AI/Machine Learning and Sol-Gel Derived Hybrid Materials: A Winning Coupling

Experimental research in the field of science and technology of polymeric materials and their hybrid organic-inorganic systems has been and will continue to be based on the execution of tests to establish robust structure-morphology-property-processing correlations. Although absolutely necessary, these tests are often time-consuming and require specific efforts; sometimes, they must be repeated to achieve a certain reproducibility and reliability. In this context, the introduction of methods like the Design of Experiments (DoEs) has made it possible to drastically reduce the number of experimental tests required for a complete characterization of a material system. However, this does not seem enough. Indeed, further improvements are being observed thanks to the introduction of a very recent approach based on the use of artificial intelligence (AI) through the exploitation of a "machine learning (ML)" strategy: this way, it is possible to "teach" AI how to use literature data already available (and even incomplete) for material systems similar to the one being explored to predict key parameters of this latter, minimizing the error while maximizing the reliability. This work aims to provide an overview of the current, new (and up-to-date) use of AI/ML strategies in the field of sol-gel-derived hybrid materials.