Migrated from the old Google Sites wiki (Nov 2026). Text and numbers were carried over; photos and screenshots were not. If you have them, add them to assets/wiki/ and link them here. Check numbers against your own run before relying on them.
Recipe
- On the ALD tool, open the folder named for Zihuan Liu. Recipe: AL2O3_300cycles.
- Expected result: about 60 nm of Al₂O₃ from 300 cycles.
Thickness vs. cycles
| Cycles | Measured thickness | Implied rate |
|---|---|---|
| 300 | ≈ 60 nm (ellipsometer) | ≈ 0.2 nm/cycle |
| 2000 | ≈ 200 nm | ≈ 0.1 nm/cycle |
| 700 | result shown in photos on the old page |
The original page quotes roughly 0.1 nm per cycle and notes that growth is not linear in cycle count. The 300-cycle and 2000-cycle numbers above disagree with a single rate, so measure your own film before relying on either.
To do: whoever next runs this recipe, add your cycles, measured thickness and date to the table.
What it looks like
After the backside SiO₂ of an SOI wafer is removed and 60 nm of Al₂O₃ is deposited, the backside turns noticeably darker. Confirm thickness on the ellipsometer.
Edge deposition on the front side
Some Al₂O₃ wraps around the wafer edge onto the front side.
- Usually not a problem: a Piranha clean removes it.
- If your process cannot take Piranha, clip a steel plate over the front side during deposition.
Something wrong or out of date? Fix it: edit this page on GitHub. See Editing this site.