After recognizing that evaporation was playing an important role in causing chuck marks, we designed an experiment to measure the magnitude of the cooling that was caused by this active, forced, evaporation from the surface. Our experiment was to use a specially designed Infra-Red (IR) video camera to image the surface of a spinning substrate with surface layers of solvent being spin coated on the surface (even though as pure solvents they spin completely off leaving a completely dry substrate behind).

        This figure shows four different frames captured from the IR video camera at different times for a run using a relatively volatile solvent (acetone) on a glass substrate. As an IR image, the brightness indicates the relative temperature (brighter is warmer and darker is cooler). Temperature/colors are calibrated by using standard black-body emitters. Frame (a) shows the wafer before fluid of the same temperature is dispensed. Frames (b), (c), and (d) show time intervals at 1, 3, and 5 seconds into spinning, respectively. Spin speed was 500 RPM.
        Most of the observed cooling has occurred already within the first second. Then with increased spinning time, the forced air currents gradually help the wafer warm up to the temperature of the surrounding air. With a volatile solvent (as used for this experiment) the fluid residence time on the wafer surface is quite short. The center of the 4" substrate is substantially warmer than the surrounding ring. This is caused by intimate physical contact between the small metal vacuum chuck and the back of the substrate. The vacuum chuck was small enough to use with 2" diameter wafers, so lots of wafer was sticking out over the edge and therefore not affected by any contact with metal.

This work was published in the literature at this location:

D. P. Birnie, III, B. J. J. Zelinski, and D. L. Perry, "Infrared Observation of Evaporative Cooling During Spin Coating Processes", Optical Engineering, 34, 1782-1788 (1995).