| Project Detail |
Responsive soft materials, made of crosslinked polymer networks, have emerged as a promising class of systems. In particular, thermoresponsive hydrogels and microgels are widely used from drug delivery to cell engineering and art restoration. An important feature of such systems is their internal elasticity, which can be tuned according to their preparation protocol and crosslinker concentration. Despite recent progress in their modeling by means of numerical simulations, there is no way to know a priori their structure and elasticity from only their initial constituents, and hence, extensive exploration of initial parameters through simulations and experiments must be performed. To overcome this difficulty, in the present proposal MGELS -Machine-learning polymer Gel’s ELasticity and Structure- I will exploit my existing knowledge on Machine Learning (ML) methods to develop novel tools able to predict structural and elastic properties of hydrogels and microgel particles to design in silico polymer networks with desired features.
The project is based on four objectives. The first is to develop a ML-Neural Network (ML-NN) able to predict the network structure using data from molecular dynamics simulations of microgels and hydrogels already collected from the supervisor. The second is to explore structural and elastic properties of new configurations of networks, following the methods established in the host group. As a third objective, we will create a new database of polymer networks that will be made open access at the end of the project. Finally, we will extend the ML-NN approach to predict elastic properties and to identify the onset of auxetic behavior. With the accomplishment of these goals, we will be able to fully predict polymer networks properties, being very appealing for the soft matter and materials science communities. The scientific outcomes from the action and the development of new skills will help me to establish my future career in academia. |
