Direct-Write Lithography Recipes
Maskless Aligner (Heidelberg MLA150)
For CAD design tips and requirements, see these pages:
- Design Guidelines + Tips - useful info for designing your CAD files, alignment marks etc.
- CAD Files and Templates - example CAD files etc.
Photolithography Recipes for the Heidelberg MLA150. All recipes were characterized on blank Silicon wafers. For different substrate coatings/materials, you will likely need to run a focus-exposure matrix ("series" exposure mode), using our params as a starting point.
These recipes use the same spin and bake params as our contact aligner and stepper recipes, using built-in hotplates on the photoresist spinner benches.
Any I-Line PR is usable, although we only characterized a select few below. Run your own Focus-Exposure Matrix to characterize a new PR.
Note: On this tool, it is common to have to run a Focus-Exposure Matrix (aka. FEM), via "Series" exposure mode, for each new layer structure and/or critical feature size. The layer structure can strongly affect the exposure parameters. See the FEM Analysis Tips page for how to choose the proper exposure parameters.
The MLA Troubleshooting > Out-of-Focus Exposures section can help you avoid bad exposures, please read it!
Positive Resist (MLA150)
We found that positive PR's worked well with the 405nm laser - the 375nm laser would likely also work. Sub-micron features are possible. Overexposure is recommended to avoid stitching and horiz/vert discrepancies; compensate for widening/narrowing using CD Bias as described here.
Note: calibrations done with the "HIMT design" will show higher dose, due to using only dark-field line/space patterns.
| Resist | Spin Cond. | Bake | Thickness | Laser (nm) | Exposure Dose (mJ/cm2) | DeFocus | Rehydrate* | PEB** | Developer | Developer Time | Comments |
|---|---|---|---|---|---|---|---|---|---|---|---|
| AZ4110 | 4 krpm, 30s | 95°C, 60s | ~ 1.1 µm | 405 | 240 | 5 | none | AZ400K:DI 1:4 | 50s | Used HIMT design (good for isolated lines 0.8-1um) | |
| AZ4330 | 4 krpm, 30s | 95°C, 60s | ~ 3.3 µm | 405 | 320 | 6 | none | AZ400K:DI 1:4 | 90s | Used HIMT design | |
| AZ4620 | |||||||||||
| SPR 955-1.8 | 4 krpm, 30s | 95°C, 90s | ~ 1.8 µm | 405 | 210 | 10 | 110°C, 90s | AZ300MIF | 60s | Used UCSB design (1um dense lines) | |
| SPR 220-3.0 | 2.5 krpm, 30s | 115°C, 90s | ~ 2.7 µm | 405 | 325 | - 4 | 115°C, 90s | AZ300MIF | 60s | Used HIMT design. 0.6-0.9µm line/space. | |
| SPR 220-7.0 | 3.5 krpm, 30s | 105°C/2min
Cool 1min |
~ 7.0µm | 375 | ~550mJ | -20 | >1hr | 115°C, 90s | AZ300MiF | 70s | Rehydration after exposure is necessary, to prevent bubbles at PEB. |
| SPR 955-CM0.9 | 3 krpm, 30s | 95°C, 90s | ~ 0.9 µm | 405 | 250 | - 7 | 110°C, 90s | AZ300MIF | 60s | Used HIMT design | |
| THMR-3600HP | 1.5 krpm, 45s;
250 rpm/s |
100°C, 60s | 0.430µm | 405 | 180–220 | -4 | 100°C, 60s | AZ300MiF | 20s | ~0.4nm line/space:
lower dose for clear-field, higher dose for dark-field. | |
| *Rehydration: After exposure, leave sample in lab air (ok to cover in tray, with tinfoil). Allows water vapor in air to diffuse into PR to complete chemical reaction.
**PEB: Post-exposure bake: after exposure, before develop | |||||||||||
Negative Resist (MLA150)
We found that all the negative PR's we tested required the 375nm in order to be fully exposed with reasonable dose/time.
| Resist | Spin Cond. | Bake | Thickness | Laser (nm) | Exposure Dose (mJ/cm2) | DeFocus | PEB* | Flood** | Developer | Developer Time | Comments |
|---|---|---|---|---|---|---|---|---|---|---|---|
| AZ5214** | 6 krpm, 30s | 95°C, 60s | ~ 1.0 µm | 375 | 35 | - 5 | 110°C, 60s | 60" | AZ300MIF | 60s | Used UCSB design. Good for up to ~1.3um open line space. |
| AZnLOF2020 | 4 krpm, 30s | 110°C, 60s | ~ 2.1µm | 375 | 340 | - 3 | 110°C, 60s | none | AZ300MIF | 90s | Used UCSB design. Good for 2um open line space. |
| SU-8 2075 | ~70µm | 375 | Extremely viscous. Pour into a wide-mouthed bottle, dispense directly from bottle. Replace napkin at end. | ||||||||
| *PEB: post-exposure bake. For AZ 5214-IR, this performs Image Reversal
** To use AZ5214 as a negative PR requires Flood Exposure with the MA6 or MJB aligner after PEB, before developing. See here for a basic AZ5214 process, it is different than typical negative resists. | |||||||||||
Greyscale Lithography (MLA150)
AZ4620 is the manufacturer-recommended PR for greyscale litho.
Please see the MLA150 - Greyscale Design Guidelines & Limitation
| Resist | Spin Cond. | Bake | Thickness | Laser | Exposure Dose (mJ/cm2) | Focus Offset | Rehydrate* | PEB** | Developer | Developer Time | Reflow*** | Comments |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AZ4620 | ? krpm/30” | 95°C, 60” | AZ300MIF | 60s | To Be Added | |||||||
| SPR 220-7.0 | 3.5 krpm, 30s | 105°C/2min
Cool 1min |
~ 7.0µm | 375 | ~624mJ to clear large mm-area,
520mJ to clear ~5µm lines. |
-20 | ≥1hr | 115°C, 90s | AZ300MiF | 70s | TBD | Author Credit:
|
| Notes on SPR 220-7 Greyscale: Rehydration after exposure is necessary, to prevent bubbles at PEB.
Stitching leaves ridges in Y-direction with ~5% height of removed PR depth. Can be reduced via reflow, but significantly affects PR profile. Author Credit: Patrick Curtis, 2022 & Biljana Stamenic 2023 & Demis D. John 2023; Please see our publications policy. | ||||||||||||
| *Rehydration: After exposure, leave sample in lab air (ok to cover in tray, with tinfoil). Allows water vapor in air to diffuse into PR to complete chemical reaction.
**PEB: Post-exposure bake: after exposure, before develop ***Reflow: To smooth out stitching lines. Will change sharp vertical profiles considerably, only good for gradually-sloped profiles. | ||||||||||||
E-Beam Lithography Recipes (Raith EBPG 5150+)
Electron Beam Resists Available
EBL Resists are custom-mixed according to user resolution needs - contact Bill Mitchell to get a bottle.
Currently available at UCSB are:
- PMMA: (950K in anisole, 950K in MIBK, 495K in anisole, 50K in anisole)
- High-resolution positive polymer-based resist with relatively poor sensitivity (resolution scales directly and sensitivity scales inversely with molecular weight);
- Poor plasma etch resistance, hence used primarily to fabricate metal lines via liftoff processes (via a bi-layer resist scheme...low MW on bottom, high MW on top for single lines, or vice-versa for T-gate fabrication);
- Utilizes an inert solvent developer (usually MIBK:IPA mixtures)
- P(MMA-MAA) copolymer: (low MW methyl-methacrylate (MMA) and methacrylic acid (MAA) copolymers in ethyl lactate)
- A positive polymer-based resist with poor resolution but with significantly higher sensitivity than the higher MW PMMA resists above
- Used primarily as the top layer in a bi-layer resist scheme for T-Gate fabrication, and utilizes inert solvent developer (MIBK:IPA mixtures)
- CSAR-62: ZEP-equivalent resist manufactured in Germany at much more competitive pricing!
- High-resolution polymer-based positive resist with very good sensitivity and excellent etch resistance
- Can be used in both metal lift-off processes (slight overexposure results in an excellent undercut profile) and various dry-etch processes for pattern transfer to the underlying substrate
- Utilizes inert solvent developers (e.g., n-amyl acetate for higher sensitivity and good resolution or MIBK:IPA mixtures for increased LER performance)
- maN-2403: negative polymer-based resist (that is NOT chemically amplified)
- Very good resolution (down to the 40-50nm range) and sensitivity
- Exhibits excellent dry-etch resistance
- Developed using a dilute basic solution (e.g., metal-ion-free developers such as AZ-300MIF)
- HSQ: negative resist that is based on spin-on glass material (ie, not polymer-based)
- Extremely good resolution (features below 10nm can be resolved)
- Etch resistance is high in Cl-based chemistries since HSQ reduces to a porous SiOx structure after exposure and development
- Sensitivity and contrast are very dependent on developer solution used and are usually poor
- Standard AZ300MIF developer solutions have decent sensitivity (100's of uC/cm2 at 100kV) but extremely poor contrast
- Stronger (and toxic!) 25%TMAH solutions have much better contrast but poor sensitivity (1000's of uC/cm2 at 100kV)
- "Salty" developer solutions using 1wt% NaCl dissolved in either 4wt% NaOH or AZ300MIF solutions have the best contrast but reduce sensitivity significantly (10,000's of uC/cm2 at 100kV)
- DUV resists: UV6, UV210 UVN-30: chemically amplified polymer-based resists
- High resolution and excellent sensitivity (clearing doses below 100uC/cm2 at 100kV)
- UV6 used mostly in optimized t-gate resist structures
- Developed using a dilute basic solution (e.g., metal-ion-free developers such as AZ-300MIF)
- Can be "double exposed" by ASML DUV Stepper and EBL.
- Recommended to produce ASML alignment marks first for double exposure methods.
- See ASML litho recipes for datasheets.
EBL Exposure Recipes
To Be Added - BEAMER simulation is always required for small (<<micron) features.