May 28, 2008 - Melting range, microstructure, mechanical properties and spreadabililty of Zn-(4$6 mass%)Al-(1$6 mass%)Cu alloys were investigated.
Materials Transactions, Vol. 49, No. 7 (2008) pp. 1531 to 1536 Special Issue on Lead-Free Soldering in Electronics IV #2008 The Japan Institute of Metals
Characteristics of Zn-Al-Cu Alloys for High Temperature Solder Application Seong-Jun Kim1; * , Keun-Soo Kim2 , Sun-Sik Kim2 , Chung-Yun Kang3 and Katsuaki Suganuma2 1
Department of Adaptive Machine Systems, Graduate School of Engineering, Osaka University, Suita 567-0871, Japan Institute of Scientific and Industrial Research, Osaka University, Ibaraki 567-0047, Japan 3 Department of Materials Engineering, Pusan National University, 30 Jangjeon, Geumjeong, Busan, 609-735, Korea 2
Melting range, microstructure, mechanical properties and spreadabililty of Zn-(46 mass%)Al-(16 mass%)Cu alloys were investigated. Liquidus temperature was targeted between 655 and 675 K, and solidus temperature was targeted to 645 K. The liquidus temperature of the Zn-Al-Cu solders increased with Cu contents, but it decreased with Al contents. Microstructures of the Zn-Al-Cu solders consisted of primary "-phase (CuZn4 ), -phase (Zn matrix), - eutectic phase (Zn-Al eutectic) and "- eutectic phase (Zn-Cu eutectic), irrespective of the Al and Cu contents. Increasing the Al and Cu contents, hardness and tensile strength increased, but elongation decreased. The Al content played an important role in improving the spread ratio, the Cu content had no significant influence on the spread ratio. [doi:10.2320/matertrans.MF200809] (Received January 9, 2008; Accepted April 10, 2008; Published May 28, 2008) Keywords: zinc-aluminum-copper, lead-free solder, high temperature, melting range, microstructure, tensile property, spreadability
1.
Introduction
The application of power electronics has been extended to a variety of automotive, aerospace, and energy production industries.1,2) With the miniaturization and increasing power of power electronics, high temperature operation has become a serious issue. As a response to the growing demand of high temperature operation, next generation power semiconductors such as SiC and GaN, and packaging materials such as AlN and Si3 N4 have been developed for application at temperatures in excess of 573 K.3–5) The development of a high temperature solder that can function at an ambient temperature above 573 K is expected to enable significant improvements in power electronics. For a solder alloy design, it is necessary to consider the proper melting range. The soldered interconnections must remain intact without melting for service temperatures of 573623 K. Thus, the solidus temperature of the high temperature solder needs to be 2025 K higher than the maximum operating temperature of 623 K, and the melting range is recommended to be narrower than 25 K to avoid the formation of solidification defects. To meet these requirements, the solidus temperature for the high temperature solder is recommended to be above 643 K and the liquidus temperature below 673 K. Only a few candidates exist for high temperature lead-free solder application, including Au-(Sn, Si, Ge), Bi-Ag, and Zn based alloys. Since Au and Bi based alloys have several serious problems such as their high cost, formation of a massive intermetallic compound and brittle nature,6–9) Zn based alloys are thought to be a good choice. For this reason, several studies on Zn-(Sn, Al, Mg, Ga) based high temperature solders have been reported.10–14) Vianco defined an ultra high temperature solder working properly between 573 K and 623 K and suggested a Zn-Al based alloy for the possible *Graduate
Student, Osaka University
alloy system.15) It has been reported the tensile strength, creep resistance, dimensional stability and corrosion resistance of Zn alloy are improved by the addition of Al/Cu elements.16–18) However, little research has been reported on Zn based solder alloyed with Al and Cu with a melting range of 643673 K for high temperature use. In this paper, we examined the proper chemical compositions of Zn-Al-Cu alloy to meet the requirements of the melting range, and investigated fundamental characteristics including microstructure, mechanical properties and solderability for high temperature solder application. 2.
Experimental Procedures
Considering an operating temperature of 573623 K, the proper melting range of high temperature solder was set as a solidus temperature above 643 K and a liquidus temperature below 673 K. From Zn-Al-Cu ternary phase diagrams, it was estimated the liquidus temperature of Zn-Al-Cu alloy in the composition range of Zn-(03 mass%)Al-(16 mass%)Cu exceeds the limit of 673 K that we set19) (hereafter, the composition unit of mass% is omitted in the notation). Therefore, in this study, we studied the chemical composition range of Zn-(46)Al-(16)Cu. The alloys used in this study were prepared from pure Zn (99.96), Al (99.99) and Cu (99.99) using the following procedures. Initially, Zn and Cu pure elements were melted in graphite crucibles using a high frequency induction furnace to produce a Zn-10Cu alloy. The melt was kept at 823 K for 300 s in an inert Ar gas atmosphere to ensure complete dissolution, and then cast into Cu molds coated with boron nitride. The Zn-10Cu, pure Al and pure Zn were prepared to meet the target composition, and melted together in the same way using the induction furnace in an Ar gas atmosphere. Molten metal was kept at 773 K for 300 s and poured into the Cu molds with an internal diameter of 12 mm and a height of 100 mm. Table 1 shows the chemical compositions of Zn-Al-Cu alloys for the
1532
S.-J. Kim, K.-S. Kim, S.-S. Kim, C.-Y. Kang and K. Suganuma Chemical compositions of casted Zn-Al-Cu solder alloys.
Alloy (mass%)
Zn
Al
Cu
Pb
Cd
Zn-4Al-3Cu
Bal.
4.208
3.142
0.011
—
Zn-5Al-3Cu
Bal.
5.116
3.098
0.008
0.001
Zn-6Al-3Cu
Bal.
6.037
3.241
0.010
—
selected samples, which were analyzed using inductively coupled plasma (ICP) mass spectroscopy. The melting range was investigated using a differential scanning calorimeter (DSC) for a 20 mg specimen and a cooling rate of 1 K/min in a purified Ar gas atmosphere. The cross-section of the cylindrical solder alloy was polished and etched using 5% HCl and 95% methanol solution. Microstructural examination was carried out for the phase and composition analysis using optical microscopy (OM), scanning electron microscopy (SEM), X-ray diffraction (XRD), electron probe micro-analysis (EPMA) and transmission electron microscopy (TEM). Samples for TEM were prepared by ion beam thinning. Solderability of the Zn-Al-Cu solders was evaluated with a spread ratio through a spreading test. The spreading tests were performed in an Ar gas atmosphere. For the substrate, a 99.99Cu sheets 1 mm thick with areas of 15 15 mm were used. After polishing, the Cu sheets were degreased in a 10% water solution of HCl, and then cleaned in ethanol and dried in air. A piece of the solder (5 mm t1 mm; 100 mg) was laid on the Cu sheets, heated to 703 or 723 K in the solder bath, and held for 30 s. The area of the solder spread was measured three times per alloy, and the average value was selected for multiple regression analysis. Vickers hardness was measured for a 9.8 N load and 10 s holding. A specimen for the tensile test was prepared by directly casting the cylindrical shape following the ASTM E8m standard. A tensile test was executed with a crosshead speed of 2 mm/min (gauge length: 25.4 mm, strain rate: 1:31 103 s1 ). 3.
700 mass% Al Liquidus Solidus
Chemical composition (mass%)
690
Temperature, T / K
Table 1
4%
5%
6%
680
670
660
650
640 0
3 2 5 4 Cu content, x / mass%
1
6
7
Fig. 1 Variations of liquidus and solidus temperatures as functions of the Al and Cu contents. Table 2 Melting range of the selected Zn-Al-Cu alloys as the candidates for the high temperature lead free solder. Chemical composition (mass%)
Solidus
Liquidus
Melting
temperature
temperature
Range
(K)
(K)
(K)
Zn
Al
Cu
95
4
1
644.4
661.1
16.5
94
4
2
642.5
664.4
21.7
93
4
3
643.2
670.3
26.9
93
5
2
644.4
657.8
13.2
92
5
3
643.5
664.0
20.3
91
5
4
643.2
671.0
27.6
91 90
6 6
3 4
642.8 641.6
653.4 660.7
10.4 18.9
89
6
5
643.3
669.4
25.9
Results and Discussion
3.1 Melting range Figure 1 shows variations of liquidus and solidus temperatures as functions of the Al and Cu contents. The solidus temperature for all samples was maintained around 643 K. This solidus temperature is quite similar to the ternary eutectic temperature of a Zn-Al-Cu system as shown in Fig. 2 (Zn-7.05Al-3.85Cu, 645 K).19) The liquidus temperature increased with the Cu content, however it decreased with the Al content. The melting range of Zn-Al-Cu alloys increased to 50 K and decreased to 10 K as the Al content decreased and the Cu content increased. To meet the above melting range requirements, nine alloy compositions of Zn-Al-Cu solder were selected as shown in Table 2, and further characteristics of were examined only for these nine candidates. The nine candidate alloys are marked in Fig. 2, showing the results of the melting range measurement are in agreement with the phase diagram.19)
3.2 Microstructure Figure 3 shows the microstructure of Zn-Al-Cu solders as a function of Al and Cu contents. Regardless of the Al/Cu content, the microstructure consisted of three phases, i.e., white dendritic phase (A), gray dendritic phase (B) and black eutectic phase (C), as marked in Fig. 3(a). For determination of the phases that appeared in the alloys, XRD (Fig. 4) and EPMA (Fig. 5) were carried out. XRD patterns indicated "-phase (CuZn4 ), -phase consisting of Zn-(0.81.5)Al(1.21.5)Cu and -phase consisting of Al-(1732)Zn existed in the Zn-Al-Cu alloys. Quantitative EPMA results showed the white dendritic phase (A) was Zn-18.4Cu and the gray dendritic phase (B) was Zn-1.3Al-1.5Cu. Combining the EPMA and XRD results, the white phase (A) was primary "-phase (CuZn4 ) and the gray phase (B) was -phase (Zn matrix). Eutectic phase (C) observed in OM (Fig. 3) had two different eutectic structures, i.e., lamellar (C1) and nonlamellar (C2) structures, in the SEM observation as shown in
Characteristics of Zn-Al-Cu Alloys for High Temperature Solder Application
(b)
Cu
773
Al
3K
3K
3K
10
t%
67
/a
(Al2Cu)
71
69
ε (CuZn4)
η (Zn)
Zn
95
85 80 90 Zn content, x / at%
L+α' 673 623
L+ε+η
η +α'
ε+ η +α'
573
5
α'(Al)
L+ε L+ η
20
t, x ten on
20
Temperature, T / K
10
con ten t, x
/a
L
723 c Cu
t%
wt% 4% 5% 6% Al
Al
Zn-5Al-4Cu (mass%)
ε+ η
(a)
1533
ε+ η +α η +α 523 15 10 5 20 Cu 0 Al content, x / at%
Zn Cu 9.2
Fig. 2 Ternary phase diagram of Zn-Al-Cu system: (a) Zn rich region (Zn-(030 at%)Al-(030 at%)Cu) and (b) 91 mass%Zn vertical section (b, marked as a dotted line in (a)).19)
(c)
(b)
(a)
C A A B
B 20µ m
C Fig. 3
Optical microstructures of Zn-x%Al-3Cu solder alloys: (a) 4Al, (b) 5Al, and (c) 6Al.
Intensity, I / cps
η (Zn, HCP) α (Al, FCC) ε (CuZn4, HCP)
mass%
Zn-6Al-3Cu Zn-5Al-4Cu Zn-5Al-3Cu
B A
5µ m
A
B
Zn
82.6
97.2
Al
-
1.3
Cu
18.4
1.5
Total
100.0
100.0
Zn-5Al-2Cu Zn-4Al-3Cu
30
40
50 60 70 80 Diffraction angle, 2θ / degree
Fig. 5 EPMA quantitative analysis of Zn-5Al-3Cu solder alloy.
90
Fig. 4 X-ray diffraction patterns of Zn-Al-Cu alloys (Cu K).
Fig. 6. Figure 7(a) shows the TEM micrograph of the fine lamellar structure (C1), and Figs. 7(d) and (f) show the electron diffraction patterns of a dark phase and bright phase, respectively. The electron diffraction pattern showed the lamellar structure was composed of dark -phase (Al-rich, FCC) and bright -phase (Zn-rich, HCP), which were expected to be Zn-Al eutectic phase from the ternary phase diagram.19) Figure 7(b) is the TEM micrograph of the nonlamellar structure (C2), and Fig. 7(e) shows the electron
diffraction pattern of the dark phase in the micrograph. The electron diffraction pattern of the bright phase in this coarse eutectic was the same as the pattern of -phase (Fig. 7(f)). From the pattern analysis, the non-lamellar structure was composed of "-phase (CuZn4 , HCP) and -phase (Zn-rich, HCP). This "- eutectic phase was found to be Zn-Cu eutectic phase from the Zn-Al-Cu ternary phase diagrams.19) -phase in Fig. 7(b) (marked as a dotted square) was observed at a higher magnification as shown in Fig. 7(c). From the results of the microstructure observation and analysis, regardless of the Al and Cu contents, the microstructure of Zn-(46)Al-(15)Cu alloy consisted of primary-" phase (CuZn4 ), -phase (Zn matrix), - eutectic phase (fine lamellar eutectic) and "- eutectic phase (coarse
1534
S.-J. Kim, K.-S. Kim, S.-S. Kim, C.-Y. Kang and K. Suganuma
C2
(a)
(b)
(c)
C1
C1
C2
C1 5µm
C2 Fig. 6
SEM micrographs of eutectic phases in Zn-xAl-3Cu alloys: (a) 4Al, (b) 5Al, and (c) 6Al.
(c)
(b)
(a)
(c)
η -phase η -phase α-phase ε-phase 500nm
(d)
500nm
(e)
- - 242 022 220
100nm
(f)
_ 1011 _ 0001 1010 0000
_ _ 0111 1101 _ 0000 1010
Fig. 7 TEM micrographs and electron diffraction patterns of eutectic phases in Zn-6Al-3Cu alloys: TEM micrographs of - phase (a), "- phase (b), higher magnification of " phase (c), Electron diffraction pattern of -phase(Al, Z¼½101) (d), "-phase(CuZn4, Z¼½12 10) (e), and -phase(Zn, Z¼½1213) (f).
80
Area fraction, Af / %
eutectic). The fraction of each phase, however, changed with Al and Cu contents as shown in Fig. 8. From the Zn-Al-Cu ternary phase diagram as shown in Fig. 2, an increase in Al content means the composition of an alloy approaches the ternary eutectic in the composition range of 46Al. In the view of microstructure, decreases in the "-phase and -phase fractions, and an increase in the eutectic phase (- and "- phase) fraction are expected. Figure 8(a) shows an increase in Al content results in an increase of in the eutectic phase fraction and a decrease in the phase fraction, but little change is observed in the area fraction of "-phase. With a decrease in the Cu content, the composition of Zn-Al-Cu alloy approaches the eutectic line of " and phases. Thus, the decreases in "-phase and eutectic phase fractions and an increase in the -phase fraction are expected. Figure 8(b) shows that "-phase and eutectic fractions phase increase and the -phase fraction decreases with Cu content. The area
(a)
60
(b) Matrix η-phase Matrix η-phase
40 Eutectic phases
Eutectic phases
Primary ε−phase
Primary ε−phase
20
0 2
3 4 5 6 7 80 1 2 3 4 5 6 Cu content, x / mass% Al content, x / mass%
Fig. 8 Area fraction of cross section of Zn-Al-Cu alloys: (a) Zn-xAl-3Cu alloys (x ¼ 37Al) and (b) Zn-6Al-xCu alloys (x ¼ 15 Cu).
Characteristics of Zn-Al-Cu Alloys for High Temperature Solder Application
120 110 100
120 100 80 60 10
90
1
11
300
(b)
10
290 280 Au-20Sn∗
270
Elongation / %
Tensile strength, σ / MPa
5% 6%
(a)
9
1
2
3
4
5 5 6 0 1 2 Cu contents, x / mass%
280 270 260 250 10
15
20
25
30
35
(b)
66
64
62
60
7
Bi-2.5Ag:34MPa∗
0
35
(a)
8
6
260
30
290
68
Spread ratio, Sr / %
mass% Al 4%
25
(b) y = 0.31Af + 207.5
300
Fig. 11 Effect of the area fraction of eutectic phases on Vickers hardness (a) and tensile strength (b).
6
2 3 4 5 Cu contents, x / mass%
20
310
Area fraction of eutectic phases, Af / %
Fig. 9 Effect of Al and Cu content on the Vickers hardness for Zn-(45)Al-(15)Cu alloys.
250
15
mass% Al 4% 5% 6%
80 0
310
(a) y = 1.33Af + 80.94
Tensile strength, σ / MPa
140 Vickers hardness, H / HV
Vickers hardness, H / HV
130
1535
mass% Al 4% 5% 6%
0 1 2
mass% Al 4% 5% 6%
3 4 5 6 70 1 2 3 4 5 6 7 Cu content, x / mass%
Fig. 12 Spread ratio of Zn-Al-Cu alloys at 703 K (a) and at 723 K (b).
3
4
5
6
Fig. 10 Effect of Al/Cu content on the tensile property for Zn-Al-Cu alloys: (a) tensile strength and (b) elongation .20,21)
fraction of each phase changed with Al and Cu contents, as shown in Fig. 8, in agreement with the Zn-Al-Cu ternary phase diagram. 3.3 Mechanical properties Figure 9 shows the variation in Vickers hardness as a function of alloying content. The hardness of Zn-Al-Cu solders increased with the Al and Cu contents, being about 94123 HV. The hardness of Zn-Al-Cu alloy was lower than that of Au-20Sn (200.8 HV); however, it was sufficiently higher than the prior reported hardnesses of Bi and Pb based solders (Bi-2.5Ag: 15 HV and Pb-5Sn: 8.9 HV).20) Figure 10 shows the variation in tensile properties as a function of alloying content. The tensile strength increased and the elongation decreased as the Al and Cu contents increased. Zn-Al-Cu alloys showed sufficiently high tensile strengths of 255300 MPa, which were higher than Pb-5Sn (15 MPa), Bi-2.5Ag (34 MPa) and Au-20Sn (275 MPa) tensile strengths. The elongation was lower than that for Pb-5wtSn (39%); however, it was higher than that for the other high temperature lead free solder candidates (Bi-2.5Ag: 1.5% and Au-20Sn: 4.2%).20)
The variations in the tensile property correlated to the microstructural behavior as a function of the Al and Cu contents. As the Al and Cu contents increased, the phase decreased and the eutectic phases (- and "- eutectic phases) increased (Fig. 8). Figure 11 indicates relationships of the Vickers hardness and tensile strength with the area fraction of the eutectic phases. The Vickers hardness and tensile strength have linear relationships with the area fraction of the eutectic phases. Thus, the mechanical properties of the Zn-Al-Cu solders strongly depend on the area fraction of fine eutectic phases. 3.4 Spreadability Solderability of the Zn-Al-Cu solder alloy was tested by measuring the spread ratio. Spread tests were conducted at 703 K and 723 K, which were 32 K and 52 K higher than the maximum liquidus temperature among the nine candidates. The spread ratio was calculated from the area difference of the 5 mm diameter solder disk prior to and post heating: Spread ratio (%) ¼ ðAt Ai Þ 100=Ai ;
ð1Þ
where Ai is the initial plan area of a solder alloy (2:5 2:5 mm2 ) and At is the total plan area wetted by the molten solder alloy. Figure 12 shows the influence of alloying content on the spread ratio for the Zn-Al-Cu solders. As the Al content
1536
S.-J. Kim, K.-S. Kim, S.-S. Kim, C.-Y. Kang and K. Suganuma
increased, the spread ratio improved. However, the Cu content had an insignificant effect on spreadability. A higher test temperature produced a larger spread ratio. The addition of Al to the Zn alloy and high temperature has been known to increase the fluidity of the alloy;18) therefore the spread ratio increased with Al content and test temperature. The Zn-AlCu solders had spread ratios between 62 and 66%, which were lower than those of Pb-5Sn based solders (8590%). However, from the results of prior research, it should be improved using process techniques such as using a metal coating, mechanical assistance of the movement of the molten solder, and the use of flux.15) 4.
Conclusion
In the present work, Zn based solders were examined using alloying elements, i.e., 46Al and 16Cu for high temperature solder applications. The solder candidates having the appropriate melting range of 645673 K were Zn-4Al-(13)Cu, Zn-5Al(24)Cu and Zn-6Al-(35)Cu. Regardless of the Al and Cu contents, the microstructure of the Zn-Al-Cu solders consisted of the primary "-phase (CuZn4 ), matrix -phase (Zn matrix), - eutectic phase (Zn-Al eutectic) and "- eutectic phase (Zn-Cu eutectic). The fraction of each phase varied with the alloying content. As the Al and Cu contents increased, the eutectic phases increased. The increase in eutectic phases was associated with the increases in Vickers hardness and tensile strength, and with the decrease in elongation. The increase in the Al content improved the solderability expressed as a spread ratio. Through the evaluation of fundamental characteristics of Zn-Al-Cu solders, it is concluded that Zn-Al-Cu alloys have potential as high temperature lead-free solders in view of their melting range. Comparing to the other high temperature lead-free solder candidates (Bi-2.5Ag and Au-20Sn), Zn-AlCu solders have superior mechanical properties. Considering the solderability of solder can be improved by process technologies including inert gas and/or joining pressure, the elongation of the solder is thought to be a very important property of high temperature solders for the reliability of the soldered joint. Zn-4Al-(23)Cu solders have the highest elongation value of about 9% and a proper melting range close to 25 K. Therefore, we suggest Zn-4Al(23)Cu as a suitable composition range for high temper-
ature solder application. For the evaluation of a composition’s potential as a solder alloy, further works on joining characteristics are required. Acknowledgements This work carried out under the support of NEDO project, ‘‘R&D of Alternatives to High temperature High Lead Solder’’. REFERENCES 1) P. L. Dreike, D. M. Fleetwood, D. B. King, D. C. Sprauer and T. E. Zipperian: IEEE Trans. Components, Packaging, and Manufacturing Technology, Part A 17 (1994) 594–609. 2) K. C. Reinhardt and M. A. Marciniak: Proc. 3rd Int. High-Temperature Electronics Conf., (Albuquerque, NM, 1996) pp. 9–15. 3) P. G. Neudeck, R. S. Okojie and L. Y. Chen: Proc. of IEEE 90 (2002) 1065–1076. 4) H. Ueda: Proc. The 17th Inter. Symp. Power devices and IC’s, (IEEE, 2005) p. 147. 5) M. Bratcher, R. J. Yoon and B. Whitworth: Proc. 3rd Int. HighTemperature Electronics Conf., (Albuquerque, NM, 1996) pp. 21–26. 6) K. Suganuma: Solid State and Mater. Sci. 5 (2001) 55–64. 7) C. Y. Liu and K. N. Tu: J. Mater. Res. 13 (1988) 37–44. 8) J. H. Kim, S. W. Jeong and H. M. Lee: J. Electron. Mater. 31 (2002) 557–563. 9) K. S. Kim, S. H. Hur and K. Suganuma: Microelectron. Reliab. 43 (2003) 259–267. 10) J. E. Lee, K. S. Kim, K. Suganuma, J. Takenaka and K. Hagio: Mater. Trans. 46 (2005) 2413–2418. 11) J. E. Lee, K. S. Kim, K. Suganuma, M. Inoue and G. Izuta: Mater. Trans. 48 (2007) 584–593. 12) M. Rettenmayer, P. Lambracht, B. Kempf and C. Tschudin: J. Electron. Mater. 31 (2002) 279–285. 13) T. Shimizu, H. Ishikawa, I. Ohnuma and K. Ishida: J. Electron. Mater. 28 (1999) 1172–1174. 14) M. Nobumasa and N. Shuichi: Japan patent 2005–143983. 15) P. T. Vianco: Welding Journal 81 (1999) 51–55. 16) Y. H. Zhu: J. Mater. Sci. 36 (2001) 3973. 17) Y. H. Zhu and F. E. Goodwin: J. Mater. Res. 10 (1995) 1927. 18) H. Schumann: Metallographie (Wiley-VCH Verlag, 1991) p. 681. 19) G. Petzow and G. Effenberg: Ternary Alloys, A Comprehensive Compendium of Evaluated Constitutional Data and Phase Diagrams, vol. 4 (Weinheim, Federal Republic of Germany, New York, 1988) 92–11. 20) J. E. Lee: Developments of Zn-Sn based high and low temperature lead free solders and their reliability analysis, (Ph. D. Dissertation of Osaka Univ., Osaka, 2007). 21) J. H. Kim, S. W. Jeong and H. M. Lee: Mater. Trans. 43 (2002) 1873–1878.
