Si target composition has dramatic impact on morphology
Last week, we ran several wafers with a silicon target from Refining Systems (Las Vegas). In addition to the source material being different, the target bond was a silver elastomer vs. the typical indium bond that AJA uses. Below is a TEM micrograph of a comparison between the RS silicon and AJA silicon. Both these wafers were deposited last week. Both were annealed at 1000 C RTP 100 C/s. The RS silicon is slightly thicker ~ 20 nm.
The morphology of these two films are dramatically different. Both targets are rated intrinsic. Unfortunately, AJA does not provide an impurities spec so we can’t compare it to the RS target.
Based on these results, we are confident that Si target composition is a major contributor to the determination of pore morphology. The next step is to identify a series of Si target vendors and ask for target specs with different dopants and purities. Lesker is next on the list.
Interestingly, I was not able to model the optical properties of the the RS film by ellipsometry. It showed some uncharacteristic absorption peaks not normally seen in a-Si.