Effects of Silver Coating Covered with Copper Filler on Electrical Resistivity of Electrically Conductive Adhesives

Transcription

1 Materials Transactions, Vol. 51, No. 1 (21) pp to 1789 Special Issue on Lead-Free and Advanced Interconnection Materials for Electronics #21 The Japan Institute of Metals Effects of Silver oating overed with opper Filler on Electrical Resistivity of Electrically onductive Adhesives Hiroshi Nishikawa 1, Saya Mikami 2; *, Koichi Miyake 3, Akira Aoki 4 and Tadashi Takemoto 1 1 Joining and Welding Research Institute, saka University, Ibaraki , Japan 2 Graduate School of Engineering, saka University, Suita , Japan 3 Mitsui Mining & Smelting o., Ltd., Ageo , Japan 4 Hikoshima Smelting o., Ltd., Shimonoseki 75-93, Japan Electrically conductive adhesives (EAs) are usually composed of conductive metal fillers and polymeric resin. For the metals fillers, silver is the most commonly used due to its high electrical and thermal conductivities, and chemical stability. Recently copper can be a promising candidate for conductive filler metal due to its low resistivity, low cost and good electro-migration performance. In this study, to overcome the problem of high electrical resistance associated with the oxidation of copper, copper particles were coated with silver, and the silver-coated copper was tested as a filler metal. In particular, the effect of silver coating on the electrical resistivity of EAs just after curing and after reliability tests was investigated. It was found that the electrical resistivity of EA using silver-coated copper filler was much lower and more stable than that of EA using pure copper filler after curing and after reliability tests. [doi:1.232/matertrans.mj212] (Received April 16, 21; Accepted July 22, 21; Published September 8, 21) Keywords: copper filler, electrically conductive adhesive, silver coating, electrical resistivity, oxidation 1. Introduction It is known that polymeric adhesives could be used to join a large variety of material combinations: metal-metal, metal-plastic, metal-composite, metal-ceramic plastic-plastic and so on. In electronics industry, electrically conductive adhesives (EAs) have been investigated for use in microelectronics packaging and interconnections as a lead-free solder substitute due to such advantages as low curing temperature and simple process. They are usually composed of conductive metal fillers and polymeric resin. The metal fillers provide mainly electrical and thermal properties of EA. The polymeric resins provide mechanical strength of EA and mechanical interconnections between surface mounted device components and a substrate. For the polymeric resins, thermosetting resins such as epoxy, polyimide and polyurethane are commonly proposed. For the metals fillers, gold (Au), silver (Ag), copper () and nickel (Ni) in various sizes and shapes are examined. Among them, silver is the most commonly used due to its high electrical and thermal conductivities, and chemical stability. However the cost of conductive adhesives filled with silver fillers is much higher than the conventional lead-free solders. Then, current EAs still have some limitations in terms of electrical and mechanical properties as replacements for the lead-free solder. Many research efforts have been focused on the improvement of electrical conductivity and the reliability of EA joints. 1 9) Additionally, the replacement of expensive silver fillers by other metal fillers or new materials has been examined for wider applications of EAs. 1 14) For example, S. K. Kang 1) investigated a low temperature conductive adhesive that was developed with bismuth-tin (Bi-Sn) coated copper powders dispersed in a polymer matrix. It was indicated that conductive adhesives containing Bi-Sn could *Graduate Student, saka University potentially replace solder in the future. H. P. Wu 11) developed a high-conductive EA with lower weight fraction filler using silver nanowires. W. Lin 14) reported that conductive adhesive filled with silver plating nano-graphite has broad and potential applications. Then, copper can be a promising candidate for conductive filler metal due to its low resistivity, low cost and good electro-migration performance. However, a problem for copper fillers is oxidation of the copper surface and deterioration of the EA electrical properties. There are basically two approaches for the surface treatment of copper fillers to prevent oxidation of the metal surface. ne is an inorganic material coating and the other is an organic coating. However, the thermal stability of these coatings is a concern because most lose their effectiveness when exposed to curing conditions and the various environmental conditions. In this study, EAs using surface-modified copper fillers were developed. In particular, to overcome the problem of high electrical resistance associated with the oxidation of copper, copper particles were coated with silver, and the silver-coated copper was tested as a filler metal. The effect of silver coating on the electrical resistivity of EAs just after curing and after reliability tests was investigated. 2. Materials and Experimental To investigate the effect of silver coating of copper particles on the electrical properties of EAs, four different types of copper particles were examined and four different types of conductive adhesives were prepared. The purity of copper for these copper particles was greater than 99.5 mass%, and the main contaminant of the copper particles was oxygen. The components are shown in Table 1. onductive adhesive EA-1 was mainly composed of pure copper fillers with a spinous shape and phenolic resin as a polymer matrix. EA-2 was composed of silver-coated copper fillers with a spinous shape and resin. This silvercoated copper contains 1 mass% silver. EA-3 was com-

2 1786 H. Nishikawa, S. Mikami, K. Miyake, A. Aoki and T. Takemoto Table 1 Basic composition of conductive adhesives used in this study. onductive adhesive EA-1 EA-2 EA-3 EA-4 Metal filler Pure Ag ratio (mass%) Shape of metal filler Spinous Spinous Spherical Spherical Metal content (mass%) 8.6 Polymer matrix Phenolic resin pad FR-4 5mA A V onductive adhesive 25 Unit : (mm) Fig. 1 Schematic illustration of four-point probe method for measurement of electrical resistance. 5 (c) (d) Fig. 2 SEM images of particles used in this study: pure copper fillers with a spinous shape, 1 mass% silver-coated copper filler with a spinous shape, (c) 5 mass% silver-coated copper filler with a spherical shape, (d) 1 mass% silver-coated copper filler with a spherical shape. posed of silver-coated copper fillers with a spherical shape and resin. This silver-coated copper contains 5 mass% silver. EA-4 was composed of silver-coated copper fillers with a spherical shape and resin. This silver-coated copper contains 1 mass% silver. In this study, silver coating was achieved on copper particle by using electroless plating. Then in the case of 1 mass% silver, the thickness of the silver coating is theoretically about 9 nm. In the case of 5 mass% silver, the thickness of the silver coating is theoretically about 42 nm. In all adhesives, the total content of metal fillers was fixed at 8.6 mass%. All of the conductive adhesives were cured at 448 K for 1 h in a convection coven in air. The four-point probe method, shown in Fig. 1, was applied to measure the electrical resistance of a specimen with specific dimensions, and the electrical resistivity of conductive adhesives was calculated from the resistance. Ten specimens were tested for each type of conductive adhesive. Then, some specimens were subjected to high-temperature exposure at 398 K up to 2 h and 1 h in air. For the other cured specimens, a temperature/humidity aging test at 358 K and 85% relative humidity (RH) was conducted for 15 h. The resistance of each specimen was measured periodically and the resistivity was calculated accordingly. The change rate of the electrical resistivity was determined from the following eq. (1), hange rate of electrical resistivity (%) ¼ð f i Þ= i 1 ð1þ where i is average electrical resistivity of EA before aging, and f is average electrical resistivity of EA after aging. Morphology of each copper particle was observed by using scanning electron microscopy (SEM). After hightemperature exposure test, auger electron spectroscopy was applied to clarify the composition of the copper particle surface in the conductive adhesive. 3. Results and Discussion The morphology of copper particle was observed through SEM. Figure 2 shows SEM images of the copper fillers used in this study. As can be seen from Fig. 2,, the mean particle size for copper fillers with a spinous shape is approximately 2.5 mm. From Fig. 2(c), (d), the mean particle size for copper fillers with a spherical shape is approximately 5 mm. In order to investigate the effect of silver coating on the electrical resistivity of EAs, Fig. 3 shows electrical resistivity of each EA after curing at 448 K for 1 h. Regardless of the particle shape, the electrical resistivity of conductive

3 Effects of Silver oating overed with opper Filler on Electrical Resistivity of Electrically onductive Adhesives 1787 Electrical resistivity, ρ / Ω. cm [ 1-5 ] 2 1 EA-1 EA-2 EA-3 EA-4 Fig. 3 Effect of silver coating for particles on electrical resistivity of electrical conductive adhesives just after curing at 448 K for 1 h. hange rate of electrical resistivity (%) : EA-1 : EA-2 ring temperature: 448 K : EA-3 ring time: 1 h : EA-4 Holding temperature: 398 K Holdin time, t / h Fig. 4 Effect of silver-coating on electrical resistivity of EAs during high-temperature exposure test : : : : Fig. 5 Atomic concentration of, and for pure copper particle in EA-1: just after curing at 448 K for 1 h, after high temperature exposure at 398 for 1 h. adhesives EA-2, EA-3, and EA-4, which were filled with the silver-coated copper particles, is much lower than that of EA-1, the pure copper particles. EA-2, EA-3, and EA-4 showed the electrical resistivity of the same order of magnitude (1 4 cm) except EA-1. The electrical resistivity of conductive adhesives EA-3 and EA-4, the silver-coated copper particle with the spherical shape, is furthermore lower than that of conductive adhesive EA-2, the silver-coated copper filler with the spinous shape. The electrical resistivity of EA-4 was lowest value in all EAs. It was concluded that the electrical resistivity of cured conductive adhesives are strongly affected by the surface condition and shape of the filler metal. Figure 4 shows the change rate of the electrical resistivity of the conductive adhesives as a function of the hightemperature exposure time. The exposure conditions were 398 K for 2 h and 1 h in air. In this figure, as the test time increases, the electrical resistivity of conductive adhesive EA-1 shows a significant increase after only 1 h and the resistivity could not be measured. The change rate of the electrical resistivity of EA-2, EA-3, and EA-4 is much slower. Even after 1 h, the electrical resistivity of EA-2, EA-3, and EA-4 is comparatively stable. This is in contrast to the oxide formation on the surface of the pure cupper particles. The change rate of electrical resistivity of EA-4 is slowest in all EAs and only EA-4 remains stable with change rate of less than 2% after high-temperature exposure for 2 h. In other words, conductive adhesive EA-4, the spherical 1% silver-coated copper filler, shows much more stable electrical resistivity than EA-2, the spinous 1% silver-coated copper filler, and EA-3, the spherical 5% silver-coated copper filler. This could be explained by referring to the uniformity and thickness of silver coating onto copper particles. Therefore, the stable electrical resistivity of EA-4 might be due to the stability of silver coating at 398 K in air, which could prevent further oxidation of copper surface. The metal oxides of pure copper and silver-coated copper particles in conductive adhesives were investigated by Auger electron spectroscopy. Auger electron spectrums for the particles in the conductive adhesives are shown in Figs. 5 and 6. Figure 5 plots the atomic concentration profiles of copper, oxygen, and carbon for the pure copper particle of EA-1 just after curing (Fig. 5) and after hightemperature exposure at 398 K for 1 h (Fig. 5) as a function of sputter time. Figure 6 shows the atomic concentration profiles of copper, silver, oxygen, and carbon for the silver-coated copper particle of EA-4 just after

4 1788 H. Nishikawa, S. Mikami, K. Miyake, A. Aoki and T. Takemoto Ag : : Ag : Ag : : Ag : Fig. 6 Atomic concentration of, Ag, and for silver-coated copper particle in EA-4: just after curing at 448 K for 1 h, after high temperature exposure at 398 for 2 h. curing (Fig. 6) and after high-temperature exposure at 398 K for 2 h (Fig. 6) as a function of sputter time. The particles in the conductive adhesive are covered with the polymeric resin, and so carbon derived from the polymeric resin is detected for all sputter time in Figs. 5 and 6. In the case of pure copper, shown in Fig. 5, the atomic concentrations of oxygen is comparatively low just after curing. n the other hand, after high-temperature exposure for only 1 h, the atomic concentrations of copper and oxygen increase up to 2 s. This means copper oxide exists near the surface of the pure copper particle. The electrical resistivity of EA-1 is strongly affected by the copper oxide formed on the surface during the high-temperature exposure. In contrast, in the case of silver-coated copper, shown in Fig. 6, both after curing and after high-temperature exposure, the atomic concentration of oxygen near the surface increases very slightly or not all, and so this means the copper oxide doesn t at first exist near the surface of the particle. It could be confirmed that silver coating remains near the surface after high-temperature exposure for 2 h. Therefore, as shown in Fig. 4, EA-4, the spherical 1% silver-coated copper filler, shows good electrical resistivity and stability even after high-temperature exposure for 2 h. Figure 7 shows the change rate of the electrical resistivity of the conductive adhesives during the temperature/humidity aging test. The conditions for this test were 358 K and 85%RH for 15 h. As in the high-temperature exposure test, the electrical resistivity of EA-1 shows increases quickly with increasing aging time and, after only 1 h, the resistivity could not be measured. In contrast, the change rate of the electrical resistivity of EA-2 and EA-4 is much slower. EA-2 and EA-4, the silver-coated copper filler, show improved stability during aging when compared to EA-1, the pure copper filler. This may be due to the antioxidation capability of silver-coated copper at elevated temperature and humidity conditions. Then, in the case of EA-4, it was found that the change rate of the temperature/ humidity aging test at 358 K and 85%RH was slower than that of the high-temperature exposure at 398 K, shown in Fig. 4. hange rate of electrical resistivity (%) onclusion Holdin time, t / h To overcome the problem of high electrical resistance due to the oxidation of copper filler, copper particles were coated with silver, and the silver-coated copper was evaluated as a filler metal. Especially, the effect of the silver coating on the electrical resistivity of a conductive adhesive just after curing and after reliability tests was investigated. It was found that the electrical resistivity of EAs using silver-coated copper filler was much lower and more stable than that of EA using pure copper filler after curing and after reliability tests. As a result of Auger electron spectroscopy, it was clear that copper oxide exists near the surface of pure copper particles after high-temperature exposure for only 1 h. The electrical resistivity of EA using pure copper was strongly affected by the copper oxide formed near the surface. The silver coating provides good oxidation resistance for copper particles. Acknowledgment : EA-1, : EA-2, : EA-4 ring temperature: 448 K ring time: 1 h Holding conditions: 358 K/85 % RH Fig. 7 Effect of silver-coating on electrical resistivity of EAs during temperature/humidity aging test. We would like to thank N. Terada from Harima hemical Inc., for kindly providing us the phenolic resin used in this study.

5 Effects of Silver oating overed with opper Filler on Electrical Resistivity of Electrically onductive Adhesives 1789 REFERENES 1) M. G. Perichaud, J. Y. Deletage, H. Fremont, Y. Danto and. Faure: Microelectron. Reliab. 4 (2) ) D. Lu and. P. Wong: Int. J. Adhesion Adhesive 2 (2) ) H. K. Kim and F. G. Shi: Microelectron. J. 32 (21) ) M. Yamashita and K. Suganuma: J. Electron. Mater. 31 (22) ) K. S. Moon, S. Liong, H. Li and. P. Wong: J. Electron. Mater. 33 (24) ) W. J. Jeong, H. Nishikawa, H. Gotoh and T. Takemoto: Mater. Trans. 46 (25) ) W. J. Jeong, H. Nishikawa, D. Itou and T. Takemoto: Mater. Trans. 46 (25) ) H. Jiang, K. S. Moon, J. Lu and. P. Wong: J. Electron. Mater. 34 (25) ) F. M. oughlan and H. J. Lewis: J. Electron. Mater. 35 (26) ) S. K. Kang, S. Buchwalter and. Tsang: J. Electron. Mater. 29 (2) ) H. P. Wu, J. F. Liu, X. J. Wu, M. Y. Ge, Y. W. Wang, G. Q. Zhang and J. Z. Jiang: Int. J. Adhesion Adhesive 26 (26) ). F. Goh, H. Yu, S. S. Yong, S. G. Mhaisalkar, F. Y.. Boey and P. S. Teo: Mater. Sci. Eng. B 117 (25) ) T. Liang, W. Guo, Y. Yan and. Tang: Int. J. Adhesion Adhesive 28 (27) ) W. Lin, X. Xi and. Yu: Synth. Met. 159 (29)