Meta
About the Group
User Tools
Sidebar
mla150
mla150
mla150 is a maskless exposure tool. It can expose photoresist to 0.6um resolution on almost any size wafer with both frontside and backside alignment. Because no mask is required, unlike traditional photolithography tools, photoresist can be exposed half an hour after designing a layout. This also allows patterning gigantic features the size of the entire wafer with micron resolution. Exposing a 6“ wafer usually takes around 1.5 hours, though this varies with exposure settings.
Process Notes
| Date | Person | Data |
|---|---|---|
| 15 Jan 2021 | Daniel | Standard 2um MiR701 on picotrack exposes best on high quality mode with 180mJ/cm2 on silicon. Don't use fast mode because it leaves streaks along the mla150 exposure lines. Exposing a full 6” wafer takes about 1.5 hours. Exposure width may vary by 0.2um across different wafers; finer characterization may reduce this uncertainty. |
| 19 Jan 2021 | Daniel | 12um AZ P4620 deposited with Wei's picotrack recipe takes about 1000mJ/cm2 on fast mode and leaves hard-to-etch film even after development. Something needs fixing. |
| 9 Feb 2021 | Daniel | Standard AZ P4620 12um recipe on picotrack1 exposes best at 1000 to 1100mJ/cm2 when using fast mode (neither high quality nor high aspect ratio mode to save time). Descum on yes-g500 helps somewhat but may not be necessary (plate configuration 0A0(G*)00FF, 400W, O2, 60s) |
| 16 Jun 2021 | Daniel | Heidelberg field service engineer tuned up tool. Now: optical autofocus is best at focus -2, pneumatic autofocus is best at focus +4, top side alignment accuracy within 300nm (spec 500nm), bottom side accuracy within 850nm (spec 1um), and required exposure dose should be 50% to 75% higher than previous recorded values. |
| 12 Jul 2021 | Daniel | Spun standard 2um MiR701 on picotrack after a primeoven recipe 2 on a fused silica wafer. Using pneumatic autofocus, a dose of 210mJ/cm2 and focus +4 exposed the pattern fairly well, +/- 0.2um, but did not do full dosage test. |
| 10 Oct 2021 | Daniel | Ran exposure test for AZ P4620. Use primeoven recipe 2 for HMDS, then spin 12um AZ P4620 with standard picotrack recipe (optionally no TEBR). Ideal exposure was at 800mJ/cm2, 0 defocus. Higher exposures will make large nitrogen bubbles in the resist (this is probably a result of heating, so it could be solved by exposing more slowly or in multiple shorter exposures, but that isn't an option on mla150); lower exposures will not completely develop. Can run through standard picotrack2 develop recipe twice for better effect (not all exposed PR is completely removed the first time; thick resists are tricky). Any further descum or hardbake will reflow the photoresist; I don't have a way around this yet. |
| 11 Oct 2021 | Daniel | Exposed AZ P4620 with < 20% ambient humidity in NanoLab (there's a humidity meter display on the wall by svgcoat3). The resist failed to develop! Apparently DNQ-based photoresists need water in the resist, delivered via ambient humidity, to expose properly. This is why the resist is supposed to sit for ~30 minutes in ambient air after spinning before exposure. Solution: wait for a time when NanoLab humidity is sufficiently high (e.g., > 40% is probably good) (it's not actively controlled because it's almost always > 40% anyway), or put wafer into humidity box near zeiss-sem. |
mla150.txt · Last modified: 2021/11/30 17:36 by dteal
