Heated glasses undergo a variety of kinetic processes that are very important both scientifically and technologically, such as crystallization. When controlled, crystallization is the key to make glass-ceramics but it must be avoided in order to make a glass. In this research we will study some microscopic aspects of crystallization. For example, we want to investigate the eluding "diffusional entity" that controls this process. We are also interested in the initial stages of formation of a critical crystal nucleus, as well as in the average time to form the first critical nucleus. In addition, we seek to develop an artificial neural network (ANN) that is able to predict the glass forming ability (GPA) of complex, multi-component compositions. GPA is inversely related to the easiness to crystallize. With this ANN and forty years of our group's accumulated knowledge, we hope to accelerate the development of new and interesting glass composition which may display unusual properties.
News published in Agência FAPESP Newsletter about the scholarship: