Photoresist Spin Coating: From Viscosity to Target Thickness and Uniformity

Every lithography step after coating assumes the resist has the right thickness everywhere. This guide explains how spin speed, viscosity and evaporation set that thickness, how to choose a resist and a program for a target, what causes edge beads, comets and bubbles, and how to measure and fit your own spin curve.

In one paragraph: Use the resist supplier’s thickness and processing window as a starting point, then measure your own spin curve. The inverse-square-root speed relationship is an approximation, not a universal exponent [1][3]. Qualify dispense, ramp, time and bowl conditions together, inspect defects, and compare thickness after a consistent softbake and cooling procedure.

1.  What Photoresist Spin Coating Does

Photoresist is the light-sensitive film that carries a pattern from a mask into a substrate, and spin coating is by far the most common way to apply it [1]. A few millilitres of resist are dispensed on a substrate that is then spun at, typically, several thousand rpm. Centrifugal force spreads the resist into a thin film, the excess is thrown off the edge, and solvent evaporates until the film is stable enough to handle [1].

The step is short: spin coating typically takes 10–20 s, and the films are smooth, reproducible and very uniform, at least on round wafers [1]. It is also wasteful of material: most of the dispensed resist is spun off, and typically only a few per cent stays on the substrate [1]. Everything that happens later in lithography, from exposure dose to developed sidewalls, assumes that the resist has the intended thickness and that the thickness is the same across the substrate [8].

This guide follows the questions in the order a process engineer meets them: which physics sets the thickness, how to pick a resist and a speed for a target, which program and environment details matter, how to keep the film uniform and free of defects, and how to measure and fit the result. General spin coating background is in Spin Coating Explained: Stages, Physics and Film Thickness.

2.  What Sets Resist Thickness

Two classic models describe the thinning of a liquid on a spinning disc. In the Emslie–Bonner–Peck model, a Newtonian liquid with no evaporation keeps thinning as long as the disc turns; the model has no final thickness [2]. Meyerhofer added solvent evaporation: outflow slows as the film thins, evaporation carries on, and comparable thinning rates mark the crossover from flow-dominated to evaporation-dominated thinning [3]. His result for the final thickness is [3]:

Later models, such as that of Flack and co-workers for polymer resists, track how the solvent concentration and the concentration-dependent viscosity change during spinning [5]. Experimental work on resist solutions, such as Lai’s study of electron resists, relates film thickness to the solution and the spin speed [4]. For practical photoresist work, the useful conclusions are:

• Speed. For the conventional photoresist processes described in the cited note, an inverse-square-root relationship approximates thickness versus spin speed [1]. Doubling the speed reduces the thickness by a factor of about 1.4 [1].
• Viscosity and solids. More viscous, more concentrated resists give thicker films at the same speed [3][5]. This is why each resist covers its own thickness range.
• Evaporation. Anything that changes how fast solvent leaves the film, such as temperature, the solvent vapour around the substrate or the air flow above it, also changes the final thickness [1][3].

The film is not dry when spinning stops. MicroChemicals gives typical residual solvent contents after spin coating of about 10 % for thin resists to 35 % for thick ones, falling to about 3–5 % only during the softbake, with more solvent remaining near the substrate in thick films [1]. The softbake that follows is covered in Soft Bake, Post-Bake and Annealing on a Hot Plate: What Each Step Does.

3.  Choosing a Resist and a Speed for a Target

A resist has a process-dependent thickness window; use its measured spin curve rather than assuming an exact inverse-square-root response. Pushing a resist far outside its range by speed alone tends to move the process to where it is less reproducible.

Arithmetic from the approximate reciprocal-square-root relation; your own resist and coater may deviate, so measure a spin curve.

Choose the resist for the target, then the speed

MicroChemicals recommends using a resist that reaches the desired thickness with conventional spin profiles, about 2,000–4,000 rpm for 20–30 s, for reasons of reproducibility [1]. MicroChemicals also notes that for many of the resists it supplies, the product designation encodes the thickness reached at 4,000 rpm (3,000 rpm for some) [1]. For very thick films, the same source names thick positive and negative resists that reach up to about 10 µm, and special thick resists reaching about 50–100 µm [1]; these examples do not establish a universal maximum thickness for spin coating.

In practice:

1. Pick a resist whose data sheet places your target inside its standard speed window.
2. Run a spin curve on your own coater and substrates to find the speed for the target (see Section 8).
3. Stay within a qualified speed window. If the target falls outside it, evaluate additional measurements or a different resist grade rather than extrapolating.

Diluting a resist

Dilution with the recommended thinner lowers viscosity and solids and therefore thickness [3][5]. It has limits. MicroChemicals notes that too strong a dilution, like an expired resist or too-warm storage, promotes particle formation, and that highly diluted resists may start to gel, leaving millimetre-sized clots of the same colour as the resist on the film [1]. Dilute only with the solvent the resist supplier recommends, in the range they recommend, and qualify the diluted resist with its own spin curve.

4.  Spin Time and When the Film Sets

Spin time matters most when the run stops before the film has set. If spinning ends while resist is still being thrown off, final thickness also depends on the spin time; the more viscous the resist and the slower the speed, the longer this window lasts [1]. MicroChemicals gives examples: a thinner resist at 3,000–4,000 rpm reaches its final thickness of about 3 µm after only a few seconds, while a thick resist needs about 20 s to reach about 7 µm [1].

• At the slowest speed you plan to use, check that the final-speed time gives repeatable thickness after softbake and that small changes in time do not change the result.
• If a thicker film is needed than the full spin gives, change the resist rather than shortening the spin. A short spin makes thickness very time-critical [1].
• Keep the time constant when you build a spin curve, so that the curve shows the effect of speed only.

5.  Writing the Spin Program

A spin program has more steps than its final speed. Each one affects coverage and uniformity:

Summarised from MicroChemicals [1]; the right values depend on the resist, substrate and coater.

The optimum dispense volume depends on the substrate size, film thickness, resist viscosity and spin profile [1], so treat it as a recipe parameter, not a constant. For details of static and dynamic dispense, see Static vs Dynamic Dispense in Spin Coating: Volume, Coverage and Material Waste; for ramps and multi-step recipes, see Spin Coating Acceleration and Multi-Step Recipes: What the Ramp Actually Does.

Textured substrates

On textured substrates, resist can pull back from the edges of features and collect in recesses, so thickness varies with the topography [1]. A two-stage profile helps: distribute a relatively large resist volume at low speed for a few seconds, then ramp to the final speed [1]. If the result is still not good enough, MicroChemicals suggests considering an alternative such as dip or spray coating [1].

6.  Coater and Room Conditions

The same program can give different thicknesses on different days because evaporation depends on the surroundings:

• Temperature. Warmer resist evaporates faster but is less viscous. Their effects can oppose each other [1], but the net thickness change and acceptable temperature variation must be established for the actual resist. For a heated substrate, see Heated Spin Coating: How Substrate Temperature Changes Evaporation and Thickness.
• Solvent vapour in the bowl. The more saturated the air above the substrate, the slower the evaporation, the longer the spin-off and the thinner the final film [1]. After the coater has been cleaned or has stood unused, coating a few dummy wafers first can bring the thickness back to a steady value [1].
• Air movement. A cover rotating with the substrate reduces air movement and evaporation and can reduce the thickness by a factor of two or more [1]. Cover geometry can affect drying [1]; use only a manufacturer-approved cover and exhaust configuration, and do not obstruct required openings or defeat interlocks.
• Bowl geometry. Inner walls should deflect spun-off resist downward, not back onto the substrate as droplets [1].
• Solvent choice. The cited note warns that resists with low-boiling solvents dry too quickly during spinning to form a uniform film [1]; use a supplier-approved formulation rather than boiling point alone as a selection rule.

Write down the room temperature and humidity with each run, and keep the coater state (clean, warmed up with dummies or not) part of the recipe.

7.  Uniformity and Coating Defects

Edge bead

An edge bead is a raised rim of resist along the substrate edge. It forms especially with thick films and with square substrates, where faster drying at the edges and corners stops the resist from spinning off [1]. In contact exposure it causes two problems: the solvent-rich bead can stick to the mask, and it holds the mask away from the rest of the film, which often leads to poor resolution, wrong dimensions or shallow sidewalls [1]. The gap effect follows from the optics of proximity printing [8], covered in Mask Aligner Exposure: Contact vs Proximity and the Resolution Limit.

The cited note discusses the following options. Use only an approved holder, dispenser and operating program; never reach into a rotating bowl. Examples include [1]:

• for round substrates, dynamic edge bead removal at about 500 rpm with an edge bead removal solvent from a fine nozzle, not a high-vapour-pressure solvent such as acetone;
• a resist that reaches the target at a sufficiently high speed, about 3,000–4,000 rpm;
• for very thick films, a higher speed for a short time;
• an abrupt speed increase at the end of the spin, when the film has dried but the bead is still liquid;
• several coats, each at a higher speed;
• a wait between coating and softbake, so that the viscosity drop at softbake temperature does not enlarge the bead, possibly with staged drying;
• an approved recessed holder whose support, gap and flow geometry have been qualified;
• qualified edge removal after rotation has stopped, where the substrate and subsequent process permit it.

Edge bead removal and backside rinse in the coater are the subject of Edge Bead Removal and Backside Rinse: EBR/BSR in Spin Processing. Square and small pieces are covered in Spin Coating Small, Square and Fragment Substrates: Chucks, Vacuum and Corners.

Comets, bubbles, particles and poor wetting

Summarised from MicroChemicals [1].

Wetting and adhesion problems usually start before the coater. Substrate cleaning, dehydration and adhesion promotion are described by MicroChemicals [7]; why films dewet is explained in Why Your Ink Dewets: Surface Energy, Contact Angle and Wetting Before You Coat, and a full cleaning sequence in Substrate Cleaning Protocol for Thin-Film Devices: Solvents, UV-Ozone, Plasma and Queue Time. A general defect guide is Spin Coating Defects: Comets, Striations, Pinholes and Edge Beads Diagnosed.

8.  Measuring Thickness and Fitting a Spin Curve

MicroChemicals describes several ways to measure resist thickness [1]:

• Optical methods based on transmission or reflection spectroscopy.
• Step and profilometer. Prepare a clean film-to-substrate step on a sacrificial sample without gouging the substrate. Measure away from displaced resist; substrate damage must not be counted as film thickness.
• Developed structures. Useful only if the developer does not noticeably thin the areas meant to remain.

Measure after the specified softbake and a consistent cooling/delay procedure, because residual solvent and measurement timing affect the result [1][6], and measure at the same points each time, for example the centre and several points at a fixed distance from the edge. Methods are compared in How to Measure Thin Film Thickness: Profilometry, Ellipsometry and Interference.

Build a spin curve

A spin curve is a set of films coated at different final speeds with everything else fixed, then measured and plotted as thickness against speed on log-log axes. The approximate reciprocal-square-root relation [1] suggests a straight line of slope near −1/2, but your measured slope is what counts; the reasons it may differ, and how many speeds and repeats are needed, are covered in Spin Speed vs Film Thickness: How to Build and Use a Spin Curve. Use the curve only inside the measured speed range.

Fit your own data: spin curve fitter

The fitter below applies these rules to your own resist films: enter each film’s final speed and its thickness after softbake. It states what the number of speeds supports, runs curvature and repeat checks, and gives a speed for a target thickness only inside the measured range. It opens with a worked example, labelled as example values, not measured data, and performs the fit in your browser.

9.  A Route from New Resist to Qualified Coating

A short route from a new resist to a qualified coating:

1. Define the target. Thickness after softbake, allowed variation, substrate size and shape, and the exposure method that follows [8].
2. Choose the resist. One whose data sheet places the target inside its standard speed window, about 2,000–4,000 rpm for the conventional resists in the cited note [1]. For chemically amplified resists, plan the post-exposure bake from the start, since it is part of the chemistry [9].
3. Prepare the substrate. Clean, dehydrate and prime as the resist supplier recommends [7].
4. Prepare the resist. Let bubbles escape; check its age and storage [1].
5. Write the program. Dispense, spread if textured, fast ramp on smooth substrates, a final-speed time checked at the slowest speed (Section 4), and an edge step if needed [1].
6. Condition the coater. Clean bowl, approved cover and exhaust configuration, and qualified conditioning runs [1].
7. Build the spin curve. At least several speeds with repeats, softbake each film the same way, measure at fixed points [1][6].
8. Fix the recipe. Record speed, time, ramp, dispense volume, softbake, room temperature and humidity.
9. Check uniformity and defects. Centre-to-edge thickness, edge bead, comets; trace each defect back to Section 7.
10. Re-check after changes. A new resist lot, a dilution, a different substrate size or a cleaned coater all justify a short check run.

10.  Levon™ Spin Coaters for Photoresist

The table quotes the Levon™ product pages; “Not stated” means the page does not give the value.

Specifications as published on the Levon™ product pages on 22 October 2026, where “no load” is stated by the page. The supplied configuration is confirmed in the quotation.

How this connects to the guide: the published speed ranges of the EZ6-S and the AC200 S-PP and SE span the 2,000–4,000 rpm window that MicroChemicals gives for conventional spin profiles (Section 3), with room on either side for a spin curve; the other models state a maximum speed only. The published acceleration ranges reach well above the > 1,000 rpm/s ramp mentioned in Section 5, and the multi-step programs listed for the EZ6-S and the AC200 S-PP, SE and PE allow a spread step, a final speed and an edge step in one recipe. The EZ6-S suits small substrates and recipe development; the AC200 series lists edge bead removal and backside rinse as add-on options on several models, relevant to the edge bead issues in Section 7; the AC300-S-CTM handles fragments up to 12-inch wafers. To choose between them, see How to Choose a Lab Spin Coater: Speed, Acceleration, Substrate Size and Glovebox Fit.

11.  Keep Exploring the Lithography Knowledge Hub

Start here

• Spin Coating Explained: Stages, Physics and Film Thickness — stages and physics of spin coating
• Spin Speed vs Film Thickness: How to Build and Use a Spin Curve — building and using a spin curve
• How to Choose a Lab Spin Coater: Speed, Acceleration, Substrate Size and Glovebox Fit — choosing a spin coater

Lithography steps

• Soft Bake, Post-Bake and Annealing on a Hot Plate: What Each Step Does — softbake and post-exposure bake
• Mask Aligner Exposure: Contact vs Proximity and the Resolution Limit — contact and proximity exposure
• Exposure Dose and Puddle Development: Reading a Resist Contrast Curve — dose and development
• Edge Bead Removal and Backside Rinse: EBR/BSR in Spin Processing — edge bead removal and backside rinse

Links will be added as each article is published.

Equipment

• EZ6-S Spin Coater — 100–12,000 rpm, 10 × 10 programs
• AC200 Spin Coater Series — fragments to 8-inch, EBR/BSR options
• AC300-S-CTM Spin Coater — fragments to 12-inch wafers
• Levon™ Spin Coaters — all spin coaters

Questions about a specific spin coater can be sent through the contact form.

12.  FAQ

13.  References