices have been dramatically developed to meet faster, smaller, lighter, and more functional
requirements. A die attach bonding technology for chip level
packages, one of the device bonding technologies, has been
successfully applied to integrated circuits (ICs), insulated
gate bipolar transistors (IGBTs), light emitting diodes
(LEDs), quad flat packages (QFPs), and many other active
components. These devices are based on single- or multi-chip
Si dies mounted in complex structures comprising several
materials such as metallic electrodes, ceramics, metallic heat
spreaders, and solders. These devices are exposed to thermomechanical fatigue environments. Moreover, mismatches in
the materials properties, such as the coefficients of thermal
expansion (CTEs) and the elastic moduli, were found to
induce thermal stress. To overcome these severe conditions,
good high temperature solders, which are used as die-attach
materials, are required to be soft for the relaxation of thermal
stress, and should be thermally conductive for effective heat
dissipation.
Pb-Sn alloys, containing over 85 mass%Pb,1–3) have been
commonly used in die attach bonding as high temperature
solders. The high Pb content in these solders invariably
makes them subject to RoHS restrictions; therefore the
development of lead-free solders is a target technology.
Several high temperature lead-free solder candidates exist,
including peritectic Sn-Sb, eutectic Au-Si, Au-Sn, and Bi-Ag
alloys.4–12) Peritectic Sn-Sb alloy however, has a somewhat
low melting temperature (about 235–240 C), and the toxicity
of Sb is of some concern. Of similar concern is Bi-Ag alloy,
*Corresponding
author, E-mail: jelee@eco.sanken.osaka-u.ac.jp
which retains the very brittle nature associated with Bi. For
the Au-based alloys, Au-Si alloy is very expensive, while AuSn alloy has a very brittle nature due to the formation of
massive intermetallic compounds (IMCs). Moreover, the Aubased alloys have been known to form voids, resulting in
delamination at the interface between the Si die and the
solder.5,9) For these reasons, there is a need to develop new
types of affordable and reliable high temperature lead-free
solder materials.
Recently, the authors proposed a new guideline to establish
high temperature lead-free solders.13) In this guideline, the
solder alloy can form some liquid phase in a secondary reflow
process, at 250 C, while the volume of the liquid phase
should be controlled so as not to distort the bonding structure.
Based on this guideline, two alloy systems, Zn-Sn and Zn-In,
have become prime candidates as high temperature lead-free
solders. These alloys do not form IMCs over their entire
range of composition and temperature. In addition, Zn-Sn
alloy is relatively inexpensive. These alloys are expected to
exhibit good mechanical, electrical, and thermal conductivity
properties. In contrast to the fascinating nature of Zn-based
alloys, the poor oxidation resistance of Zn can be a major
drawback when the die attached layers are exposed to the
environment. Therefore, the objective of the present work is
to evaluate the potential of Zn-Sn and Zn-In alloys for
application as high temperature lead-free solders. The
thermal properties and phase stability during thermal and
humidity exposure are mainly examined.
2.
Experimental Procedures
Zn-xSn (x ¼ 40, 30, and 20 mass%) and Zn-30 mass%In
binary alloys were provided by Nihon Genma MFG co. Ltd.
Thermal Properties and Phase Stability of Zn-Sn and Zn-In Alloys as High Temperature Lead-Free Solder
The chemical compositions of starting materials (mass%).
Composition
Components
Zn-40Sn
Zn-30Sn
Zn-20Sn
Zn-30In
Sn
40.880
30.810
20.950
—
Pb
0.018
—
—
—
Ag
0.004
0.003
0.001
0.003
Cu
—
0.002
—
—
Bi
—
—
—
0.001
Zn
Cd
59.097
0.001
69.184
0.001
79.048
0.001
69.916
—
In
—
—
—
30.080
400
Temperature, T /°C
Table 1
300
200
100
0
The chemical compositions of the alloys are listed in Table 1.
Hereafter, the composition unit ‘‘mass%’’ is omitted. The
thermal analysis of the alloys was carried out using differential scanning calorimetry (DSC) in the temperature range
from room temperature to 400 C in Ar at a gas flow rate of
30 mL/min, and a constant heating and cooling rate of 3 C/
min. The liquid fraction change of the alloys during solidification was estimated by the calculations based on Scheil’s
model using CALPHAD software (Pandat) with the solder
database (ADAMIS). The alloys were measured to verify the
CTE and thermal deformation in the temperature ranges from
50 C to 300 C using a thermal mechanical analyzer
(TMA). Before measuring the TMA, a calibration was
performed using a standard reference material (SRM 736) of
copper in the form of a rod, 6.4 mm in diameter and 51 mm
long. The measurement for the CTE and thermal deformation
of the alloys was performed using the dilatometric method,
under a load of 9.8 mN.
Tensile specimens of the alloys (46 mm in length, 8 mm
wide and 1 mm thick) were prepared by casting the alloys at
500 C for 30 min into a steel mold. The cooling rates of ZnSn and Zn-30In ingots were measured using a K-type
thermocouple until solidification and these were approximately 28 C/s and 22 C/s, respectively. The cooling
curve is shown in Fig. 1. The specimen surfaces were
mechanically polished using 0.3 mm alumina powder and
then exposed to two condition atmospheres, i.e. 85 C/85%
relative humidity (RH) for 100 h, 250 h, 500 h, and 1000 h
and heat-treatment of 85 C for 100 h. The tensile strength
of each specimen was measured by a tensile test at a strain
rate of 3:5 104 s1 at room temperature. One data point
was obtained from the average value of ten specimens. A
focused ion beam (FIB) was employed for the processing of
the cross-section microstructures of Zn-Sn alloys. The
microstructures of the alloys were observed by scanning
electron microscopy (SEM). The element analysis and phase
identification of the alloys were carried out by using energy
dispersive X-ray spectroscopy (EDS) and X-ray diffraction
analysis (XRD).
3.
Results and Discussion
3.1 Microstructures and thermal properties of alloys
Figure 2 shows the microstructure of the as-received
alloys. In the case of Zn-Sn alloys, the dark and bright color
phases are the primary -Zn and eutectic -Sn/-Zn phases,
585
0
5
10
15
20
25
Time, t /s
Fig. 1
Cooling curves of alloys during die casting into a steel mold.
respectively. The primary -Zn phases were surrounded by
eutectic -Sn/-Zn phases, in which fine eutectic -Zn
platelets were found dispersed in a eutectic -Sn matrix. A
dimension fraction of the primary -Zn phases alloys
increased while the eutectic -Sn/-Zn phase decreased
with decreasing Sn content, as expected from the phase
diagram,14) in which the eutectic composition is Sn-8.8Zn.
On the other hand, the eutectic -Zn platelets in Zn-30In
alloy could not be found in the corresponding a eutectic -In
matrix. However, it was confirmed from the DSC results that
these alloys are composed of the primary -Zn and eutectic
-In/-Zn phases.
DSC analysis was carried out in order to understand the
thermal reaction properties of alloys on heating and cooling,
as shown in Fig. 3(a) and (b). These hypereutectic alloys
show two endothermic peaks: one peak appears at 200 C for
Zn-Sn alloys and at 145 C for Zn-30In alloy, while the other
peak ranges from 365 C to 383 C. Each endothermic peak
corresponds well to the melting temperature of the hypereutectic alloy. The lower and higher temperature peaks
correspond to the eutectic and melting temperatures, respectively. The undercooling temperatures of all alloys were
about 4–9 C for the higher temperature peak and 3 C for the
lower temperature peak. The undercooling temperature of the
alloys is lower than those of the Sn based alloys, i.e. about
20–30 C for Sn-3Ag, Sn-0.7Cu, and Sn-Ag-Cu alloys.15–17)
These DSC results coincided with those expected for the
binary phase diagrams of the alloys.14)
Figure 4 shows the change in liquid fraction of the alloys
as a function of composition during solidification. The liquid
fraction of Zn-Sn alloys decreased with decreasing Sn
content. The melting and eutectic temperatures calculated
using Scheil’s model agree well with the DSC results.
Table 2 compares the liquid fraction of the alloys calculated
using both the phase diagram and Scheil’s model on the
eutectic isotherm and at the secondary reflow temperature
(250 C), respectively. The liquid fraction of Zn-Sn alloys
calculated using the phase diagram and Scheil’s model
decreased to about 47% to 24%, and 35% to 15% with
decreasing Sn content at 250 C, respectively. The liquid
fraction of the alloys calculated using Scheil’s model is low,
at approximately 10%, as compared with that of the alloys
586
J.-E. Lee, K.-S. Kim, K. Suganuma, M. Inoue and G. Izuta
(a)
α-Zn
Zn platelets
(b)
α-Zn+β-Sn
(c)
(d)
α-Zn
α-Zn+β-In
20 µm
Fig. 2 Microstructures of as-received alloys formed at a cooling rate of 3 C/min; (a) Zn-40Sn, (b) Zn-30Sn, (c) Zn-20Sn, and (d) Zn-30In
alloys.
150
200
250
300
350
Endothermic
(a)
400
400
365 °C
374 °C
383 °C
373 °C
Zn-40Sn
Zn-30Sn
Zn-20Sn
Zn-30In
200 °C
145 °C
360
Temperature, T /°C
100
320
280
Zn-30Sn
R1 R2R4 R3
240
Zn-20Sn
E1
200
160
(b)
Zn-40Sn
Zn-30In
E2 E3
E4
120
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
197 °C
Exothermic
Mass fraction of liquid
359 °C
142 °C
100
150
Zn-40Sn
367 °C
379 °C
364 °C
200
250
300
Zn-30Sn
Zn-20Sn
Zn-30In
350
400
Temperature, T /°C
Fig. 3
Fig. 4 The temperature vs. mass fraction of liquid during solidification
of alloys calculated from Scheil’s model; E and R indicate the liquid
fraction of alloys on the eutectic isotherm and at 250 C, respectively.
DSC curves of alloys during (a) heating and (b) cooling.
calculated according to the phase diagram. Thus, it is very
useful to estimate the liquid fraction using the solidification
simulation for the die attach bonding.
Thermal Properties and Phase Stability of Zn-Sn and Zn-In Alloys as High Temperature Lead-Free Solder
Table 2 The liquid fraction of alloys calculated using phase diagrams and
Scheil’s model.
Alloys
Zn-40Sn
Zn-30Sn
Liquid
fraction
Zn-20Sn
Zn-30In
Phase diagram
(mass%)
Scheil’s equation
(mass%)
E1
44
31
R1
47
35
E2
33
22
R2
35
25
E3
22
14
R3
24
15
E4
31
20
R4
34
22
Point#
#
pointing out in the Fig. 4. E on eutectic isotherm. R at 250 C for
secondary reflow temperature.
Thermal deformation, ∆ δ /µm
100
0
Zn-In alloy
eutectic temp.
-100
Zn-Sn alloy
eutectic temp.
-200
Zn-40Sn
Zn-30Sn
Zn-20Sn
Zn-30In
-300
-50
0
50
100
150
200
250
300
Temperature, T /°C
Fig. 5 The thermal deformation curves of each alloy in the temperature
range between 50 C and 300 C.
Zn-40Sn
Zn-30Sn
Table 3
The coefficient of thermal expansion (CTE) of alloys.
Alloys
Temperature range
CTE, /106 K1
Zn-40Sn
50 200 C
29.2
Zn-30Sn
50 200 C
31.7
Zn-20Sn
50 200 C
33.2
Zn-30In
50 140 C
31.3
Figure 5 shows the thermal deformation curves of the
alloys in the temperature range from 50 C to 300 C. Here,
Zn-Sn and Zn-30In alloys linearly expanded in the temperature range from 50 C to the eutectic temperature. Thus,
the CTE of each alloy is calculated in the temperature range
of 50 C to 200 C for Zn-Sn alloys and 50 C to 140 C
for Zn-30In alloy. The results are summarized in Table 3.
The CTEs of Zn-Sn alloys increased from 29:2 106 K1
to 33:2 106 K1 with decreasing Sn content. The CTE of
Zn-30In alloy, 31:3 106 K1 , was similar to that of Zn30Sn alloy. The CTE of Zn-Sn and Zn-30In alloys are
slightly higher than that of Pb-5Sn alloy, i.e. 28:7 106 K1 .3)
In contrast, the thermal conductivity of the Zn-based alloys
is expected to be relatively higher around 100 Wm1 K1 ,
as compared with 35 Wm1 K1 for Pb-5Sn and 9
Wm1 K1 for Bi-11Ag alloys.8,18) On the other hand, all
alloys began to deform from the eutectic temperature,
indicating the formation of a liquid phase. The deformation
of Zn-Sn alloys gradually decreased with decreasing Sn
content. The deformation of Zn-20Sn and Zn-30In alloys is
slight however, even though the volume fractions of the
corresponding liquid phases were 15% and 22% at 250 C,
respectively.
3.2 Oxidation resistance of alloys
Figure 6 shows the surface microstructures of Zn-Sn and
Zn-20Sn
α-Zn
0h
587
Zn-30In
α-Zn
α-Zn+β-Sn
100h
1000h
20 µm
Fig. 6
Top surface microstructures of alloys after 85 C/85%RH exposure.
α-Zn+β-In
J.-E. Lee, K.-S. Kim, K. Suganuma, M. Inoue and G. Izuta
2Zn þ O2 (g) ¼ 2ZnO;
4/3In þ O2 (g) ¼ 2/3In2 O3 ;
Sn þ O2 (g) ¼ SnO2 ;
2Sn þ O2 (g) ¼ 2SnO;
Go85 C
Go85 C
Go85 C
Go85 C
Intensity (Arb. units)
1000h
100h
ZnO
As cast
Zn
Sn
20
(b)
30
40
50
60
30
40
50
60
1000h
In2O3
100h
ZnO
As cast
Zn
In
20
2 θ (Cu Kα)
Fig. 7 X-ray diffraction of alloys after 85 C/85%RH exposure; (a) Zn30Sn and (b) Zn-30In alloys.
ð1Þ
¼ 628:906 kJ ð2Þ
¼ 541:008 kJ ð3Þ
¼ 503:372 kJ ð4Þ
¼ 492:292 kJ ð5Þ
Go of ZnO is the smallest among all oxides. Thus, the ZnO
phase is the most stable oxide followed in turn by In2 O3 ,
SnO2 , and SnO.
Several mushroom-like nodules and cracks were detected
in Zn-30In alloy, even after 10 h as shown in Fig. 9. From the
EDS analysis, the protruding mushroom-like nodules consisted of the eutectic -In/-Zn phases. This phenomenon is
very similar to the previous report by Kim et al.19) Such
nodule formation seems to be attributed to a stress relaxation
mechanism after oxidation. A ZnO phase was observed to
form along the Zn grain boundaries, which is considered to be
•
xM(s) þ 1/2yO2 (g) ¼ Mx Oy (s)
(a)
Intensity (Arb. units)
Zn-In alloys after 85 C/85%RH exposure for 0 h, 100 h, and
1000 h. In the initial microstructures, the primary -Zn phase
formed fine dendrite morphology. In the case of Zn-Sn alloys,
fine eutectic -Zn platelets were occasionally observed in the
eutectic -Sn matrix. It is considered that the eutectic -Zn
platelets became much smaller in a eutectic -Sn matrix
owing to the large nucleation driving force cased by the rapid
cooling rate. After 100 h exposure, the surface microstructure
of Zn-Sn alloys showed no significant microstructural
change, as compared with the initial microstructure. On the
other hand, Zn-30In alloy exhibited a dramatic change, i.e. a
very serious degradation after 100 h. The surface microstructure of Zn-30In alloy revealed the formation of
numerous mushroom-like nodules and cracks. Moreover,
after 1000 h exposure, Zn-30In alloy was completely rusted
due to oxidation. On the other hand, despite increasing the
exposure time to 1000 h, the surface of Zn-Sn alloys again
showed no change. Thus, from the observed results of the
surface microstructures, Zn-Sn alloys were verified to have
superior oxidation resistance under the thermal and humidity
condition, as compared with Zn-30In alloy.
XRD analysis was employed in order to identify the
reaction product on the surface of the alloys after thermal and
humidity exposure. Figure 7 shows the XRD results of Zn30Sn and Zn-30In alloys. In the early stages of exposure, Zn30Sn and Zn-30In alloys showed no visible signs of oxide
formation. After 100 h exposure, however, the ZnO phase
was detected in the surfaces of Zn-30Sn and Zn-30In alloys,
although the peak intensity of the ZnO phase for Zn-30Sn
alloy was relatively lower than that for Zn-30In alloy.
Independent of alloy composition, following 1000 h exposure, the oxide was detected only the ZnO phase in Zn-Sn
alloys. On the other hand, the In2 O3 phase was detected in
Zn-30In alloy after 1000 h exposure.
In order to understand the oxidation reaction of each
element (In, Sn, and Zn), Figure 8 shows the Gibbs free
energy changes (Go ) in the oxidation reactions of Sn, In and
Zn in the temperature range from room temperature to
100 C. Here, it is observed that the as-formed oxide becomes
more stable with decreasing Go , where Go for each
element was calculated on a molar O2 basis at 85 C as
follows;
Gibbs free energy, ∆G / kJ mole O2-1
588
-450
2Sn + O2(g) =
-500
Sn + O2(g) =
-550
2SnO
SnO2
= 2/3In2O3
4/3In + O2(g)
-600
2ZnO
2Zn + O2(g) =
-650
-700
0
20
40
60
80
100
Temperature, T /°C
Fig. 8
Ellingham diagram of each element with oxygen.
Thermal Properties and Phase Stability of Zn-Sn and Zn-In Alloys as High Temperature Lead-Free Solder
589
Free surface
a
α-Zn+β-Sn
ZnO
In
In InIn
In Zn
0
1
2
3
Zn
Point-a
Intensity (CPS)
Intensity (CPS)
In
α-Zn
4
Zn
5
6
7
8
9 10
Energy, E/KeV
Zn
Sn
Sn
0
1
2
3
4
Zn
5
6
7
8
9 10
Energy, E/KeV
Fig. 9 Top surface microstructure and EDS point analysis of Zn-30In alloy
after 10 h exposure.
Fig. 10 Cross section microstructures processed by FIB and EDS
qualitative analysis of Zn-20Sn alloy after 1000 h exposure.
the source of compressive stresses on the soft eutectic -In/
-Zn phases. From XRD results, the ZnO phase was verified
to have a Wurtzite-type structure. The hexagonal close
packed (hcp) crystal structure of the ZnO phase is similar to
that of the Zn phase. When the Zn phase is oxidized to the
ZnO phase, the O atoms locate at the interstitial sites between
the Zn atoms. Therefore, the a and c axes of the Zn unit cell
expand from 2:66 1010 m to 3:25 1010 m, and 4:95 1010 m to 5:21 1010 m, respectively. Their lattice unit
volume expands from 3:04 1029 m3 to 4:78 1029 m3 .
This volume change in the lattice unit swells above 50 vol%
of the Zn lattice. This lattice volume expansion results in the
formation of a number of mushroom-like nodules and cracks
during relaxation of the compressive stress. Most importantly, the softness of the In accelerates the Zn oxidation of Zn30In alloy due to its large deformation.
Figure 10 shows the cross-section microstructure afforded
by FIB, and the EDS qualitative analysis of Zn-20Sn alloy
after 1000 h exposure. In these Zn-Sn alloys, there is no trace
of oxidation except for the free surface. It was found that ZnSn alloys display an excellent oxidation resistance even in
severe thermal and humidity conditions following 1000 h
exposure. It has been reported that both Zn and Zn-based
alloys exhibit excellent resistance to atmospheric corrosion
and corrosion in most natural waters.20) When thin oxide
protective layers completely cover the free surface of the
metal, the corrosion proceeds at a greatly reduced rate.20)
According to Anderson’s report, a series of 20-year exposure
tests carried out by the American Society for Testing and
Materials (ASTM), quotes the following average rates of
Zn corrosion in various environments, i.e. 0.0064 mmyr1 in
an industrial environment, 0.0015 mmyr1 in the coastal
regions, 0.0011 mmyr1 in the rural conditions, and
0.00018 mmyr1 in an arid atmospheres.21) In the present
work, Zn-Sn alloys contain a significant amount of Zn (over
60 mass% Zn) and the Sn do not serious deformation, and as
such, it is expected that the oxidation resistance of Zn-Sn
alloys improves as the Zn content increases.
Figure 11 shows the cross-section microstructures and
oxidation distance of Zn-30In alloy during thermal and
humidity exposure. The oxidation of Zn-30In alloy preferentially takes place along the Zn grain boundaries, and
proceeds deep within the alloy. The oxidation distance of Zn30In alloy increases with increasing exposure time, from
approximately 15 mm for 10 h to 390 mm for 100 h exposure.
3.3 Tensile properties of alloys
Figure 12 shows the typical nominal stress-strain curves
for the alloys. Here, the stress-strain curves of all alloys
exhibited ductile deformation, while their tensile strength
remained essentially the same. On the other hand, the
elongation of Zn-Sn alloys decreased with decreasing Sn
content. It can be said that the elongation of Zn-Sn alloys is
governed by the amount of Sn content. Meanwhile, both the
tensile strength and elongation of Zn-30In alloy was
significantly lower than those of Zn-Sn alloys.
590
J.-E. Lee, K.-S. Kim, K. Suganuma, M. Inoue and G. Izuta
80
d
Nominal stress, σ /MPa
(a)
(b)
Zn-40Sn
Zn-30Sn
Zn-20Sn
Zn-30In
Bi-2Ag
70
60
50
40
30
20
10
0
0
5
10
15
20
Nominal strain, ε /%
25
30
Fig. 12 Representative nominal stress-strain curves of alloys.
Oxidation distance, d/µm
(c)
400
Zn-30In
300
200
100
0
0
25
50
75
100
125
Exposure time, t /hr
Fig. 11 Cross section microstructures and oxidation distance of Zn-30In
alloy after (a) 10 h, (b) 25 h, and (c) 50 h exposure.
Figure 13 shows the fracture surface of the alloys after the
tensile tests. The eutectic -Sn/-Zn or -In/-Zn phases of
the alloys certainly plays a key role in the ductile deformation.
Figure 14 shows the tensile properties of the alloys during
thermal and humidity exposure. Figures 14(a), (b), and (c)
indicate the UTS, 0.2% proof stress, and elongation of the
alloys, respectively. The UTS and 0.2% proof stress of Zn-Sn
alloys slightly decreased with increasing exposure time.
After 100 h exposure, the elongation of Zn-Sn alloys abruptly
decreased, especially for Zn-20Sn. However, the elongation
of Zn-Sn alloys showed no further degradation beyond 100 h
exposure. It was already confirmed that the oxidation of ZnSn alloys occurs only on the surface region, and does not
proceed to the inside of the alloys beyond 1000 h exposure.
Moreover, the elongation of Zn-Sn alloys exposed for 1000 h
is still larger than that observed for the as-cast Bi-2Ag alloy,
which shows a brittle fracture pattern, as seen in Fig. 12.
In contrast to Zn-Sn alloys, Zn-30In alloy collapsed after
100 h exposure, revealing serious degradation (Figure 6),
such that no tensile test could be performed.
In order to investigate the effect of humidity on the
decreasing of the elongation, a thermal exposure test was
carried out at 85 C for 100 h. The tensile properties of the
heat-treated alloys (shown in Fig. 15) are compared with
those of the as-cast alloys and the alloys exposed to 85 C/
85%RH for 100 h. The tensile properties of the heat-treated
Zn-Sn alloys were similar to those of the Zn-Sn alloys tested
after thermal and humidity exposure. On the other hand, the
elongation degradation of the heat-treated Zn-30In alloy was
much smaller than that of Zn-30In after thermal and humidity
tests. From the results of the tensile tests, the degradation
origin of Zn-Sn alloys indicates the volume expansion
afforded by the CTE of each Zn-Sn alloy during exposure at
85 C. Thus, the volume expansion of the alloys seems to
play a key role in the deterioration of elongation because the
decreasing Sn content gradually diminishes the buffer ability
of Sn against the volume expansion. Therefore, the elongation deterioration of the alloys is influenced by the temperature induced volume expansion. Especially, it was found
that the mechanical properties of Zn-30In alloy are strongly
affected by the humidity rather than the temperature.
4.
Conclusions
In the present work, Zn-xSn (x ¼ 40, 30, and 20 mass%)
and Zn-30 mass%In alloys are evaluated as new high
temperature lead-free solders, with particular focus on the
fundamental thermal properties and phase stability during
Thermal Properties and Phase Stability of Zn-Sn and Zn-In Alloys as High Temperature Lead-Free Solder
591
(b)
(a)
α-Zn
α-Zn+β-Sn
(c)
(d)
α-Zn+β-In
α-Zn
Fig. 13 Fracture surface of as-cast alloys after tensile tests.
thermal and humidity exposure. The results are summarized
as follows:
(1) The microstructure of the alloys is consisted of the
primary -Zn and eutectic -Sn/-Zn phases for Zn-Sn
alloys, and the primary -Zn and eutectic -In/-Zn
phases for Zn-30In alloy.
(2) The alloys show two endothermic and exothermic
peaks. The lower and higher endothermic peaks
indicate the eutectic and liquidus temperatures of the
alloys, respectively. The undercooling of the final
solidification is negligible, only 3 C.
(3) The liquid fraction of the alloys calculated using
Scheil’s model is lower than that of the alloys
calculated according to the phase diagram by approximately 10 mass% at the eutectic temperature and
250 C.
(4) The CTEs of Zn-Sn alloys increased from 29:2 106 K1 to 33:2 106 K1 with decreasing Sn
content. The CTE of Zn-30In alloy is 31:3 106 K1 .
The deformation of the alloys occurs at the eutectic
temperature, where the deformation change of Zn-Sn
alloy decreases with decreasing Sn content.
(5) Only the surface of Zn-Sn alloys was oxidized after
1000 h exposure under 85 C/85%RH conditions. A
number of mushroom-like nodules and cracks were
formed on the upper surface of Zn-30In alloy as a result
of the volume expansion by Zn oxidation.
(6) The stress-strain curves of the alloys revealed ductile
deformation in the early stages. The UTS and 0.2%
proof stress of Zn-Sn alloys were almost similar, while
the elongation of Zn-Sn alloys decreased with decreasing Sn content.
(7) After thermal and humidity exposure, the UTS and
0.2% proof stress of Zn-Sn alloys slightly decreased
with increasing exposure time. The elongation of Zn-Sn
alloys abruptly decreased with increasing Sn content
following 100 h exposure. However, the elongation of
Zn-Sn alloys showed no further degradation until
1000 h exposure.
(8) The elongation of Zn-Sn and Zn-30In alloys was
strongly affected by the temperature and humidity
conditions, respectively.
We found that Zn-Sn alloys exhibit excellent oxidation
resistance and good mechanical properties under severe
thermal and humidity conditions, and as such, they have good
potential in die attach bonding. The CTE values of Zn-Sn
alloys are somewhat higher than those of the high temperature Pb-Sn alloys, but it is expected that these values will be
overcome by excellent heat dissipation. Also, in order to
overcome oxidation under severe conditions, it is recommended that Zn-30In alloy is appropriately designed or
encased in a protective molding in order to avoid reaction
with moisture. Further work is required to evaluate the
reliability of these alloys in thermal fatigue tests in order to
592
J.-E. Lee, K.-S. Kim, K. Suganuma, M. Inoue and G. Izuta
100
100
(a)
(a)
80
40
Zn-40Sn
Zn-30Sn
Zn-20Sn
Zn-30In
20
0
100
UTS, σ /MPa
60
0.2% proof stress, σ 0.2 /MPa
0.2% proof stress, σ 0.2 /MPa
UTS, σ /MPa
80
(b)
80
60
40
20
0
40
60
40
20
0
100
(b)
80
60
40
20
0
40
(c)
Elongation, δ /%
Elongation, δ /%
(c)
30
20
10
0
as cast alloys
85°C/85%RH - 100h
85°C - 100h
30
20
10
0
0
500
1000
Exposure time, t /hr
Fig. 14 Tensile properties of alloys as a function of exposure time in
85 C/85%RH condition; (a) ultimate tensile strength, (b) 0.2% proof
stress, and (c) elongation.
determine their capacity as reliable high temperature leadfree solder alloys for application in die attach bonding.
Acknowledgment
This work was partly supported by the NEDO Project
Zn-40Sn Zn-30Sn Zn-20Sn Zn-30In
Fig. 15 Tensile properties of alloys under three different conditions; (a)
ultimate tensile strength, (b) 0.2% proof stress, and (c) elongation.
‘‘R&D of high temperature lead-free soldering’’ and the 21st
century COE program of The Japan Ministry of Education,
Culture, Sports, Science and Technology in 2006. The
authors would also like to acknowledge Nihon Genma MFG
Co. Ltd. for providing the alloys.
Thermal Properties and Phase Stability of Zn-Sn and Zn-In Alloys as High Temperature Lead-Free Solder
REFERENCES
1) K. S. Kim, C. H. Yu, N. H. Kim, N. K. Kim, H. J. Chang and E. G.
Chang: Microelectron. Reliab. 43 (2003) 757–763.
2) W. D. Zhuang, P. C. Chang, F. Y. Chou and R. K. Shiue: Microelectron. Reliab. 41 (2001) 2011–2021.
3) Y. Nagatomo, T. Nagase and S. Shimamura: J. Jpn. Inst. Electron.
Packaging 3 (2000) 330–334.
4) B. F. Wolffenbuttel: Sensors and Actuators A 62 (1997) 680–686.
5) J. W. Jang, S. Hayes, J. K. Lin and D. R. Frear: J. Appl. Phys. 95 (2004)
6077–6081.
6) H. G. Song, J. P. Ahn and J. W. Morris, Jr: J. Electron. Mater. 30 (2001)
1083–1087.
7) J. W. R. Tew, X. Q. Shi and S. Yuan: Mater. Lett. 58 (2004) 2695–
2699.
8) J. N. Lalena, N. F. Dean and M. W. Weiser: J. Electron. Mater. 31
(2002) 1244–1249.
9) S. Terashima, T. Uno, E. Hashino and K. Tatsumi: Mater. Trans. 42
(2001) 803–808.
10) J. H. Kim, S. W. Jeong and H. M. Lee: Mater. Trans. 43 (2002) 1873–
1878.
593
11) S. S. Kim, J. H. Kim, S. W. Booh, T. G. Kim and H. M. Lee: Mater.
Trans. 46 (2005) 2400–2405.
12) M. Yokozawa, K. Hidaka, T. Ogashiwa and T. Arikawa: J. Jpn. Inst.
Electron. Packaging 3 (2000) 617–620.
13) J. E. Lee, K. S. Kim, K. Suganuma, J. Takenaka and K. Hagio: Mater.
Trans. 46 (2005) 2413–2418.
14) T. B. Massalski: Binary Alloy Phase Diagrams, 2nd ed. (ASM
International, 1992).
15) Y. S. Kim, K. S. Kim, C. W. Hwang and K. Suganuma: J. Alloys
Compd. 352 (2003) 237–245.
16) K. S. Kim, S. H. Hur and K. Suganuma: Microelectron. Reliab. 43
(2003) 259–267.
17) S. K. Kang, D. Y. Shih, D. Leonard, D. W. Henderson, T. Gosselin, S. I.
Cho, J. Yu and W. K. Choi: JOM 56 (2004) 34–38.
18) E. A. Bransdes and G. B. Brook: Smithells Metals Reference Book, 7th
Ed. (Butterworth Heinemann press, British, 1992).
19) K. S. Kim, T. Matsuura and K. Suganuma: J. Electron. Mater. 35 (2006)
41–47.
20) L. L. Shreir: Corrosion vol 1 Metal/Environment Reactions, 2nd Ed.
(Newnes Butterworths, London and Boston 1976).
21) E. A. Anderson: ASTM Special Publication No. 175 1955.