In Silico Modelling of Evaporation: Implications for Dermal Absorption and Permeability

The permeability of the skin barrier is crucial for evaluating the dermal absorption, bioavailability, and safety of cosmetics and pharmaceuticals. This thesis addresses the impact of volatile evaporation on dermal delivery, an essential factor for accurately assessing formulation efficacy and safety. An in silico model is developed to simulate how evaporation affects the dermal absorption of volatiles, which is then used in three studies. Firstly, to recreate a current and relevant in vitro permeation study. Secondly, to elucidate the permeability from experimental dermal data based on fitted partition and diffusion coefficients. Thirdly, to further the development of an in silico digital skin twin that can be used by non-modelling experts to simulate complex formulation transdermal permeation. Volatile permeants are modelled using vapour pressure, considering evaporation as a passive diffusion process. The evaporation model was then integrated with a skin permeation model and validated with published in vitro permeation test data. The model is used to generate skin permeability coefficients representative of an experimental dataset to better understand permeability. A KNIME-based toolbox is used to create an end-to-end in silico model. The workflow combines physicochemical informatics, molecular dynamics, and quantitative structure-property relationships for complex formulation simulations. The in silico model achieves a good agreement with experimental data and outperforms published in silico models from the Cosmetics Europe ADME Task Force. The mechanistic model for volatile evaporation significantly improved the prediction of dermal delivery. The end-to-end workflow, tested against published in vitro results, accurately predicts formulation and evaporation effects on percutaneous absorption. By automating multiple calculation steps the model provides a user-friendly means to conduct simulations of complex formulations. Recently published permeability datasets reveal weak correlation with hydrophobicity indicating that the Potts and Guy equation's accuracy diminishes for newer datasets, which indicate a more significant non-lipophilic permeation pathway. The growing complexity of modern datasets requires more comprehensive and accurate in silico models that can accommodate nuances of newer experimental data, and the effects of evaporation. To accurately calculate permeability from an in silico perspective the fitting of the in silico model greatly enhances the accuracy and takes into account complex in vitro conditions and the influence of the specific skin sample used.