We have deposited Aluminium nitride over gate oxide (SiO2) in a MOSFET and taken Aluminium contacts. It is generally accepted that fixed oxide charges are reduced on treating the sample with forming gas (10% H2 and 90% N2). What effect will it have on the stacked dielectric layer? Will it affect the number of surface sites on AlN? Generally, we do annealing for such dielectrics in pure N2 ambient. However, we need to sinter aluminium contacts which require forming gas. If we do not supply hydrogen, will sintering take place properly? What is the role of hydrogen in the process? In your opinion, which will be better - annealing in pure N2 ambient or annealing using forming gas?
1) AlN is stable in hydrogen - at least up to Al melting point. So no changes are expected.
2) The forming gas under heating forms a small amount of ammonia (~0.5%). That ammonia reduce the aluminum oxide, allowing reliable sintering.
with nitrogen only, sintering will be of awfully low quality.
I think to give a decisive answer one has to run the experiment with nitrogen only and with forming gas. Then i have doubt about sintering process the SiO2 and Al N interface if really a reaction runs as described by the colleague Oleg. If ammonia will reduced Al O2 it will affect some removal of the material from the surface. In the sintering process we want to affect bonding between the atoms at the surface. I think intermixing at the interface is effective in the sintering process. I f we look at the materials at the interface we find SiO2 and Al N, both materials can be intermixed in the sense Al. has affinity to oxygen likewise Si to Nitrogen at higher temperatures.
Best wishes
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Annealing