Most spin-coating recipes are written as a final speed and a time, but every recipe also contains a ramp: the seconds in which the substrate accelerates to the speed that sets the film. During those seconds the liquid spreads, reaches the edge and starts to thin, and whether it covers the whole surface is decided largely in this stage. This guide explains what spin coating acceleration does and does not control, how to read and plan the ramp, why recipes are split into spread, thinning and drying steps, and which acceleration and program specifications are published on the Levon™ spin coater pages.
In one paragraph: the ramp changes how liquid spreads during spin-up and can affect coverage, drying history and final thickness. Photoresist guidance for smooth substrates reports that a steep ramp can suppress uncoated areas, while textured substrates may benefit from a short low-speed spread step.6 These are process-specific observations, not universal acceleration rules. Keep the ramp as a fixed, recorded condition of each recipe and spin curve, and check both coverage and thickness when it changes.
1. The short answer: what the ramp does
Spin coating acceleration controls how quickly the substrate gets from one speed to the next. A ramp of one or two seconds looks negligible beside a spin step of thirty, but spin coating is usually described in four overlapping stages: deposition, spin-up, spin-off and evaporation.8 The ramp is the spin-up stage, where coverage is won or lost. The practical summary is:
- The ramp affects coverage. It changes how the dispensed liquid spreads and how the front reaches the edge. For smooth substrates, one photoresist application note reports that a high acceleration (> 1000 rpm/s) towards the final speed suppresses uncoated areas, and that decreasing the acceleration increases the risk of what it calls resist cracks: uncoated gaps in the film.6
- Final speed is one thickness control. The ideal evaporation-controlled model predicts an inverse-square-root dependence on final speed.2 That result does not establish how strongly acceleration affects the final dry thickness of a particular ink.
- Steps separate the jobs: spread, thin, and dry or remove the edge bead, each tuned on its own.
- Acceleration is part of the recipe. “3,000 rpm for 30 s” is incomplete without the ramp of every step.
The stages of spin coating and the models behind film thickness are explained in Spin Coating Explained: Stages, Physics and Film Thickness. If you are still deciding whether spin coating is the right method for your film, start with the guide to choosing spin, dip, slot-die or blade coating.
2. Spin-up: spreading, fingering and coverage
At the end of a static dispense, the liquid sits as a puddle near the centre of a stationary substrate. When the substrate accelerates, the liquid is pulled outward. As an order-of-magnitude estimate, the viscous time scale h2/ν is about 0.1 s for a 1 mm puddle of a liquid ten times as viscous as water (illustrative values), so the liquid soon co-rotates with the substrate and flows outward, driven by the centrifugal force and resisted by viscosity.
The advancing front can break into fingers
While the liquid spreads over a dry surface, it has an advancing edge, the contact line. Experiments on spinning drops show that the spreading is axisymmetric at first and then, beyond a critical radius, the front becomes unstable and develops fingers, also called rivulets.34 In one study with an impulsive spin-up, the most unstable wavelength was independent of drop size and rotation speed over the range investigated, while changing the wetting properties of the liquid significantly changed the critical radius and, with it, the number of fingers; the instability was linked to a capillary ridge near the edge of the drop.4 Fingering is therefore not only a question of speed: wetting, set by the liquid and the substrate surface, is part of it.
Why the ramp affects coverage
If the fingers reach the edge before the liquid between them has filled in, the excess flows out along the fingers and the regions between them can stay dry: uncoated sectors or radial streaks that start at the edge of the original puddle. The ramp changes how the spreading proceeds. For photoresists on smooth substrates, a steep ramp to the final speed is reported to suppress these uncoated parts, with gentler ramps increasing the risk.6 As qualitative reasoning rather than a model result: a steep ramp drives the puddle outward while it is still thick and connected, whereas a slow ramp gives the front more time to advance unevenly.
The same application note lists other causes of uncoated areas that no ramp setting will fix: too little resist for the substrate size, film thickness, viscosity and spin profile, and resist dispensed off centre.6 If coverage fails, check the dispense before you change the ramp; a separate article, Static vs Dynamic Dispense in Spin Coating: Volume, Coverage and Material Waste, will cover dispense method and volume.
Textured substrates need a spread step
On a textured or patterned surface, the liquid must flow into features before it is thinned. Here the guidance changes: textured substrates may need a two-step profile in which a comparatively large volume is first distributed at a low speed (for example 1000 rpm) for a few seconds before the ramp to the final speed.6 The note warns that the low-speed stage should not last longer than a few seconds, to avoid premature drying.6 These numbers come from photoresist practice; for other inks they are starting points to test, not settings to copy.
3. Does acceleration change film thickness?
Acceleration can affect film thickness, but the size and direction of the effect depend on the recipe. The calculation below isolates one contribution: flow-driven thinning during a ramp. It does not predict the final dry film or rank acceleration against other process variables.
The ramp is spinning time too
For a uniform Newtonian film without evaporation, the Emslie–Bonner–Peck thinning equation is dh/dt = −(2ρω2/3η)h3.1 Here we make an illustrative quasi-steady extension by substituting a prescribed speed history ω(t). This assumes that the liquid adjusts rapidly enough to the changing substrate speed for the same instantaneous flow relation to apply. It neglects transient spin-up effects and does not model the advancing contact line, evaporation or solidification.
For a linear ramp from rest to ωf in time T: ∫ω2 dt = ωf2·T/3.
ω = 2πn/60 in rad/s, where n is speed in rpm; use h in m, η in Pa·s, ρ in kg/m3 and time in s.
Flow-only model extension under the stated assumptions; not a result from the cited paper and not a dry-film prediction.
In this picture, a linear ramp from rest thins the liquid as much as one third of the ramp time spent at the final speed. The table works through assumed values: a liquid of viscosity 10 mPa·s and density 1,000 kg/m3, a 50 µm starting layer and a final speed of 3,000 rpm.
| Ramp to 3,000 rpm | Ramp time | Wet thickness after 10 s at speed | Wet thickness after 2 s at speed |
|---|---|---|---|
| Instant (idealised) | 0 s | 2.75 µm | 6.12 µm |
| 10,000 rpm/s | 0.3 s | 2.74 µm | — |
| 1,000 rpm/s | 3 s | 2.62 µm | 5.01 µm |
| 100 rpm/s | 30 s | 1.95 µm | — |
Illustrative flow-only calculation with assumed values and no evaporation. Each row includes the stated ramp followed by an additional 10 s or 2 s at the final speed; total elapsed times therefore differ. The instant-ramp row is a mathematical reference, not an executable setting. These are wet-film calculations, not measured or predicted final dry-film thicknesses.
Three lessons follow from the arithmetic, all within the model’s assumptions:
- A short ramp adds a small contribution to the integrated speed-squared history when the final-speed hold is long. In this example, adding a 3 s ramp before the same 10 s hold changes the calculated wet thickness from 2.75 to 2.62 µm.
- The relative contribution is larger with a short hold. Before the same 2 s hold, that ramp changes the calculated wet thickness from 6.12 to 5.01 µm.
- These are not equal-total-time comparisons. The 30 s ramp plus 10 s hold lasts 40 s. At a fixed total duration, a slower ramp leaves less time at the final speed and can reverse the trend in this flow-only comparison.
When the film sets matters
Real films also lose solvent. In Meyerhofer’s model, outflow dominates early, evaporation takes over as the film thins, and the final thickness is set around the time the two become comparable, which is why the final speed and the ink, not the starting volume, control the result.2 A model coupling flow, solvent diffusion and evaporation also shows how a solvent-depleted surface layer can form.5 The reasoning for the ramp, not a measured rule, is: if the film sets long after the final speed is reached, the ramp history is largely forgotten; if the film sets during or soon after the ramp, as with volatile solvents, very short steps or thick resists spun for only a few seconds, the ramp becomes part of what sets the thickness.
Photoresist practice is consistent with this. When spin-coating is stopped while resist is still being spun off, the thickness depends on the spin time, so the same note recommends, for reproducibility, a resist that reaches the desired thickness with a conventional profile (2000–4000 rpm for 20–30 s).6 For thick films it describes steep ramps, a short plateau whose time is adjusted by trials, and a rapid deceleration;6 there the ramps are part of the thickness recipe.
4. Anatomy of a multi-step recipe
A multi-step recipe splits spin coating into steps that each do one job. Not every recipe needs every step; a single ramp to a single speed is often enough on a small, smooth, well-wetted substrate. The table shows the usual building blocks.
| Step | Purpose | Typical settings to decide | What goes wrong if it is wrong |
|---|---|---|---|
| Dispense | Place the liquid: at rest (static) or on a rotating substrate (dynamic) | Volume, position, dispense speed, delay before the next step | Uncoated areas from too little or off-centre liquid; bubbles and comets6 |
| Spread | Distribute the liquid over the whole surface, or into textures, before thinning | Low speed, ramp, a few seconds | Too long: premature drying; too short or skipped on textured surfaces: unfilled features6 |
| Thin (cast) | Reach the final speed and thin the film until it sets | Final speed, ramp from the spread step, time | Wrong thickness; time-dependent thickness if the step ends while liquid is still being spun off6 |
| Dry or edge | Continue drying, or throw off the edge bead | Speed, ramp, time; solvent edge rinse where fitted | A thick or wet edge bead; a film that flows during handling |
| Stop | Return to rest | Deceleration | Disturbing a film that is not yet set |
Spread step
The spread step is a low-speed step with a short time, used when a single ramp does not cover the surface or when the substrate is textured (section 2). Set the spread time to achieve the required coverage without allowing premature drying or viscosity increase to prevent further spreading. Evaporation can already occur during this step. Keep its settings identical across a spin curve. In one study of photoresist consumption, the pre-spin speed was optimized to reduce consumption, which suggests the early part of the program also affects how much material a coating uses.7
Thinning step
This is the step whose speed you read off a spin curve. The ramp into it is the coverage control discussed above; its time is usually chosen long enough for the film to stop thinning appreciably, so that thickness depends less on the exact stopping point; check this with a short hold-time series for your ink. A separate article, Spin Speed vs Film Thickness: How to Build and Use a Spin Curve, will cover building a spin curve.
Drying and edge steps
Photoresist practice uses the program against the edge bead: an abrupt speed increase at the end of coating, while the bead is still liquid enough to be thrown off, or solvent edge removal at about 500 rpm on circular substrates.6 Edge hardware options are listed in section 7.
5. A published two-step example
One Levon™ product page publishes a complete two-step acceptance recipe that states the ramp of each step. It applies only to the configurations named.
| Step | Target speed | Time as published | Acceleration | Ramp time (arithmetic) |
|---|---|---|---|---|
| 1. Spread | 500 rpm | 5 s | 500 rpm/s | (500 − 0) / 500 = 1 s |
| 2. Thin | 2,500 rpm | 35 s | 2,000 rpm/s | (2,500 − 500) / 2,000 = 1 s |
Recipe published on the AC200 product page for the custom AC200-EBR-CTM and AC200-COT-CTM configurations, using the specific 20 cP photoresist identified in that protocol and a prebake at 100 °C for 100 s. The 1 s ramp times are nominal arithmetic from the settings, not measured ramp traces. Confirm whether the published step times include or exclude acceleration before reproducing the sequence.
The linked acceptance protocol reports approximately 1.3 µm under its specified photoresist and prebake conditions. This article reproduces the program sequence, not a general film-thickness or uniformity guarantee.
The example shows a low-speed spread step followed by a higher-speed casting step, each with a specified acceleration. It does not include an acceleration comparison or a hold-time series, so it does not establish that 35 s is a thickness-independent endpoint.
What it does not show:
- It is an acceptance recipe for two custom configurations quoted on request, not for the standard AC200 configurations or other models.
- It covers one photoresist, prebake and measurement method; it is not a recipe for other inks or a thickness guarantee.
- It does not isolate the effect of acceleration: no comparison with other ramps is published.
6. How to choose and test the ramp
There is no universal correct acceleration; it depends on the ink, the substrate, the dispense and the coater. The method below finds one for your process.
Work out the ramp times first
As illustrative arithmetic, assuming the coater follows the set acceleration, a ramp from rest to 3,000 rpm takes 30 s at 100 rpm/s, 3 s at 1,000 rpm/s and 0.3 s at 10,000 rpm/s. Check the controller’s timing convention. A 30 s ramp followed by a 10 s hold lasts 40 s; a 10 s segment whose timer includes the ramp may end before reaching the target speed.
A ramp ladder
- Fix the rest of the recipe. Same ink batch, dispense volume and position, spread step, final speed, time at speed, deceleration, substrate preparation, atmosphere and bake.
- Choose three to five ramps spanning your coater’s range on a roughly logarithmic scale, for example 500, 1,000, 3,000 and 10,000 rpm/s if the coater supports them (illustrative values).
- Coat at least two substrates per ramp, in mixed order, so that drift in the ink or room is not mistaken for an effect of the ramp.
- Inspect coverage first: uncoated sectors, radial streaks starting near the centre, a thick or irregular edge.
- Then measure thickness at the centre and towards the edge, with the measurement method you will use in production.
- Choose the ramp that gives full coverage with margin, not the one at the edge of failure, and check that thickness at that ramp is stable from run to run.
If no ramp gives full coverage, the cause is probably upstream: too little liquid, an off-centre dispense, poor wetting, bubbles or particles.
Keep the ramp fixed in a spin curve
A spin curve compares thickness at several final speeds. If the acceleration is held fixed, the ramp time grows with the final speed (1 s to 1,000 rpm and 4 s to 4,000 rpm at 1,000 rpm/s, illustratively). If the ramp time is held fixed instead, the acceleration changes with speed. Either is defensible; choose one, record it and keep using it. Check also whether your coater’s step time includes the ramp.
Watch the substrate as well as the film
A steep ramp asks more of the vacuum hold. Check that small, thin or warped substrates are seated and held, and follow the limits of the chuck and the instrument.
7. Acceleration and program specifications
Ramp control depends on the acceleration range, the stored programs and steps, and the published time limits and timing resolution. The table is a selected comparison of the Levon™ spin coater pages, not a complete catalogue. Values belong to the model or configuration named and are not interchangeable.
| Model | Acceleration | Speed range | Programs | Published time limit / resolution |
|---|---|---|---|---|
| EZ4-S / EZ4-S-PP | 100–9,999 rpm/s | 100–9,999 rpm | 5 programs × 5 steps | Up to 3,000 s, 1 s resolution |
| EZ6-S | 100–30,000 rpm/s | 100–12,000 rpm | 10 programs × 10 steps | Up to 3,000 s, 0.1 s resolution |
| AC100-S | Up to 30,000 rpm/s | Up to 12,000 rpm (no load) | 100 recipes, up to 100 steps each | Up to 3,000 s, 0.1 s resolution |
| AC200-S-PP | 20–50,000 rpm/s | 20–10,000 rpm | 100 recipes × 100 steps (expandable) | Up to 3,000 s, 0.1 s resolution |
| AC200-SE | 20–50,000 rpm/s (no load) | 20–10,000 rpm (no load) | 100 groups of 100 steps | Up to 3,000 s, 0.1 s resolution |
| AC200-PE | Max. 50,000 rpm/s (no load) | Max. 10,000 rpm (no load) | 100 recipes × 100 steps | Up to 3,000 s, 0.1 s resolution |
| AC200-SE-PLUS | Max. 10,000 rpm/s (no load) | Max. 8,000 rpm (no load) | Not stated on the page | Up to 3,000 s, 0.1 s resolution |
| AC300-S-CTM | Up to 10,000 rpm/s (no load) | Up to 5,000 rpm (no load) | Not stated on the page | Up to 3,000 s, 0.1 s resolution |
| AC300-S-EN | Up to 1,000 rpm/s (no load) | Up to 5,000 rpm (no load) | Not stated on the page | Up to 3,000 s, 0.1 s resolution |
| IC5000-S | 100–5,000 rpm/s | 100–5,000 rpm | Up to 100 recipes of 100-step programs (expandable) | Up to 3,000 s, 0.1 s resolution |
| IC8000-S | 100–5,000 rpm/s | 100–8,000 rpm | Up to 100 recipes of 100-step programs (expandable) | Up to 3,000 s, 0.1 s resolution |
Values from the Levon™ product pages as listed on 22 October 2026, in the pages’ own wording. “No load” values are hardware limits, not process results; the ramp actually achieved with a given chuck and substrate can differ. Published maximum spin or rotation times are not automatically per-step limits. Where the source does not define the scope, confirm whether the limit applies to one step or the complete program, and whether acceleration is included. The full range is on the Levon™ spin coaters page.
How the specifications map to a recipe
- Acceleration range. The upper end matters for steep coverage ramps, the lower end for gentle spread steps. The AC300-S-EN page lists a maximum of 1,000 rpm/s (no load), compared with 10,000 rpm/s (no load) on the AC300-S-CTM page.
- Programs and steps. A spread–thin–dry recipe uses three steps or more. Larger program memories matter when several users or inks share one coater.
- Timing resolution. The listed timing increments are 0.1 s for most models and 1 s for the EZ4-S. Confirm which programmed interval the increment applies to and when its timer starts; timing resolution alone does not establish ramp tracking or dispensing-trigger accuracy.
- Edge steps in hardware. The AC200-S-PP page lists integrated backside cleaning and edge debonding; on the AC200-SE, AC200-PE, AC200-SE-PLUS, AC300-S-CTM and AC300-S-EN pages, backside rinse (BSR) and edge bead removal (EBR) can be added. Where EBR or BSR hardware is not listed, confirm availability and the supplied configuration. A programmable speed change alone does not establish a solvent edge-rinse or backside-rinse capability.
- Heated spin tables. The IC5000-S (spin table RT to 150 °C) and IC8000-S (RT to 200 °C) add temperature as a variable; ramps and spread-step times found at one table temperature should be rechecked at another.
Before you request a quote
- Your speed window: the final speeds you need, the lowest spread speed, the steepest and gentlest ramps your recipes use, and the longest and shortest step times.
- Your substrates: material, shape, size and thickness, whether they are textured, and the chucks or adapters you need.
- Your environment: air or glovebox, exhaust, and whether a heated spin table is needed.
- Your recipe control: number of programs and steps, timing resolution, and any edge bead or backside rinse steps.
- Your measurement: how you will check coverage and thickness after a ramp change.
Send these through the contact form or the Get Quote request on the product page.
8. Keep Exploring the Spin Coating Knowledge Hub
This article is part of the spin coating section of the ACS Material thin-film library. The other articles in the series are listed below.
More in this series
- Pillar guide: Solution-Processed Thin Films: Choosing Spin, Dip, Slot-Die or Blade Coating
- #01: Spin Coating Explained: Stages, Physics and Film Thickness
- #02: Spin Speed vs Film Thickness: How to Build and Use a Spin Curve
- #04: Static vs Dynamic Dispense in Spin Coating: Volume, Coverage and Material Waste
- #05: Spin Coating Defects: Comets, Striations, Pinholes and Edge Beads Diagnosed
- #06: Spin Coating Small, Square and Fragment Substrates: Chucks, Vacuum and Corners
- #07: Heated Spin Coating: How Substrate Temperature Changes Evaporation and Thickness
- #08: How to Choose a Lab Spin Coater: Speed, Acceleration, Substrate Size and Glovebox Fit
Start here
- Spin Coating Explained: Stages, Physics and Film Thickness — the four stages of spin coating, the thickness models and the main defects.
- Solution-Processed Thin Films: Choosing Spin, Dip, Slot-Die or Blade Coating — when spin coating is the right method, with the coating method selector.
Spin coating equipment
- Levon™ spin coaters — compact, programmable, servo and heated spin coaters.
- EZ4-S compact spin coater, EZ6-S benchtop spin coater and AC100-S programmable spin coater — programmable ramps and multi-step recipes on small pieces and wafers.
- AC200 programmable servo spin coater and AC300-S-CTM servo spin coater — servo-driven coaters for larger wafers.
- IC5000-S and IC8000-S heated spin coaters — programmable recipes with a heated spin table.
9. FAQ
What does acceleration do in spin coating?
It controls how quickly the substrate reaches each new speed and changes the spreading and drying history during spin-up. It can affect coverage and final thickness. The relative importance of acceleration, final speed, time and formulation must be established for the process being used.
Does spin coating acceleration change film thickness?
It can. Changing the ramp changes the time spent at intermediate speeds and can change spreading, evaporation and the state of the film when the final speed is reached. Measure the dry-film response with the other recipe conditions controlled; the flow-only illustration in section 3 is not a dry-film prediction.
What is a good acceleration for spin coating?
There is no universal value. For photoresists on smooth substrates, one application note reports that more than 1000 rpm/s towards the final speed suppresses uncoated areas. For other inks and substrates, test a few ramps with everything else fixed and choose one with a margin for coverage.
How long should a low-speed spread step be?
Long enough to cover the substrate and fill any texture, short enough to avoid premature drying that interferes with spreading. For photoresists on textured substrates, the cited note suggests a few seconds at low speed (for example 1000 rpm); test other inks.