Surface topology of amorphous silicon films deposited at different substrate biases
We have always speculated that applying a substrate bias during the silicon deposition “smooths” the film by creating a nicer film. With Graham’s help, I took a few AFM scans of films deposited at three different biases: 00, 05, and 50W.
(a) 00 W: 0.337 nm RMS (b) 05 W: 0.145 nm RMS (c) 50 W: 0.123 nm RMS
As expected, a film deposited with no bias is the roughest. Application of a low bias smooths the film considerably and a further increase of power only yields a marginal improvement in roughness.
When you get new tips, I would redo the 0W condition, since it has substantial nonuniformity.
IN addition to the roughness differences, the 0W film appears to have much finer structure, which may be even more critical than the roughness.
There might be 1 or 2 more unused cantilevers in the box I left in the lab.
That 0W scan caught my eye, too. There is an awfully abrupt transition from the ‘textured’ to the smooth surface at the bottom of the scan. Also the background roughness completely disappears about 2/3 of the way down the scan. Are you sure the tip was tracking for the whole scan?
I believe the tip was in repulsive for the entire scan and the phase image showed no large fluctuations. I agree, though, that the 0W is the outlier. I’ll do that one again.
Where in the lab?