Superhydrophobic Polymer Topography Design Assisted by Machine Learning Algorithms

• DOI: 10.1021/acsami.1c04473
• Publication Date: 2021-06-15T13:32:49Z
• Authors: Qiang Wang; Jarrett J Dumond; Jarren Teo; Hong Yee Low
• Keywords: Laplace pressure; Micrometer; Design of experiments

Superhydrophobic surfaces have been largely achieved through various surface topographies. Both empirical and numerical simulations reported to help understand design superhydrophobic surfaces. Many such successful also using bioinspired biomimetic designs. Despite this, identifying the right texture meet requirements of specific applications is not a straightforward task. Here, we report hybrid approach that includes experimental methods, simulations, machine learning (ML) algorithms create maps for polymer Two objectives investigate properties were maximum water contact angle (WCA) Laplace pressure. The parameters geometries an isotropic pillar structure in micrometer sub-micrometer length scales. finite element method (FEM) was validated by data employed generate labeled dataset ML training. Artificial neural network (ANN) models then trained on database topographic (width W, height H, pitch P) with corresponding WCA ANN yielded series nonlinear relationships between pressure substantial differences Design span topography provide optimal or tradeoff parameters. This research demonstrates potential as rapid tool exploration.