Interface nanochemistry effects on stainless steel diffusion bonding
- PDF / 351,536 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 98 Downloads / 245 Views
I. INTRODUCTION
DIFFUSION bonding has been widely applied to metals, such as aluminum and titanium alloys and stainless steels, and to silicon and other solid-state electronic materials, where it is generally termed direct wafer bonding, to join materials for special purposes where relatively large contact areas are involved, and where other familiar deposition or joining processes such as chemical vapor deposition or welding are not generally applicable.[1,2,3] For metals, theoretical descriptions of the process include time-independent local plastic deformation of asperities due to microscopic roughness on the substrate surfaces upon initial contact, followed by time-dependent deformation and mass-transport processes that eliminate the interfacial voids between the aperities.[4,5] For semiconductors, the lattice defects created by even local plastic deformation are undesirable, thus requiring smoother initial substrate surfaces, and emphasis is given to control of the interface substrate chemistry, which strongly influences the kinetics of interface bonding. The presence of ⬃3-nm-thick films of amorphous SiOx on (001) Si and vicinal 6H (0001) SiC enhanced the bonding kinetics relative to the same substrates with the oxide removed by thermal desorbtion, when bonded in ultrahigh vacuum (UHV) at 1100 ⬚C and 1 MPa.[6] Of the experimental variables used for metal bonding, which are pressure, temperature, time, ambient conditions, substrate surface roughness, and substrate surface chemistry, the effects of the latter three variables on diffusion bonding are the least investigated. Differences between theoretically predicted bonding kinetics and the experimentally observed rates for microduplex stainless steels containing ⬃25 pct Cr have been attributed to substrate surface oxides, although experimental measurements of substrate surface chemistry prior to bonding have not been reported.[7] Other studies of surface roughness and void distributions on diffusion-bonded interfaces indicated that these parameters may also have been M.J. COX, Graduate Student, and R.W. CARPENTER, Professor, are with the Center for Solid State Science, Science and Engineering of Materials, Arizona State University, Tempe, AZ 85287-1704. M.J. KIM, Professor, is with the Department of Materials Science, University of North Texas, Denton, TX 76203. Manuscript submitted July 13, 2001. METALLURGICAL AND MATERIALS TRANSACTIONS A
responsible for the disagreement.[5,8] A distinction between the roles of substrate surface oxides and roughness would be useful. We have used special equipment for the measurement and control of substrate surface oxidation to investigate this effect. For the present work, we conducted bonding experiments on AISI 304 austenitic stainless steel (⬃18 pct Cr) in a very well controlled chemical environment with low applied loads on specimens with very small substrate surface roughness, to evaluate the effect of substrate surface oxides on interface bonding. We used a unique multichamber UHV bonding instrument that contains an Au
Data Loading...
