AI-Guided Design and Property Prediction for Zeolites and Nanoporous Materials

6 Computer Generation of Hypothetical Zeolites

Estefania Argente¹, Soledad Valero¹, Alechania Misturini², Michael M.J. Treacy³, Laurent Baumes⁴, and German Sastre²

¹ Valencian Research Institute for Artificial Intelligence (VRAIN); Universitat Politècnica de València, Valencia, Spain² Instituto de Tecnología Química UPV-CSIC; Universitat Politècnica de València, Valencia, Spain³ Department of Physics, Arizona State University, Tempe, Arizona, USA⁴ ExxonMobil, Clinton, New Jercey, USA.

6.1 Introduction

Zeolites account for 250 years of history. However, it was only from the 1930s that the science of zeolites really took off [1]. Experimentally speaking, the synthesis of a new zeolite framework that was designed computationally still represents a formidable challenge. For decades, synthesis of zeolites was considered to be essentially an unpredictable task, although it was obvious from the increasing number of new zeolites being synthesized, that the simple rule of tetrahedral coordination and microporosity allowed a very large family of materials. Crystallographic principles were the only guiding rule for a while to generate zeolites not yet synthesized, called hypothetical. In this sense, a large effort has been devoted to predict new zeolitic crystal structures. Started by the pioneering work of Wells [2], the topology-based approaches have been further applied by Smith [3] and O’Keeffe [4,5]. It was not until 1996 that two breakthroughs came onto the scene. Methods to represent ...