https://nanyte.com/photoresists/hmds · last updated 2026-09-28
- Manufacturer
- MicroChemicals GmbH (application-note publisher; HMDS itself is a commodity chemical sold by many suppliers, not a single manufacturer's proprietary product)
- Chemistry
- Ancillary (not photoimageable)
- Exposure dose
- Not exposed — a vapour-phase prime
- Substrate
- Primer
- Resist
Spin coating
The manufacturer does not publish a spin curve for HMDS (Hexamethyldisilazane).
- HMDS must not be spin-coated as a liquid — the only correct application is vapor-phase, via a nitrogen-carrier 'bubbler' held at room temperature feeding a heated (75-120°C) substrate, where HMDS bonds chemically as a monolayer ("Correct Use of HMDS", Fig. 47-48, p.2).
- If liquid HMDS is spun on instead, the resulting layer is only physically (not chemically) bound; ammonia released from it during the resist softbake diffuses into and cross-links the resist near the substrate, suppressing development and degrading resolution and profile ("Incorrect Application of HMDS", Fig. 49, p.2-3).
- For the same reason, the note advises against applying HMDS in the same spin coater used for resist coating (slow HMDS evaporation can contaminate later coating runs); if spin-application is unavoidable, it recommends a 100-120°C water-desorption bake beforehand, a 100-120°C thermal-activation bake afterward, and strict spatial separation from resist-coating equipment (p.3).
- There is no spin-speed-vs-thickness relationship for a vapor-deposited chemisorbed monolayer, so HMDS has no spin curve to publish.
- Adhesion
- HMDS is itself the adhesion promoter: it is applied to the bare substrate before the imaging resist is coated.
Prime
- Prime
- 75–120 °C
- Notes
- Vapor-phase application: HMDS vapor is carried by dry nitrogen from a room-temperature bubbler onto a substrate held at 75-120°C, where it bonds chemically as a monolayer. That 75-120°C is the substrate temperature during the prime itself, not a bake of a coated film — HMDS goes on bare substrate, before any resist. No dwell time is published. If HMDS can only be spin-coated — documented, but explicitly the wrong way to apply it — bake at 100-120°C beforehand to desorb water, then again at 100-120°C to activate after application. Independently of which adhesion promoter is used, a substrate dehydration bake of roughly 120°C for a few minutes (above 140°C on oxidized surfaces, for maximum adhesion) precedes the prime.
SOURCE: Correct Use of HMDS / Incorrect Application of HMDS, p.2-3; Adsorbed Water, p.1
Exposure
HMDS (Hexamethyldisilazane) is not photoimageable. It is a vapour-phase prime applied to the bare substrate before the imaging resist is coated; nothing in this step is exposed or developed.
Development
Not published for this resist: Developer, Dilution, Time, Method, Rinse — characterize on-tool.
Hard bake, etch & strip
Not published for this resist: Hard bake, Descum, Etch resistance, Stripper, Storage — characterize on-tool.
Where it's used
Practical notes from the datasheet
HMDS (hexamethyldisilazane) is a vapor-phase adhesion promoter, not a photoimageable resist: it converts a hydrophilic, OH-terminated oxide surface (native or thermal SiO2, quartz, glass, most base metals) into a hydrophobic, resist-wetting surface by chemically bonding a monolayer of non-polar trimethylsilyl groups, releasing ammonia as a byproduct (p.2). The only correct application is vapor-phase, via a nitrogen-carrier bubbler at room temperature feeding a substrate heated to 75-120°C; spin-coating liquid HMDS instead yields only a physically bound layer whose ammonia, released during the subsequent resist softbake, diffuses into and cross-links the resist near the substrate, suppressing development and degrading resolution — a failure mode this note documents explicitly as 'Incorrect Application' (p.2-3). On noble metals without native oxide (gold, platinum), HMDS and other organic adhesion promoters show little to no adhesion benefit because they cannot chemically bond to the surface; base metals such as Al, Cr and especially Ti already adhere well without a promoter (p.3-4). The process here follows a general MicroChemicals application-note chapter ('01 Chapter — Basics of Microstructuring: Substrate Preparation') rather than a single manufacturer's technical datasheet for a specific HMDS product or grade — HMDS itself is a commodity chemical sold by many suppliers.
Troubleshooting
Common failure modes for HMDS (Hexamethyldisilazane), 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.
After HMDS priming the resist won't clear to the substrate and leaves scum in the openings. Is the HMDS to blame?
It can be, if it was spun on as a liquid. Spin-coated HMDS leaves a comparatively thick, only physically bound film. During the resist softbake that excess releases ammonia, which diffuses into the resist near the substrate and cross-links it, so the development rate drops and through-development can become impossible. The cure is vapour priming: HMDS carried by dry nitrogen from a room-temperature bubbler onto a substrate held at 75-120 °C, where it bonds chemically as a monolayer.
SOURCE: MicroChemicals application note
MicroChemicals application note "Substrate Preparation: Cleaning and Adhesion Promotion" (Basics of Microstructuring, no revision date printed) — Correct Use of HMDS; Incorrect Application of HMDS, Figs. 47-49, p.2-3. MicroChemicals application note "Photolithography Trouble Shooter" — HMDS Applied Correctly?
Development got slower on wafers we never primed. Can HMDS still be the cause?
Yes, if HMDS is ever dispensed in the same spin coater used for resist. HMDS evaporates slowly, so its vapour lingers in the bowl and penetrates resist films coated there long afterwards; the softbake then partially cross-links them, lowering the development rate and costing profile and resolution. Never apply HMDS in a coater that also coats resist. Other airborne cleanroom chemicals such as ammonia can do the same thing via the gas phase.
SOURCE: MicroChemicals application note
MicroChemicals application note "Substrate Preparation: Cleaning and Adhesion Promotion" (Basics of Microstructuring, no revision date printed) — Incorrect Application of HMDS, p.2-3. MicroChemicals application note "Photolithography Trouble Shooter" — HMDS Applied Correctly?; Contamination from Air
We have no vapour-prime oven and can only spin HMDS. How do I limit the damage?
Do three things. Bake the substrate at 100-120 °C before applying HMDS to drive off adsorbed water. Bake again at 100-120 °C afterwards to activate it, i.e. to bond it chemically to the surface. And keep HMDS application strictly separated from resist coating so the fumes cannot reach the resist. Vapour application onto a heated substrate remains the correct method.
SOURCE: MicroChemicals application note
MicroChemicals application note "Substrate Preparation: Cleaning and Adhesion Promotion" (Basics of Microstructuring, no revision date printed) — Correct Use of HMDS, p.2; Incorrect Application of HMDS, p.3
HMDS does nothing for resist adhesion on gold. What should I use instead?
On noble metals with no native oxide, such as gold and platinum, HMDS and other organic adhesion promoters show little or no improvement because they cannot bond chemically to the surface. A 10-20 nm titanium or chromium adhesion layer promises a clear improvement, and such a thin metal film can be etched away after development or after resist removal. Base metals such as Al, Cr and especially Ti generally give very good resist adhesion without any promoter.
SOURCE: MicroChemicals application note
MicroChemicals application note "Substrate Preparation: Cleaning and Adhesion Promotion" (Basics of Microstructuring, no revision date printed) — Special Problems Regarding Resist Adhesion: On Metals, p.3-4. MicroChemicals application note "Photolithography Trouble Shooter" — Inferior Resist Adhesion: On Noble Metals?
Resist dewets or peels on an HMDS-primed wafer anyway. What else should I check?
The surface underneath the prime. Adsorbed water needs a bake of about 120 °C for a few minutes, or above 140 °C on oxidised surfaces for maximum adhesion, and coating should follow as soon as the wafer is back at room temperature because water re-adsorbs quickly. After an HF dip, incompletely removed oxide leaves a highly hydrophilic surface that priming will not rescue; only complete oxide etching or a bake at about 700 °C restores it. Allresist also reports adhesion falling off substantially above 60 % relative humidity.
SOURCE: MicroChemicals application note
MicroChemicals application note "Substrate Preparation: Cleaning and Adhesion Promotion" (Basics of Microstructuring, no revision date printed) — Adsorbed Water, p.1; After HF Etching, p.3. MicroChemicals application note "Photolithography Trouble Shooter" — Uncoated Substrate Areas after Resist Coating?; Inferior Resist Adhesion. Allresist "FAQs concerning photoresists" (revised 03.03.2016) — Q4 and Q5
Sources & disclaimer
- MicroChemicals GmbH (application-note publisher; HMDS itself is a commodity chemical sold by many suppliers, not a single manufacturer's proprietary product) — HMDS (Hexamethyldisilazane) datasheet · accessed 2026-07-10
- Kawai and Kawakami. Characterization of SiO2 Surface Treated by HMDS Vapor and O2 Plasma with AFM Tip. Journal of Photopolymer Science and Technology (2003). doi:10.2494/photopolymer.16.665Silicon oxide substrates are silane-coupling treated with HMDS and separately exposed to oxygen plasma, then probed with an AFM tip to characterize the resulting surface. Measures the surface change that HMDS priming is performed to produce, on the oxide that resist is subsequently coated onto.
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.
