to calibrate our new DC/RF sputter system, we have deposited Pt film on SiO2/Si wafer (100 nm SiO2 on Si) and measured the thickness using ellipsometry. the Pt film thickness supposed to be around 270 nm according to other samples, but the ellipsometry data shows the thickness around 50 nm, also thee are two sharps peaks raw data. are those peaks normal? is there any better model to fit the data?
Thanks
Why not use a surface profiler to calibrate the film thickness? It was found that Elliposometry sometimes gives you not realistic readings which very much depend on the model used in the software. Please be aware of that the thickness values come from fitting parameters not measurements. The surface profiler gives you the measured values of thin film thickness.
Ellipsometry only works on films that are transparent enough that some of the light reaches the bottom of the film and is reflected by the film/substrate interface. Without reflection at that interface, there is no way for the technique to be sensitive to thickness.
My guess is 120nm Pt is too thick for ellipsometry, but you can google the optical constants of Pt and check. Generally, I'd recommend surface profiling for metals.
No way to use ellipsometry to find reliable thickness of Pt thicker than 30nm.
I agree with Xuhua Wang , When I would do ellipsometry on materials I would always have a physical sample typically AFM in my case. This allows you to verify the thickness to see if the ellipsometry model is correct. If you're doing a metal you can simple either pattern resist to make a step or just put a piece of tape on your sample during the dep.
-Adam
If possible you can try X-ray reflectivity of the Pt film to get a good estimate of the thickness.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators