30 nm Ar sputt. temp. series
As a follow up to my 15 nm post-Si deposition Ar sputter experiment, I fabricated a set of six wafers with 30 nm Si films this week: three with the Ar sputter process and three without. I annealed these at 1000, 1050 and 1100 C with a ramp rate of 100 C/s for 60 s.
The first, second and third row represent the 1000, 1050 and 1100 C anneals, respectively. The first column are TEM images of films without the Ar sputter process and the second column represent films with the Ar sputter.
The size distributions indicate that the most noticeable difference occurs at the lowest RTP temperature, 1000 C. The density of pores of the Ar sputtered film is ~1.5x more than the non-Ar sputtered film. Curiously, the trend does not continue at higher annealing temperatures and the two films look very similar with respect to density and size at the 1050 and 1100 C RTP.
At this point, I think we conclude the Ar bombardment is perturbing pore growth under certain conditions. My next set of experiments will involve depositing the silicon using a bias gradient.
That is, I will start depositing the first 10 nm without a substrate bias and then turn on the bias for the last 5 nm of growth. In the past we have seen that a Si film without bias does not form many pores. Here, we are limiting the nucleation of crystals in the top 5 nm of Si which will then grow into the “no bias” region. I am hoping this will further increase crystal grain and pore size.