Beruflich Dokumente
Kultur Dokumente
Review
Department of Microtechnology and Nanoscience, Bionano Systems Laboratory, Chalmers University of Technology, and Sino Swedish Microsystem
Integration Technology (SMIT) Center, Kemivagen 9, Se-412 96, Gothenburg, Sweden
b Key State Lab for New Displays and System Applications and SMIT Center, Shanghai University, Box 282,
Yanchang Road 149, Shanghai 200072, PR China
Received 22 November 2006; received in revised form 1 March 2007; accepted 2 March 2007
Available online 12 March 2007
Abstract
The ternary SnCuCo system eutectic composition was obtained by means of CALPHAD (CALculation of PHAse Diagram) methodology and
it was found to be 0.4% Co and 0.7% Cu (wt%) with a melting point of 224 C. The tensile behavior of this alloy was investigated at different strain
rates (105 , 104 and 103 s1 ) and compared to both Sn37Pb and Sn4.0Ag0.5Cu. The Sn4.0Ag0.5Cu alloy depicts the highest ultimate tensile
strength (UTS) followed by the Sn37P and finally the Sn0.7Cu0.4Co system. The elastic modulus was also higher for the Sn4.0Ag0.5Cu
followed by the Sn0.7Cu0.4Co and last the Sn37Pb. The microstructure of the Sn0.7Cu0.4Co alloy is composed of two types of intermetallic
phases, (Cu,Co)6 Sn5 and (Co,Cu)Sn2 dispersed in a Sn-rich matrix. The microstructure of this alloy proved to be very stable, after aging at 150 C for
24 h. The eutectic Sn0.7Cu0.4Co solder alloy can therefore be a very good alternative for the SAC alloys for surface mount technology applications.
2007 Elsevier B.V. All rights reserved.
Keywords: Tensile properties; Lead-free; Solder alloy; SnCuCo; Strain rate
Contents
1.
2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1. Calculation of eutectic point of the SnCuCo system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2. Solder specimen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3. Test procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.1. Elongation to failure (EF) tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.2. Strain rate change (SRC) tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1. Microstructural analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2. Stability of the microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3. Elongation to failure (TF) tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.1. Strain hardening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4. Strain rate change (SRC) tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.5. Fracture surface analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Corresponding author at: Room A518, Bionano Systems Laboratory, MC2, Chalmers University of Technology, Kemivagen 9, 412 96 Goteborg, Sweden.
Tel.: +46 31 772 3074; fax: +46 31 772 3622.
E-mail address: cristina.andersson@chalmers.se (C. Andersson).
0925-8388/$ see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2007.03.028
98
98
98
98
99
99
99
99
99
100
100
102
103
104
104
105
105
98
1. Introduction
Many lead-free solder alloys have already been proposed as
alternatives for the conventional eutectic Sn37Pb, and many
of them have already been greatly emphasized [13]. Lead-free
solders span from binary to ternary systems, and are generally
based on tin (Sn) with small additions of alloying elements such
as silver (Ag), copper (Cu), zinc (Zn), indium (In), etc. Some
popular binary systems already investigated include SnAg,
SnCu, SnZn, SnBi, and SnIn, while the ternary systems
include SnAgCu, SnAgBi, SnZnBi, among others. As
a replacement for the Sn37Pb solder alloy for surface mount
technology (SMT) applications, the International Electronics
Manufacturing Initiative (iNEMI) recommended in the year of
2000 the ternary Sn3.9Ag0.6Cu (SAC) alloy, based on availability, patent considerations and compatibility issues [4]. The
new lead-free alloys present, however, some disadvantages
compared to the Sn37Pb alloy. Apart from issues related to
physical and mechanical properties, the material cost is also a
disadvantage. The majority of all lead-free solder alloys are more
expensive than the lead containing ones, mainly because many of
these new alloys contain valuable and expensive metals, such as
silver.
The main purpose of this work was to investigate the
tensile properties of the newly developed Sn0.7Cu0.4Co leadfree solder alloy at different strain rates and to compare the
results with both Sn4.0Ag0.5Cu and eutectic Sn37Pb. The
microstructure of this alloy was also thoroughly investigated in
the context of this work.
By using the proposed alloy composition, and by substituting
the very expensive Ag for the much cheaper Co, will result in
cost savings for the whole electronics manufacturing industry.
As shown in Fig. 1, the price of 1 kg of Sn4.0Ag0.5Cu metal
alloy is 2.49 times that of Sn0.7Cu0.4Co, and 3.7 times that
of Sn37Pb. One kg of Sn0.7Cu0.4Co, on the other hand, is
merely 1.5 times that of Sn37Pb. The different element prices,
in USD/kg, used to calculate the alloy prices were the following: 9.23 for Sn, 7.81 for Cu, 359.25 for Ag and 33.63 for
Co [5].
Fig. 2. Binary phase diagram at the Sn-rich corner for CoCu interaction.
2. Experimental
2.1. Calculation of eutectic point of the SnCuCo system
In order to find the eutectic composition of the SnCuCo solder system, the
SnCo binary system was thoroughly assessed using CALPHAD (CALculation
of PHAse Diagram) methodology. The composition of the Sn-rich eutectic point
of the SnCuCo system was found to be 0.4 wt% Co and 0.7 wt% Cu with a
melting temperature of 224 C (see Fig. 2). All the detailed calculations have
been previously published in [6].
99
3. Results
6
8
60
10
12
30
Fig. 4. SEM/EDX elemental mapping of solder matrix in Sn4.0Ag0.5Cu bulk solder alloy (as-solidified): (a) SEM picture of the area analyzed, (b) elemental
distribution of Sn, (c) elemental distribution of Ag, and (d) elemental distribution of Cu.
100
Fig. 5. SEM/EDX elemental mapping in Sn0.7Cu0.4Co bulk solder alloy: (a) SEM picture of the area analyzed, (b) elemental distribution of Sn, (c) elemental
distribution of Co, and (d) elemental distribution of Cu.
Fig. 5 shows the EDX-elemental mapping analysis performed on the microstructure of the Sn0.7Cu0.4Co alloy,
showing the IMCs and including Sn, Cu and Co elements.
According to the EDX analysis, the atomic composition of the
Co-rich particles was Sn:Co:Cu, 70:26:4 at% corresponding to
the CoSn2 , with the substitution of Cu (solid solution) into the
CoSn2 phase, resulting in (Co,Cu)Sn2 . Different Co-rich particles were analyzed and all showed some Cu substitute, in
the range of 2.655.86 at%, inside the CoSn2 phase. The Curich intermetallics showed an atomic ratio between Sn:Cu:Co of
59.5:34.5:6, and can therefore be denoted as the (Cu,Co)6 Sn5
IMC phase with the substitution of Co into the Cu6 Sn5 phase.
The fact that Co does not form any intermetallics with either
Ag or Cu, and that the only intermetallics found between the
Co and Sn were rod-like CoSn2 has also been acknowledged
by other researchers [13,14]. The reason for this behavior was
the fact that Co has little solubility in the -Sn matrix, Ag and
Cu. This behavior was also confirmed by using CALFAD, which
showed that the only two types of stable IMCs that can exist in the
eutectic SnCuCo alloy are the phases Cu6 Sn5 and CoSn2 [5].
According to Fig. 5, it is also difficult to distinguish a shape
difference between the Cu6 Sn5 and the CoSn2 IMCs found in
the Sn0.7Cu0.4Co alloy. In contrast to the SAC alloys, where
there is a visible and evident difference in shape between the
Ag3 Sn and the Cu6 Sn5 IMCs, a chemical analysis is necessary for the SnCuCo alloy in order to identify the different
IMCs.
101
Fig. 6. Optical micrographs at 100 magnification of the microstructure of: as-solidified (a) SAC, (b) Sn0.7Cu0.4Co and (c) Sn37Pb, and after aging at 150 C
for 24 h (d) SAC, (e) Sn0.7Cu0.4Co and (f) Sn37Pb.
Fig. 7. Three engineering stressstrain curves for three Sn0.7Cu0.4Co samples tested at 105 , 104 and 103 s1 , respectively.
Table 2
Maximum load, UTS, yield stress, E-modulus and uniform strain for the three alloys tested
Sn37Pb
Sn4.0Ag0.5Cu
Sn0.7Cu0.4Co
103 s1
104 s1
105 s1
103 s1
104 s1
105 s1
103 s1
104 s1
105 s1
2052
42.7
31.3
20
1697
35.4
26.6
(28)
1225
25.5
22.5
(30)
2271
47.3
33.7
40
1892
39.4
28.8
(43.8)
1585
33
23
(50)
1460
30.4
19.1
37.1
1116
23.4
18.1
(40.7)
861
17.3
10.9
(42.6)
9.8
9.5
8.6
7.1
7.5
102
Table 3
Average UTS, yield stress and E-modulus, and ratios between the three alloys
tested, for the strain rate of 103 s1
Alloy
UTS
Ratio
Yield stress
Ratio
E-Modulus
Ratio
Sn37Pb
Sn4.0Ag0.5Cu
Sn0.7Cu0.4Co
42.7
47.3
30.4
1
1.11
0.71
31.3
33.7
19.1
1
1.08
0.61
20
40
37.1
1
2
1.86
(1)
where is the tensile strength (UTS), the strain rate, m the strain
rate sensitivity (given by the slope of each one of the lines) and
C is a constant. The m values calculated for each alloy were the
following: 0.078 for SAC, 0.10 for Sn0.7Cu0.4Co and 0.112
for Sn37Pb. These values are in good agreement with those
presented by [7], where m values of 0.09 and 0.08 were calculated for the eutectic Sn37Pb and for the Sn3.5Ag0.75Cu,
respectively. The m value describes the capacity to resist necking
[20]. From these values we could then expect that the lead-free
solders would present a lower resistance to necking compared
to the Sn37Pb, however, there are other factors that also affect
(2)
where and are the true stress and the true strain, k the strength
coefficient and n is the strain-hardening exponent which can be
obtained from a loglog plot of true stress versus true strain. The
strain hardening exponent is equal to zero for perfectly plastic
solids, equal to 1 for perfectly elastic solids, and for the majority
of all metals this value is between 0.1 and 0.5.
Table 4 lists the n and k values for the three solder alloys
tested. As shown in Table 4, the highest strain hardening exponent is depicted by the SAC followed by the Sn0.7Cu0.4Co
and last the Sn37Pb. This behavior is in good agreement with
the values of the shear modulus measured, which was higher for
the SAC, followed by the Sn0.7Cu0.4Co and last the Sn37Pb
alloy.
Table 4
Strain hardening exponent n and strength coefficient k for the three alloys tested
Fig. 8. Engineering tensile stress vs. strain rate for the Sn37Pb, SAC and
Sn0.7Cu0.4Co alloys.
103 s1
Sn0.7Cu0.4Co
Sn4.0Ag0.5Cu
Sn37Pb
0.13
26.9
0.26
36.5
0.11
49.5
103
Table 5 lists the stress exponent w and the strain rate sensitivity m for each solder alloy tested, plus the m values obtained
from the EF tests. As shown in Table 5, the m values calculated from the EF tests are slightly lower compared to the ones
obtained from the SRC tests. The stress exponents w for the
three solder alloys are quite high, with values between 7.94 for
the Sn0.7Cu0.4Co and 12.48 for the SAC. The w value for
the Sn0.7Cu0.4Co is very close to that of pure tin, 7.6 [23],
which is a result of the high tin content of the Sn0.7Cu0.4Co
alloy, namely 98.9 wt%.
Fig. 9. True stress vs. true strain data according to the SRC method for the
Sn0.7Cu0.4Co alloy (initial strain rate of 105 s1 ).
Fig. 10. Strain rate vs. true stress obtained from the SRC tests, for the three
alloys tested.
log(2 /1 )
,
log(2 /1 )
(3)
Sn37Pb
Sn4.0Ag0.5Cu
Sn0.7Cu0.4Co
SRC
EF
12.48
7.94
8.73
0.080
0.126
0.115
0.078
0.102
0.112
104
4. Conclusions
For the tensile strain rates investigated, and the test parameters used, the following conclusions can be drawn: the strain rate
has a substantial effect on the measured tensile strength of the
solder material; an increase in strain rate results in an increase
in strength.
For all the strain rates analyzed the strongest alloy is
the Sn4.0Ag0.5Cu, followed by the Sn37Pb and the
Sn0.7Cu0.4Co. The highest E-modulus is also shown by
the SAC followed by the Sn0.7Cu0.4Co alloy and last the
Sn37Pb.
The microstructure of the Sn0.7Cu0.4Co alloy is composed of two different intermetallic phases, the (Co,Cu)Sn2
phase and (Cu,Co)6 Sn5 dispersed in a Sn-rich matrix. The
Sn0.7Cu0.4Co alloy presents a very stable microstructure (for
the aging conditions used).
From the strain rate sensitivity values calculated, it can be
concluded that both lead-free solders present a lower resistance
to necking compared to the Sn37Pb, however, the highest strain
hardening exponent n, which decreases with increased strain
rate, is depicted by the SAC followed by the Sn0.7Cu0.4Cu
and finally the Sn37Pb. These n values also prove that the most
ductile alloy is the Sn37Pb, followed by the SAC and finally
the Sn0.7Cu0.4Co.
The fracture surface of all the alloys analyzed in this work is
dependent of strain rate. With slower strain rate (105 s1 ), the
fracture surface shows a more typical ductile fracture surface
with dimples. There is also an increase in area reduction (at the
necked region) with decreased strain rate. The alloy showing
the largest area reduction is the Sn37Pb, followed by the SAC
and last the Sn0.7Cu0.4Co. This corroborates once again that
the eutectic Sn37Pb is more ductile compared to the other two
alloys.
Acknowledgments
The authors are grateful for the financial support provided
by the EU project, Flex-Eman under the contract no: COOPCT-2003-507983 and the Swedish Foundation for International
Cooperation in Research and Higher Education (STINT), and
last but not least all the SMIT-centre member companies.
References
[1] M. Abtew, G. Selvaduray, Mater. Sci. Eng. 27 (2000) 95141.
[2] J.H.L. Pang, B.S. Xiong, T.H. Low, Int. J. Fatigue 26 (2004) 865872.
[3] C.M.L. Wu, D.Q. Yu, C.M.T. Law, L. Wang, Mater. Sci. Eng. R44 (2004)
144.
[4] http://www.nemi.org/projects/ese/if assembly.html.
[5] http://www.metalprices.com/FreeSite/metals/ag/ag.asp.
[6] L. Liu, C. Andersson, J. Liu, J. Electron. Mater. 33 (9) (2004) 935939.
[7] Zugproben, DIN 50 125, ISO 6892-1984.
[8] Standard methods of tension testing of metallic foil, E345-93, Annual Book
of ASTM standards, vol. 03.01, 1993.
[9] Standard methods of tension testing of metallic materials, E8-93, Annual
Book of ASTM standards, vol. 03.01, 1993.
[10] A. Sharif, Y.C. Chan, M.N. Islam, M.J. Rizvi, J. Alloys Compd. 388 (1)
(2005) 162167.
[11] P.L. Liu, J.K. Shang, J. Electron. Mater. 29 (5) (2000) 622627.
105
[12] K.S. Kim, S.H. Huh, K. Suganuma, Mater. Sci. Eng. A333 (1/2) (2003)
223236.
[13] K.S. Kim, S.H. Hug, K. Suganuma, Microelectron. Reliab. 43 (2003)
259267.
[14] P. Sun, C. Andersson, X. Wei, Z. Cheng, D. Shangguan, J. Liu, Proceedings
of the Conference on High Density Microsystem Design and Packaging and
Component Failure Analysis (HDP06), Shangai, China, 2006.
[15] I. Shohji, T. Yoshida, T. Takahiko, S. Hioki, Mater. Sci. Eng. A366 (2004)
5055.
[16] I. Shohji, T. Yoshida, T. Takahashi, S. Hioki, J. Mater. Sci.: Mater. Electron.
15 (2004) 219223.
[17] T.J. Kilinski, J.R. Lesniak, B.I. Sandor, in: J.H. Lau (Ed.), Solder Joint
ReliabilityTheory and Applications, Van Nostrand Reinhold, NY, 1991.
[18] H. Nose, M. Sakane, Y. Tsukada, H. Nishimura, J. Electron. Pack. 125
(2003) 5966.
[19] W.J. Plumbridge, C.R. Gagg, J. Mater. Sci.: Mater. Electron. 10 (1999)
461468.
[20] F. Lang, H. Tanaka, O. Munegata, T. Taguchi, T. Narita, Mater. Charact.
54 (2005) 223229.
[21] W.D. Callister Jr. (Ed.), Materials Science and EngineeringAn Introduction, third ed., John Willey & Sons, Inc., NY, 1994 (Chapter: Dislocations
and Strengthening Mechanisms, pp. 149180).
[22] J.H. Holloman, Trans. Am. Inst. Min. Metall. Petrol. Eng. 162 (1945)
268272.
[23] R.J. McCabe, M.E. Fine, Metall. Mater. Trans. 33A (2002) 15311539.
[24] F. Zhu, H. Zhang, R. Guan, S. Liu, Proceedings of the IEEE 6th International
Conference on Electronic Packaging Technology, 2005, pp. 466470.