When I measure the lifetime of wafers, using the instrument from Sinton, I find a strange result.
The group with HF etching to remove native oxide on Si wafers has higher lifetime compared to with native oxide on both sides. I have implemented the experiments many times, the result does not vary much.
This is not what I am expecting at the first place. Without native oxide on wafer, it should have more recombination centers and thus lower lifetime compared to passivation surface (SiO2 on surfaces).
Anyone has a reasonable explanation to it?
This is normal. The native oxide is not a good oxide, definitely. When you do a HF-dip, you remove this shity oxide and, moreover, all the dangling bonds becomes terminated by hydrogen from the HF solution. This "passivation" is not a real one, as the Si-H bonds formed like that will not survive (less than 1 hour).
So, the lifetime is better on the wafer because: bad native oxide and because the others are H-passivated.
Best regards.
this is typical effect when you do lifetime measurements. What you measure is an effective lifetime, which is affected by bulk and surface recombination. when you you do HF treatment, you change surface lifetime, thus changing effective lifetime. the effect is stronger for thin wafers. the lifetime is dominated by the surface lifetime for thin epitaxial films.
there are many good publications on this subject.
i recommend you start with D. Schroder book - Semiconductor Material and Device Characterization.
- Badges
- Science topic