Leveling of thin films over uneven substrates during spin coating

The planarization, that is leveling, of 1–4 μm thick liquid epoxy films over 25–200 μm wide isolated trenches on a silicon substrate during spin coating is determined by photochemically hardening the film and measuring the film profiles over the trenches with a profilometer. The profiles are quantitatively described by a simple lubrication theory that takes advantage of the thinness of the film compared to the feature width, the narrowness of the feature width compared with the distance of the feature from the substrate center, and the rapidity of change in the film profile compared to the overall rate of centrifugally driven film thinning. For a fixed ratio of film thickness to trench depth hf/d, the experimental data fall on a single curve when planarization is plotted against a dimensionless parameter Ω2≡ρω2w3r0/γhf, where ρ is the density of the liquid film, ω is the substrate angular velocity, w is the trench width, r0 is the radial position of the trench, and γ is the film surface tension. For a fixed value of Ω2, planarization improves with decreasing hf/d ratios. Theoretical planarization versus Ω2 curves agree reasonably well with the experimental data. Data for a positive photoresist film form a similar curve to that formed by the epoxy data, but at lower planarization values because of spinning solvent evaporation that causes film shrinkage and degradation of planarization. The effects of evaporation can be accounted for by dividing the spin-coating process into two stages. In the first the film remains fluid and the profile equilibrates with the overall film thickness. In the second the film is immobile and thins only by evaporation.