SU-8 process recipe (spin-coating, soft bake, dose, PEB, development)
Spin-coating speed, soft bake, substrate-corrected dose, exposure time, PEB and development, by SU-8 product, with their sources.
Result
Spin coating
1628 rpm
Coating speed for the target thickness.
Provenance: read from the datasheet figure then interpolated, see the uncertainty
MicroChem, SU-8 2000 Permanent Epoxy Negative Photoresist — Processing Guidelines for SU-8 2025, 2035, 2050 and 2075, aucune révision imprimée, fichier SU-8-2075-revA.pdf (UCSB Nanofab), SHA-256 e82d7ef9e46e2c8e…, Figure 1, Spin Speed versus Thickness, consulted on 2026-07-30
- Interval
- 1473 – 1819 rpm
- Reading uncertainty
- ± 1 %
- Agreement between manufacturer editions
- ± 10 %
Process recipe: SU-8 2025 · 50 µm(read from the 45–80 µm band)
Solid bar: minimum duration of the step. Light bar: up to the published maximum. The total does not count the exposure, with no irradiance entered, nor the temperature ramps, cool-downs and handling.
Exposure
Notes2
- Enter your lamp’s measured irradiance to obtain an exposure time. Without it, only the dose can be given. The time depends on your equipment, not on the resist.
- Other SU-8 products also cover this thickness, sometimes with different recipes because they come from a different datasheet. This is why the product is asked for before the thickness.
Product characteristics
- Viscosity
- 4500 cSt
- Solids content
- 68.55 %
- Density
- 1.219 g/mL
- Thickness range read from the figure
- 22 – 79.55 µm
What the datasheet warns about4
- The datasheet presents these values as a starting point to be optimised, not as a validated recipe for a given process.
- The 25–40 µm band is also covered by the thin-formulation datasheet, with a different soft bake (4–5 min at 95 °C instead of 5–6). Use the datasheet matching the product actually in use.
- With a long-pass filter removing UV below 350 nm, the datasheet states that exposure time must be increased by roughly 40 %.
- A 1.5–2× range signals variable reflection: start at the lower bound and refine with an exposure matrix.
Domain of the datasheet: Thicknesses from 25 to 225 µm, on silicon, i-line (365 nm) exposure. The 65 °C PEB step is optional (stress reduction).
Value traceability2
- MicroChem, SU-8 2000 Permanent Epoxy Negative Photoresist — Processing Guidelines for SU-8 2025, 2035, 2050 and 2075, Ver. 4, Tables 2, 3, 5, 6 · consulted on 2026-07-28
- MicroChem, SU-8 2000 Permanent Epoxy Negative Photoresist — Processing Guidelines, Rev. 1, 11-01-06 et Ver. 4 (valeurs identiques dans les deux), Table 4 · consulted on 2026-07-28
Other products covering this thickness9
Their recipes come from other datasheets and may differ.
SU-8 50 · SU-8 2035 · SU-8 2050 · SU-8 2075 · SU-8 3005 · SU-8 3010 · SU-8 3025 · SU-8 3035 · SU-8 3050
Calculation, assumptions and sources
OverviewFor an SU-8 family, a product and a thickness, the spin-coating speed and the datasheet recipe: soft bake, substrate-corrected exposure dose, exposure time, PEB, development.
For an SU-8 family, a product and a thickness, the spin-coating speed and the datasheet recipe: soft bake, substrate-corrected exposure dose, exposure time, PEB, development. The speed is a tabulated value, an interpolation between published points or a figure reading, with its uncertainty. Recipe parameters are never interpolated, and no speed extrapolation is made outside the published domain.
Method & formulascorrected dose = dose on silicon × substrate multiplier
Soft bake, PEB and development are read from the manufacturer’s table. The corrected dose and the exposure time are computed: reference dose × tabulated factor, then division by the measured irradiance. Interpolation, which is flagged, concerns only the spin speed.
Datasheet doses assume a silicon substrate. On another substrate, UV reflected at the interface passes back through the film and the absorbed energy changes. The SU-8 2000 datasheets publish a multiplier: 1.5× for glass, Pyrex and ITO; 1.5 to 2× for gold, aluminium, copper, nickel and titanium. A range reflects the dependence of reflectivity on surface condition. Start at the lower bound and refine with an exposure matrix.
For the classic series, each datasheet publishes a table of three exact speeds per product; an interpolation between two points is flagged. For the SU-8 2000 and SU-8 3000 series, the thickness–speed curve exists only as a vector figure, whose markers are read from the PDF content stream to a hundredth of a point. Two uncertainties are declared separately. The reading is 1 %. The agreement between manufacturer editions is 10 %, measured against the tables of the Rev. 2/02 brochures (median deviation 2.4 %, 93 % of deviations under 10 %, extreme 23.2 % for SU-8 2035 at 3000 rpm). The displayed interval propagates the wider of the two.
The datasheet gives a dose, in mJ/cm². The exposure time is dose ÷ irradiance: 150 mJ/cm² takes 15 seconds at 10 mW/cm². Irradiance is measured with a radiometer at i-line (365 nm), the reference wavelength of the SU-8 datasheets. Without an irradiance, only the dose is given.
- Soft bake
- Bake after coating, which evaporates solvent and densifies the film. Too short and the film sticks to the mask; too long and it becomes brittle and adheres poorly.
- Exposure
- UV exposure. In SU-8, a negative resist, it forms a strong acid in the illuminated areas; the pattern is not yet fixed.
- PEB (post-exposure bake)
- Bake that triggers crosslinking catalysed by this acid and makes the exposed areas insoluble. An insufficient PEB gives patterns that lift off during development.
- Development
- Dissolution of the un-crosslinked areas. Published times are approximate: the dissolution rate depends on agitation and temperature.
- Dose
- Energy per unit area, in mJ/cm². A property of the resist and thickness, independent of equipment.
- Irradiance
- Power per unit area delivered by the lamp, in mW/cm². A property of the equipment, to be measured with a radiometer.
Validity domainNot covered: the SU-8 TF 6000 and SU-8 3000CF DFR series, for which only the safety data sheet is available; SU-8 XFT 75 and 100, whose reading uncertainty is given in absolute micrometres, which the model does not handle.
Not covered: the SU-8 TF 6000 and SU-8 3000CF DFR series, for which only the safety data sheet is available; SU-8 XFT 75 and 100, whose reading uncertainty is given in absolute micrometres, which the model does not handle. The datasheets present their values as a starting point, to be optimised with an exposure matrix.
Choice of productThe 25–40 µm range is covered by two different datasheets: thin SU-8 2000 formulations (2000.5 to 2015) and thick formulations (2025 to 2075).
The 25–40 µm range is covered by two different datasheets: thin SU-8 2000 formulations (2000.5 to 2015) and thick formulations (2025 to 2075). Their soft bakes at 95 °C last 4 to 5 minutes for the former, 5 to 6 minutes for the latter. The product sets the datasheet and the recipe; it is asked for before the thickness. The classic series tables are also indexed by product. Where two datasheets overlap, the value shown is that of the chosen product’s datasheet, with its revision.