On the formation of the ordered phases CuAu and Cu 3 Au at a copper/gold planar interface
- PDF / 2,888,183 Bytes
- 7 Pages / 612 x 792 pts (letter) Page_size
- 55 Downloads / 158 Views
H E crystallographic order-disorder transformations which o c c u r over s e v e r a l composition r a n g e s in gold-copper alloys1 have been much studied classical reactions. Numerous references can be found dealing with the equilibrium structures, temperature and composition limits of stability, kinetics and thermodynamics of the order-disorder transformations, and arrangement of the ordered domains in the Cu3Au, CuAu, and CuAu3 superlaltices. The differences in crystallographic structure and the transformations between the s e v e r a l superlattices which can form at the same composition, namely CuAu(I) and CuAu(II) and also Cu~Au(I) and Cu~Au(II), have also been thoroughly investigated. T h e s e studies are normally c a r r i e d out on bulk or powder alloy samples of uniform composition which are fully homogenized at elevated temperature, quenched to a s t a t e of disorder, and then annealed at temperatures below about 400°C to study the ordering reaction. Another mode whereby this reaction can o c c u r and which has implications in current electronics technology is via diffusion in layered structures of copper and gold, which are present in connector components, hybrid circuit lead f r a m e s , and other metallizalions. The occurrence of diffusion at low temperatures, primarily via defect paths,2'3 and also at somewhat high temperatures, but s t i l l under 400°C, via vacancy-atom exchange through the crystal lattice4'~ is an unavoidable process. During this diffusion process, regions, which are initially near the gold/copper interface, are going to develop which have the proper compositions to be able t o form the o r d e r e d compounds. T u and B e r r y6 were a m o n g the f i r s t t o report the appearance of layers of Cu3Au, CuAu3 and, for longer annealing t i m e s , CuAu at the interface of copper-gold couples following diffusion anneals. They concluded that the ordered compounds form as sequential layers parallel to the original copper/bold interface with the Cu3Au phase appearing first, followed by the CuAu3
phase. The CuAu phase f o r m s much l a t e r in the r e a c tion and grows at the expense of the CuAua phase. They also concluded that atomic transport for growth o c c u r s by rapid boundary diffusion but that the rate limiting step in growth is the l o c a l atom movements that take place in the crystal lattice behind the advancing i n t e r face to s e c u r e o r d e r . The findings of Borders 7 differ from those presented above. No evidence of a growing l a y e r of any of the intermetallic compounds is reported. He suggests a model whereby the intermetallic phases grow into the parent g r a i n s of the gold and copper f i l m s from g r a i n boundaries which s e r v e as the diffusion paths. The r e sult is patches of Cu3Au in the co~3per film which are more dense near the original Cu/Au interface than within the parent copper film. This is in contrast to a distinct " l a y e r " as advanced by T u and Berry.6 H e also suggests that the CuAu3 phase is distributed uniformly as
Data Loading...
