AFM of film roughnesses
Chris asked me to do some scans of samples last week but I didn’t have time to get to them. I was able to scan them today since I was setting the AFM up for my class anyway.
I took 1um square scans (512×512) at 1Hz for all samples. The target and set point voltages were the same for all samples (1V and ~850mV, respectively). After scanning, I chose a 150nmx150nm square mask to calculate the roughness (the red box in the images below). The program also calculates the roughness of the entire scanned area. In summary, roughness increased from new ITC to pnc1 to old ITC but all had RMS roughness of <1nm.
pnc1 stack#6:
ITC2 (New 40nm):
ITC1(Old 20nm):
Update from comment(Feb 23) – Old bare Si wafer from SQI (200um thick)
nice results… that beg a question- what is the rms value given by the program actually measure? First, what would be the rms roughness of a silicon wafer? I’d like to see the number the program gets. Because 1 nm or less is really small, especially if we have pores… wouldn’t the AFM tip drop each time it sees a pore- or maybe it depends on the pore size, since the tip has a finite size? Also, for the sake or argument, let’s say we have identical pores of a given diameter and the AFM tip “sees” them. What if you do a Fourier analysis and remove the frequency components corresponding to the pore size- what would be the rms roughness?
I didn’t label these very well. It was just a baseline on the ITC deposition process. ITC1 and ITC2 were just oxides, and the pnc1 stack was the 3-layer stack, but not annealed…
Yep, no pores in any of these scans, just surface details.
I looked into the roughness calculations with this tool and couldn’t find any details. I suppose I could try it on my own in MATLAB.
I am going to attach an old scan of a bare 200um SQI wafer – it gave me a RMS roughness of ~0.4nm. It’s hard to compare roughness measurements from different days, though. Certainly, the tip size and sharpness matters, so I’m not sure what RMS roughness I would get with that wafer today. For the 3 scans in this post, I used a brand new tip and scanned them back to back, so there shouldn’t be too much difference in the tip size/shape.