https://nanyte.com/photoresists/su-8-2005 · last updated 2026-09-28
- Manufacturer
- Kayaku Advanced Materials
- Tone
- negative
- Chemistry
- Epoxy (SU-8 type)
- Thickness
- 4.6–7.8 µm
- Developer
- SU-8 Developer; ethyl lactate or diacetone alcohol may also be used
- Applications
- MEMS structural · High aspect ratio · Electroplating / molding
Cross-checked — two independent extractions agree on the spin curve and the single-value figures.
- Substrate
- Resist
- Exposed
Spin coating
Data points
| Series | rpm | µm | Published dose |
|---|---|---|---|
| SU-8 2005 | 1000 | 7.8 | 110–140 mJ/cm² · 6–15 µm row |
| 1500 | 6.4 | 110–140 mJ/cm² · 6–15 µm row within 10 % of a bracket edge | |
| 2000 | 5.8 | — | |
| 2500 | 5.3 | — | |
| 3000 | 5.1 | — | |
| 3500 | 4.7 | 90–105 mJ/cm² · 3–5 µm row within 10 % of a bracket edge | |
| 4000 | 4.6 | 90–105 mJ/cm² · 3–5 µm row within 10 % of a bracket edge |
Values are the manufacturer’s starting points, not a guarantee — characterize on your own tool. Series digitized from a published figure are approximate (±10 %).
Published dose is the datasheet’s own thickness bracket, quoted as printed — the full table and its citation are in the exposure section. A thickness the table does not cover shows nothing, and a thickness within 10 % of a bracket edge shows every bracket it could fall in.
read from figure — the green round markers legended 'SU-8 2005' on Figure 1b 'SU-8 2000 Thickness vs. Spin Speed', p.2 of the SU-8 2000, 2000.5-2015 Technical Data Sheet, ver. 1, August 2020. Seven plotted markers at 500 rpm steps; no numeric table accompanies the chart. Reads are calibrated against the 5 µm axis ticks of the embedded chart image.
- From 7.8 µm at 1000 rpm the curve flattens quickly — above 2000 rpm every extra 500 rpm takes off only a few tenths of a micron, ending at 4.6 µm at 4000 rpm.
- Kayaku's recommended program for the series: dispense 1 ml of resist per inch (25 mm) of substrate diameter, spread at 500 rpm for 5-10 s at 100 rpm/s, then spin at 2000 rpm for 30 s at 300 rpm/s, choosing the final speed from the spin chart.
- Remove the edge bead with EBR PG so the mask can seat in close contact.
- Adhesion
- HMDS not required — 'Adhesion promoters are typically not required.' Clean and dry the substrate first — piranha (H2SO4 and H2O2) with a DI rinse, or RIE or an oxygen barrel asher. For applications that include electroplating, an HMDS pre-treatment is recommended.
Soft bake
- Soft bake
- 95 °C · hotplate
- Notes
- Times are set by film thickness: 3-5 µm → 2 min and 6-15 µm → 2-3 min at 95 °C. Use a level hotplate — convection ovens are not recommended because a skin can form and trap solvent. To check the bake, cool the wafer to room temperature and return it to the hotplate; if the film wrinkles, bake a few more minutes and repeat until it does not.
SOURCE: Table 2 'Soft Bake Times', p.3 of the SU-8 2000, 2000.5-2015 Technical Data Sheet, ver. 1, August 2020
Exposure dose
SU-8 2005's dose is published against film thickness, not as a single number. Read the row for the film you coat and run a dose array around it.
- As published
- Dose follows the film you spin, not the grade name: 3-5 µm films take 90-105 mJ/cm² and 6-15 µm films 110-140 mJ/cm². On glass and most metals multiply by 1.5-2.
- Post-exposure bake
- 95 °C · hotplate
| Film thickness | Dose |
|---|---|
| 0.5–2 µm | 60–180 mJ/cm² |
| 3–5 µm | 90–105 mJ/cm² |
| 6–15 µm | 110–140 mJ/cm² |
| 16–25 µm | 140–150 mJ/cm² |
| 26–40 µm | 150–160 mJ/cm² |
SOURCE: SU-8 2000 2000.5-2015 Technical Data Sheet
Table 3 'Exposure Dose', p.3 of the SU-8 2000, 2000.5-2015 Technical Data Sheet, ver. 1, August 2020 — indexed by film thickness and shared by the six grades that sheet covers; no wavelength is attached to the table.
Development
- Developer
- SU-8 Developer; ethyl lactate or diacetone alcohol may also be used
- Method
- immersion, spray or spray-puddle
- Rinse
- Spray and wash the developed image with fresh SU-8 developer for about 10 s, then air-dry with filtered pressurised air or nitrogen.
- Developer family
- Solvent
Not published for this resist: Dilution, Time — characterize on-tool.
SOURCE: SU-8 2000 2000.5-2015 Technical Data Sheet
Development section and Table 6 'Development Times for SU-8 Developer', p.4 of the SU-8 2000, 2000.5-2015 Technical Data Sheet, ver. 1, August 2020
Hard bake, etch & strip
- Hard bake
- 150–250 °C · 5–30 min (optional — when the resist stays on the device)
- Descum
- Not published — characterize on-tool
- Etch resistance
- Not characterised as an etch mask. Once cross-linked the film is described as chemically and thermally stable — 315 °C at 5 % weight loss, a glass transition of 210 °C — and extremely difficult to remove with solvent strippers.
- Stripper
- Minimally cross-linked SU-8 2000 swells and lifts off in Remover PG at 50-80 °C for 30-90 min; with a 30-100 nm OmniCoat underlayer the lift-off is clean. Fully cured or hard-baked SU-8 2000 cannot be removed without OmniCoat. To rework fully cross-linked films: piranha, plasma ash, RIE (200 W, 80 sccm O2, 8 sccm CF4, 100 mTorr, 10 °C), laser ablation or pyrolysis.
- Storage
- Store upright in tightly closed containers in a cool, dry place away from direct sunlight, light, acids, heat and ignition sources, at 4-21 °C (40-70 °F). Shelf life is thirteen months from date of manufacture.
SOURCE: Hard bake (cure) section, p.5 of the SU-8 2000, 2000.5-2015 Technical Data Sheet, ver. 1, August 2020
Where it's used
Practical notes from the datasheet
SU-8 2005 is the grade for permanent epoxy structures around five microns, thicker than SU-8 2002 can reach. Its curve is flat above about 2000 rpm, which makes the thickness forgiving of small speed errors but means spin speed is a weak lever; to go meaningfully thicker, step to SU-8 2007 or 2010 rather than slowing down. A 5 µm film straddles the sheet's 3-5 and 6-15 µm rows, so bake about 2 minutes at 95 °C, expose in the 90-140 mJ/cm² band on silicon and more on glass or metals, post-bake 2-4 minutes, and develop 1-3 minutes in SU-8 Developer. The cured film is permanent — use OmniCoat underneath if it has to be released.
Grades in this family
Other grades in the SU-8 2000 series line differ mainly in coating thickness:
Troubleshooting
Common failure modes for SU-8 2005, answered from the manufacturer's datasheet and application notes. These are starting points — your substrate, tooling and environment shift the specifics, so calibrate on-tool.
SU-8 2005 images well on silicon but washes off gold and copper at the same dose. Why?
Metal substrates need more exposure. The datasheet's Table 3 doses are for silicon (1X); gold, copper, aluminium, nickel, NiFe, titanium and silicon nitride take 1.5–2X, and glass, Pyrex and ITO 1.5X. For SU-8 2005 the silicon starting point is 90–105 mJ/cm² in the 3–5 µm row or 110–140 mJ/cm² in the 6–15 µm row, depending on where your spin speed puts the film (about 4.6–7.8 µm across 4000–1000 rpm). Rerun the exposure matrix on the actual substrate.
SOURCE: SU-8 2000 (2000.5–2015) Technical Data Sheet
Kayaku SU-8 2000 (2000.5–2015) Technical Data Sheet, ver. 1, Aug 2020 — Figure 1b, p.2; Table 3 Exposure Dose, p.3; Table 4 Exposure Doses for Various Substrates, p.4
Can I soft bake SU-8 2005 in a convection oven instead of on a hotplate?
Kayaku advises against it. In a convection oven a skin can form on the resist that blocks solvent from leaving, so the film ends up incompletely dried or needs extended bake times. Use a level hotplate with good temperature control and uniformity: 2 min at 95 °C for 3–5 µm films, 2–3 min for 6–15 µm. Confirm dryness by cooling the wafer, returning it to the hotplate and watching for wrinkles; if it wrinkles, bake a few more minutes and repeat.
SOURCE: Kayaku SU-8 2000 (2000.5–2015) Technical Data Sheet, ver. 1, Aug 2020 — Soft Bake and Table 2 Soft Bake Times, p.3
My contact mask won't sit flat on the SU-8 2005 wafer and resolution suffers. What's wrong?
Check the edge bead. Spin coating builds up a thicker rim of resist at the substrate edge; it contaminates the hotplate and holds the mask off the film. Remove it with a small stream of solvent (EBR PG) at the wafer edge, from the top or the back — most automated coaters can be programmed to do this. With the bead gone, the mask can come into close contact, which the datasheet credits with improved resolution and aspect ratio.
SOURCE: Kayaku SU-8 2000 (2000.5–2015) Technical Data Sheet, ver. 1, Aug 2020 — Edge Bead Removal (EBR), p.3
My SU-8 2005 is behaving differently from last year's bottle. Could it have aged?
Check the date of manufacture first: the published shelf life for SU-8 2000 is thirteen months from that date. Then check how the bottle was kept. Kayaku specifies upright, tightly closed containers in a cool, dry place out of direct sunlight at 4–21 °C (40–70 °F), away from light, acids, heat and ignition sources.
SOURCE: Kayaku SU-8 2000 (2000.5–2015) Technical Data Sheet, ver. 1, Aug 2020 — Storage, p.6
Sources & disclaimer
- Kayaku Advanced Materials — SU-8 2005 datasheet (SU-8 2000, 2000.5-2015 Technical Data Sheet, ver. 1, August 2020) · accessed 2026-09-10
Manufacturer datasheet values are starting points; optimal parameters depend on your substrate, equipment and environment. Product names and trademarks belong to their respective owners. NANYTE is not affiliated with the manufacturers listed. Last updated 2026-09-28.