Development of Contact Angle Prediction for Cellulosic Membrane

Contact angle (CA) of a membrane determines its application. Accurate CA prediction models are available. Nevertheless, detailed membrane roughness properties and thermodynamic data of the interaction between the membrane and water droplet are required. These data are not easily accessible, and it is not available for newly developed material. This study aims to apply Artificial Neural Network to estimate the CA by using pure water flux, membrane porosity and its pore size as inputs. This model was tested on two type of filtration processes: dead end (DE) and cross flow (CF). The results showed that the prediction for DE achieve an overall accuracy of 99 with a sample size of 53 data sets. The prediction for CF could be done by using DE + CF model with a maximum R2 at training stage of 0.9456. In conclusion, a novel statistical solution to predict CA for cellulosic membrane was developed with high accuracy. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.