Posted by Roger Brennan (188.8.131.52) on July 01, 2005 at 03:37:31:
In Reply to: Question about Bismuth posted by Lee,DongKyu on June 28, 2005 at 00:21:08:
Hello Dong Kyu Lee,
I was delighted to receive an inquiry about bismuth. I vaguely remembered the diffusivity of Bi being reported as part of a list of n-type dopants in the RTI Diffusion Report (1964). Since Bismuth’s diffusivity is less that any of the popular dopants, it could be very useful.
Thank you for supplying the solid solubility limit of 8E17 at 1320 C. (Too bad it isn’t higher!) Do you plan to anneal at that high a temperature? Probably not -- so I assume you will get a lower carrier concentration. Still the concentration should be adequate for n-wells, diffused resistors, bases in pnp transistors and a lot of other things.
Perhaps the quickest way to find out if you can activate any of the bismuth is to implant it into a lightly doped p-type wafer and then four-point probe it after anneal. We can do this if you wish but you probably have your own four-point probe.
Things to watch out for:
1) Very minimum anneals (such as “flash” annealer can produce) can leave residual damage (so I’m told) in the region of the junction. This can cause false 4PP measurements.
2) Excessive bismuth, not substitutionally inserted could depress the mobility and cause other interferences.
At any rate, once you get a reliable 4PP measurement, it is probably time to get a spreading resistance profile. The sheet resistance calculated from the SRP should agree fairly well with the sheet resistance from the four-point probe. The SRP has proven useful for silicon doped with boron, phosphorus, arsenic, gallium, and antimony. It is also believed that it is okay for aluminum and indium although there is no data. So SRP will probably work for bismuth as well.
As for stripping oxide, we can take your samples "as is" and usually bevel and run them without difficulties. We have HF to strip oxide and aqua regia to strip metals if they are a problem. The only material we request you remove is passivation nitride. We do not have a way of stripping it. It is too thick and hard for us to bevel easily and smoothly so scratches on the bevel occur frequently.