This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Transcription

1 This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier s archiving and manuscript policies are encouraged to visit:

2 Microelectronics Reliability 49 (2009) Contents lists available at ScienceDirect Microelectronics Reliability journal homepage: Kirkendall voids formation in the reaction between Ni-doped SnAg lead-free solders and different Cu substrates Y.W. Wang, Y.W. Lin, C.R. Kao * Department of Materials Science and Engineering, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei City, Taipei 10617, Taiwan article info abstract Article history: Received 25 July 2008 Received in revised form 24 September 2008 Available online 4 November 2008 The reactions between Sn2.5Ag solder doped with different levels of Ni (0 0.1 wt.%) and two different types of Cu substrates, electroplated Cu and high-purity Cu substrates, were studied. The main objective was to investigate the effect of Cu substrate and the effect of Ni additions on the formation of Kirkendall voids within the Cu 3 Sn phase. Reaction conditions included one reflow and subsequent aging at 160 C for up to 2000 h. After reflow, Cu 6 Sn 5 was the only reaction product observed for all the different solders and substrates used. During aging, both Cu 6 Sn 5 and Cu 3 Sn formed. Nevertheless, Kirkendall voids were observed only when the electroplated Cu was used, and was not observed when high-purity Cu was used. It was proposed that impurities in electroplated Cu helped the nucleation of these voids. The Ni additions made the Cu 3 Sn layer thinner. For the case of the electroplated Cu substrates, the amount of Kirkendall voids decreased correspondingly with the Ni additions. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction At temperatures greater than C, the solid-state reactions between Cu and many Sn-based solders produce two reaction products, Cu 6 Sn 5 and Cu 3 Sn [1]. The growth of Cu 3 Sn tends to induce the formation of Kirkendall voids, while the growth of Cu 6 Sn 5 does not [1 4]. Zeng et al. [4] reported that a large number of Kirkendall voids formed when electroplated Cu was reacted with the eutectic PbSn solder at C. These voids located not only at the Cu/Cu 3 Sn interface but also within the Cu 3 Sn layer. The formation of Kirkendall voids is not limited to the eutectic PbSn solder. Kirkendall voids were also reported to form when electroplated Cu was reacted with SnAgCu solders [5], or pure Sn [1]. The relatively rapid diffusion of Cu out of the Cu 3 Sn layer could be the major contributing factor for the formation of these voids. It was indeed shown that although both Cu and Sn were mobile within Cu 3 Sn, the Cu flux was somewhat greater than that of the Sn flux (three times greater at 200 C) [6 8]. The formation of Kirkendall voids accompanying the Cu 3 Sn growth raises serious reliability concerns because excessive void formation increases the potential for brittle interfacial fracture [2,4,5,9,10]. Since the formation of Kirkendall voids is associated with the growth of Cu 3 Sn, it is reasonable to assume that the amount of Kirkendall voids can be reduced by reducing the Cu 3 Sn thickness. Recently, It was shown that Ni addition to Sn3.5Ag (3.5 wt.% Ag, balance Sn) in amounts as minute as 0.1 wt.% could * Corresponding author. Tel./fax: addresses: crkao@ntu.edu.tw, kaocr@hotmail.com (C.R. Kao). substantially retarding the Cu 3 Sn growth [11 13], raising the hope that Ni addition can alleviate the problems caused by Kirkendall voids. In addition to the effect of Ni addition, it was further pointed out that the type of Cu substrate used also had a marked effect on the Kirkendall voids formation [1]. It was reported that Kirkendall voids formed when electroplated Cu was used, while no such void was detected when OFHC (oxygen-free high conductivity) Cu substrate was used [1]. The objective of this study was to investigate the effect of Cu substrate and the effect of Ni additions on the formation of Kirkendall voids within the Cu 3 Sn phase. Specifically, the reactions between Sn2.5Ag solder doped with different levels of Ni (0 0.1 wt.%) and two different types of Cu substrate, the electroplated Cu and the OFHC Cu substrates, were studied. 2. Experimental Solder balls with four different compositions, Sn2.5Ag, Sn2.5Ag0.01Ni, Sn2.5Ag0.03Ni, and Sn2.5Ag0.1Ni were prepared from % purity elements. The solder ball compositions were verified by using ICP-AES (inductively coupled plasma atomic emission spectroscopy), and it was estimated that the reported compositions had a maximum wt.% uncertainty. For a given solder ball composition, there was no measurable ball-to-ball variation in composition. Each solder ball had a mass of 10 mg. The solder balls were placed on Cu substrates with 600 lm diameter wettable openings. The solder could not wet the Cu substrate outside this opening, and the solder as a result formed a /$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi: /j.microrel

3 Y.W. Wang et al. / Microelectronics Reliability 49 (2009) sphere only over this wettable opening. Two types of Cu substrates were used, electroplated Cu and OFHC Cu. The Cu plating solution was CuSO 4 -based and did not contain the C 6 H 12 O 6 S 4 Na 2 additive. The reflow temperature profile had a peak temperature of 235 C and 90 s duration during which the solder was molten. The nominal ramp rate and cooling rate were both 1.5 C/s. After reflow, the samples were subjected to solid-state aging at 160 C for 500, 1000 or 2000 h. The samples were then mounted in epoxy, sectioned by using a low-speed diamond saw, and metallurgically polished in preparation for characterization. The reaction zone for each sample was examined using an optical microscope and a scanning electron microscope (SEM). The compositions of the reaction products were determined using an electron microprobe, operated at 15 kev. In microprobe analysis, the concentration of each element was measured independently, and the total weight percentage of all elements was within 100 ± 1% in each case. For every data point, at least four measurements were made and the average value was reported. 3. Results Fig. 1a1 a4 shows the backscattered electron micrographs for the samples using the electroplated Cu substrate with different Ni additions that were reflowed for one time. The corresponding micrographs for the OFHC substrates are shown in Fig. 2a1 a4. For both types of Cu substrates, the Cu 6 Sn 5 phase is clearly visible. At this stage, the Cu 6 Sn 5 phase was the only phase observed at the interface, and Cu 3 Sn was not observed. Comparing Fig. 1a1 a4 to their counterparts in Fig. 2a1 a4, one concludes that these two different substrates produced essentially the same microstructure. In other words, the substrate effect was not important in the as-reflow condition. Figs. 1a1 a4 and 2a1 a4 also shows that the additions of Ni to Sn2.5Ag made the amount of Cu 6 Sn 5 at the interface increase substantially. For Sn2.5Ag without any Ni addition, the Cu 6 Sn 5 layer was thin, continuous, and void-free, consistent with the classical scallop microstructure reported in the literature [14 16]. With the addition of 0.01 wt.% Ni, Cu 6 Sn 5 still retained the scallop microstructure. Nevertheless, with the addition of 0.03 wt.% Ni, Cu 6 Sn 5 became thicker, and had many voids between these Cu 6 Sn 5 grains. As had been pointed out in our previous study [11,12], these voids were originally occupied by trapped solder. During sample preparation, the trapped solder in these voids was etched away. When the Ni addition increased even more, the amount of Cu 6 Sn 5 near the interface increased correspondingly. The above observation showed that Cu 6 Sn 5 formed at the interface without regard to the Ni additions. The alloy additions, however, did change the amount of Cu 6 Sn 5 and its microstructures at the interface. Fig. 1b1 d4 shows the micrographs for the samples using the electroplated Cu substrate with different Ni additions that were aged for 500, 1000 and 2000 h. The corresponding micrographs for the OFHC substrates are shown in Fig. 2b1 d4. In all cases, two intermetallic compounds, the Cu 6 Sn 5 -based phase and the Cu 3 Sn-based phase, formed. These two compounds both exhibited a dense layered structure. The voids within the Cu 6 Sn 5 phase formed during the reflow had disappeared after 500 h of aging. Figs. 1 and 2 show that, for all combinations of Ni additions and Cu substrates, both Cu 6 Sn 5 and Cu 3 Sn thickened with aging. During aging, the most striking feature is that the additions of Ni substantially reduced the Cu 3 Sn to Cu 6 Sn 5 thickness ratio. For samples with the same aging, the Ni additions had different effects on the growth of Cu 6 Sn 5 and Cu 3 Sn. Ni additions to Sn2.5Ag increased the Cu 6 Sn 5 thickness substantially, as shown in Fig. 3. The Ni additions, however, decreased the Cu 3 Sn thickness substantially, as shown in Fig. 4. In short, the types of the Cu substrate did not seem to have a clear effect on the growths kinetics of Cu 6 Sn 5 and Cu 3 Sn. Fig. 1. Backscattered electron micrographs for Sn2.5Ag alloys with different Ni additions that were reflowed once over the electroplated Cu substrates and then aged at 160 C for 0, 500, 1000, or 2000 h. The symbol SA denotes the Sn2.5Ag alloy, and SA0.01Ni denotes Sn2.5Ag0.01Ni, etc.

4 250 Y.W. Wang et al. / Microelectronics Reliability 49 (2009) Fig. 2. Backscattered electron micrographs for Sn2.5Ag alloys with different Ni additions that were reflowed once over the OFHC Cu substrates and then aged at 160 C for 0, 500, 1000, or 2000 h. The symbol SA denotes the Sn2.5Ag alloy, and SA0.01Ni denotes Sn2.5Ag0.01Ni, etc. In addition, both types of Cu substrates produced Cu 6 Sn 5 and Cu 3 Sn with very similar microstructures. Although different Cu substrates produced the same intermetallic compounds at the interface with very similar microstructure and growth kinetics. One key difference is the existence of the Kirkendall voids within the Cu 3 Sn phase when the electroplated Cu was used, as shown in Fig. 1. Zoom-in micrographs for Fig. 1d1 d4 showing the details of Kirkendall voids are presented in Fig. 5. For all the solder compositions used, the Kirkendall voids started to appear at an aging time between h. In contrast, there was no such void observed when the OFHC Cu was used even after aging for 2000 h, as shown in Fig. 2. With the electroplated Cu substrate, the amount of the Kirkendall voids increased with the aging time. For Sn2.5Ag without any Ni addition, 2000 h of aging at 160 C produced a nearly continuous series of Kirkendall voids near the Cu/Cu 3 Sn interface, shown in Fig. 5a. Such a pronounced degree of void formation without question posted a serious reliability concern for solder joints. With Ni additions, the Cu 3 Sn layer became thinner, and the amount of Kirkendall voids decreased correspondingly, as shown in Fig. 5. In other words, Ni addition did have its effectiveness in reducing the amount of the Kirkendall voids. 4. Discussion This study showed that the reaction products were always Cu 6 Sn 5 during reflow, and Cu 6 Sn 5 +Cu 3 Sn during aging at 160 C. Fig. 3. Total thickness of Cu 6 Sn 5 formed near the interface versus the aging time at 160 C. Fig. 4. Thickness of Cu 3 Sn formed near the interface versus the aging time at 160 C.

5 Y.W. Wang et al. / Microelectronics Reliability 49 (2009) Fig. 5. Zoom-in micrographs for Fig. 1d1 d4 showing enlarged views of Kirkendall voids. Although the compound formation did not change with the Ni addition, the amount of compound at the interface did increase substantially with Ni additions, as shown in Figs. 1 and 2. Our recent preliminary data show that for a given Ni addition, different cooling rates produced different amounts Cu 6 Sn 5 near the interface, with slower cooling rate producing thicker Cu 6 Sn 5. This observation seems to suggest that a major part of the Cu 6 Sn 5 actually formed during the solidification of the solder. A faster cooling rate corresponded to less time for Cu diffusion, and as a result the dissolved Cu atoms precipitated out as Cu 6 Sn 5 evenly throughout the entire joint. Consequently, less Cu 6 Sn 5 located near the interface. Conversely, a slower cooling rate corresponded to more time for Cu diffusion, and as a result the dissolved Cu atoms could diffuse back to the interface to precipitate out as Cu 6 Sn 5, and more Cu 6 Sn 5 located near the interface. Similar cooling rate effect had also been reported in a very recent study on the reaction between Sn and CuNi alloy substrates with various compositions [17]. An interesting observation was that additions Ni was able to reduce the Cu 3 Sn thickness substantially. It was effective even after aging at 160 C for 2000 h, which surpassed the typical industrial high temperature storage requirement of 150 C for 1000 h. The reason why Ni addition was effective in reducing the Cu 3 Sn thickness is unclear at this moment. Several theories have been proposed, including thermodynamic arguments [18] and kinetic arguments [12,19]. More studies are needed to elucidate this point. The reduced Cu 3 Sn thickness with the Ni addition was helpful in reducing the amount of voids within Cu 3 Sn, because the formation of these voids was directly related to the growth of the Cu 3 Sn layer. This study shows that the types of Cu substrate used have a prominent effect on the observation of Kirkendall voids within the Cu 3 Sn layer. The reason why the Cu substrate had such a strong effect deserves further discussion. Other than the Kirkendall voids, the two different substrates, electroplated and OFHC, behaved very similarly, in term of the intermetallic growth kinetics and microstructure. Since the growth kinetics and microstructure were controlled to a large degree by the interdiffusion of the elements, it seems reasonable to assume that as far as diffusion-related phenomena were concerned the two types of substrate behaved the same. In short, the Cu substrates simply served as the end-members of the diffusion couples, and maintained a constant Cu thermodynamic activity at the Cu/Cu 3 Sn interface. If the diffusion-related phenomena were the same for both types of substrate, then what was the reason why the Kirkendall voids were observed for one type of substrate, but not for the other? It was very likely that the impurities in electroplated Cu helped the nucleation of the voids. Assuming all the atoms in Cu 6 Sn 5 and Cu 3 Sn diffused by vacancy mechanism, one notes that there were two necessary conditions for the formation of the Kirkendall voids. The first was that the total out-going atomic flux was greater than the total in-coming atomic flux for a certain controlled volume. There was consequently a net in-coming vacancy flux into the controlled volume due to the un-balanced atomic flux. These in-coming vacancies were at the same time destroyed at vacancy sinks that existed in any materials, such as dislocations. The vacancy concentration was determined by the in-coming vacancy flux minus those vacancies destroyed by various vacancy sinks. The vacancy concentration had to be higher than a certain critical value for the nucleation of the voids to occur. A vacancy concentration higher than this critical value was the second necessary condition for the formation of the voids. In electroplated Cu, impurities and crystal imperfections are imbedded during the electroplating process. These impurities and imperfections might play key roles in assisting the void nucleation. These impurities and imperfections might lower the critical vacancy concentration value by serving as the heterogeneous nucleation sites for void nucleation. This was probably the reason why Kirkendall voids were only observed for electroplated Cu even though the diffusional characteristics, expressed in term of growth kinetics and phase microstructure, were quite similar for both substrates. Very recently, it was pointed out that the electroplating additive indeed had a very strong effect on the void formation and its distribution within the Cu 3 Sn layer. It was indicated that the electroplating additive assisted the formation of the voids by accelerating the void nucleation [20]. The identity of Cu 6 Sn 5 is in fact a rather complicated issue. At temperatures lower than 186 C, Cu 6 Sn 5 exhibits a monoclinic g structure [21,22]. Recent studies, nevertheless, showed that Ni was able to stabilize the high temperature hexagonal g structure. Since Ni was observed in Cu 6 Sn 5 in this study, these Cu 6 Sn 5 might in fact have the high temperature hexagonal g structure. A recent phase diagram study, that has detailed information on this issue, has been published very recently [23]. 5. Summary (1) Kirkendall voids formed only when electroplated Cu was used. No such void was observed when OFHC Cu was used. (2) Apart from the Kirkendall voids, the two different Cu substrates behaved very similarly, in term of intermetallic thickness and microstructure.

6 252 Y.W. Wang et al. / Microelectronics Reliability 49 (2009) (3) In the case of electroplate Cu substrate, Kirkendall voids formed after aging at 160 C for more than 500 h. The amount of voids increased with the aging time. (4) The additions Ni was helpful in retarding the Cu 3 Sn growth, thus reducing the amount of Kirkendall voids when the electroplated Cu was used. Acknowledgements This work was supported by the National Science Council of R.O.C. through Grant NSC E MY3. The EPMA analysis performed by Ms. S.Y. Tsai of National Tsing Hua University is also acknowledged. References [1] Laurila T, Vuorinen V, Kivilahti JK. Interfacial reactions between lead-free solders and common base materials. Mater Sci Eng 2005;R49:1 60. [2] Ahat S, Sheng M, Luo L. Microstructure and shear strength evolution of SnAg/ Cu surface mount solder joint during aging. J Electron Mater 2001;30(10): [3] Vianco PT, Rejent JA, Hlava PF. Solid-state intermetallic compound layer growth between copper and 95.5Sn 3.9Ag 0.6Cu solder. J Electron Mater 2004;33(9): [4] Zeng K, Stierman R, Chiu TC, Edwards D, Ano K, Tu KN. Kirkendall void formation in eutectic SnPb solder joints on bare Cu and its effect on joint reliability. J Appl Phys 2005;97: [5] Chiu TC, Zeng K, Stierman R, Edwards D, Ano K. Effect of thermal aging on board level drop reliability for Pb-free BGA packages. In: Proceedings of 2004 electronic components and technology conference, IEEE; p [6] Oh M. Doctoral thesis, Lehigh University; [7] Oberndorff P. Doctoral thesis, Technical University of Eindhoven; [8] Paul A. Doctoral thesis, Technical University of Eindhoven; [9] Mei Z, Ahmad M, Hu M, Ramakrishna G. Kirkendall voids at Cu/solder interface and their effects on solder joint reliability. In: Proceedings of 2005 electronic components and technology conference, IEEE; p [10] Date M, Shoji T, Fujiyoshi M, Sato K, Tu KN. Impact reliability of solder joints. In: Proceedings of 2004 electronic components and technology conference, IEEE; p [11] Tsai JY, Hu YC, Tsai CM, Kao CR. A study on the reaction between Cu and Sn3.5Ag solder doped with small amounts of Ni. J Electron Mater 2003;32(11): [12] Ho CE, Yang SC, Kao CR. Interfacial reaction issues for lead-free electronic solders. J Mater Sci Mater Electron 2007;18: [13] Wang YW, Lin YW, Tu CT, Kao CR. Effects of minor Fe, Co, and Ni additions on the reaction between SnAgCu solder and Cu. J Alloy Compd, accepted for publication. [14] Yang W, Felton LE, Messler Jr RW. The effect of soldering process variables on the microstructure and mechanical properties of eutectic Sn Ag/Cu solder joints. J Electron Mater 1995;24(10): [15] Choi S, Bieler TR, Lucas JP, Subramanian KN. Characterization of the growth of intermetallic interfacial layers of Sn Ag and Sn Pb eutectic solders and their composite solders on Cu substrate during isothermal long-term aging. J Electron Mater 1999;28(11): [16] Choi WK, Lee HM. Effect of soldering and aging time on interfacial microstructure and growth of intermetallic compounds between Sn 3.5Ag solder alloy and Cu substrate. J Electron Mater 2000;29(10): [17] Vuorinen V, Yu H, Laurila T, Kivilahti JK. Formation of intermetallic compounds between liquid Sn and Various CuNi x metallizations. J Electron Mater 2008;37(6): [18] Gao F, Takemoto T, Nishikawa H. Effects of Co and Ni addition on reactive diffusion between Sn 3.5Ag solder and Cu during soldering and annealing. Mater Sci Eng 2006;A420: [19] Yu H, Vuorinen V, Kivilahti JK. Effect of Ni on the formation of Cu 6 Sn 5 and Cu 3 Sn intermetallics. In: Proceedings of the 2006 electronic components and technology conference, IEEE; p [20] Kim JY, Yu J. Effects of residual impurities in electroplated Cu on the Kirkendall void formation during soldering. Appl Phys Lett 2008;92: [21] Vuorinen V, Laurila T, Mattila T, Heikinheimo E, Kivilahti JK. Solid-state reactions between Cu(Ni) alloys and Sn. J Electron Mater 2007;36(10): [22] Nogita K, Nishimura T. Nickel-stabilized hexagonal (Cu, Ni) 6 Sn 5 in Sn Cu Ni lead-free solder alloys. Scripta Mater 2008;59: [23] Schmetterer C, Flandorfer H, Luef CH, Kodentsov A, Ipser H. Cu Ni Sn: a key system for lead-free soldering. J Electron Mater doi: /s [available on-line].