Observation-scale dependence of apparent dynamic contact angle in droplet impact on a hydrophilic silicon wafer surface

Abstract Droplet spreading on hydrophilic solid surfaces is important in engineering processes such as semiconductor cleaning and coating, where the initial wetting behavior influences process performance. The contact angle extracted from experimental images or numerical interfaces is a finite-scale apparent quantity, and its value can depend on the observation scale used for interface fitting. This dependence is especially important on strongly hydrophilic surfaces, where the interfacial slope near the contact line is small. In this study, the observation-scale dependence of the apparent dynamic contact angle was investigated for a water droplet impacting a UV-treated hydrophilic silicon wafer. The results of high-speed imaging experiments were compared with volume-of-fluid simulations using an experimentally derived dynamic contact-angle boundary condition. The same local curve-fitting procedure and the same finite observation-scale definition were applied to both experimental images and numerical interfaces. The measured apparent contact angle depended on the observation scale; increasing the observation scale tended to reduce measurement noise, whereas an excessively large observation scale reduced the locality of the fitted interface region. The apparent dynamic contact angle extracted from the computed interface also depended on the grid resolution. Under the present conditions, grid refinement and selection of an observation scale containing sufficient interface-sampling points reduced the discrepancy between the apparent dynamic contact angles obtained from simulations and experiments. Applying the same local curve-fitting procedure and observation-scale definition to both data sets provides a common finite-scale framework for determining and comparing apparent dynamic contact angles from experimental images and numerical interfaces, complementing evaluation based on macroscopic spreading behavior.