Acquiring chemical intuition into the catalytic properties of UiO-type monolithic frameworks using machine learning techniques
Lead Research Organisation:
University of Cambridge
Department Name: Chemical Engineering and Biotechnology
Abstract
This project aims to enhance the knowledge about the link between the structural and catalytic properties of monolithic metal-organic frameworks. Herein, I combine the latest developed experimental and computational methods to gain insights into the effect of synthetic conditions on the structure of the obtained frameworks and the structural properties of the frameworks on their catalytic properties for biomass conversion, consecutively.
Monolithic metal-organic frameworks (MOFs) are a novel group of porous materials that were first reported in the Adsorption & Advanced Materials, University of Cambridge, led by Prof. Fairen-Jimenez. The fascination around these monolithic materials stems from their unique features, including the synthesis of nanosized MOF particles during the sol-gel synthesis method, providing a way forward to unlock a major issue in the translation of MOFs to industry: their shaping into useful materials with high porosity while maintaining excellent packing and therefore high density. All these properties, on top of the inherent unique properties of MOFs (e.g. unprecedented high surface areas, catalytic single-site dispersion, and chemical tunability) make monolithic MOFs extremely promising for catalytic applications. Here, two monolithic frameworks, UiO-66 and UiO-67, are proposed in this study to be used for biomass conversion, i.e. glyoxal conversion to glycolic acid.
This project aims to solve two important challenges: a) how to control the structure of these monolithic frameworks by tuning the synthetic conditions; b) how are the chemo-structural properties of the monolithic frameworks interrelated with their catalytic performance. This is particularly important when considering that the synthetic parameters for the sol-gel monolithic synthesis are more complicated than conventional solvothermal reactions used for standard MOFs. In this case, our knowledge about controlling monolithic structures is even less developed than the knowledge we have on the synthesis of conventional materials.
To solve the above challenges, I propose to use design of experiments (DOE) followed by machine learning (ML) techniques to gain intuition into how to: 1) control the structure of monolithic frameworks, especially UiO-type frameworks, to gain optimal structural properties, and 2) gain control over the structural properties of monolithic MOFs and their functionality for glyoxal conversion to glycolic acid. Such knowledge will be further used to design and synthesize efficient catalysts for chemical reactions.
Monolithic metal-organic frameworks (MOFs) are a novel group of porous materials that were first reported in the Adsorption & Advanced Materials, University of Cambridge, led by Prof. Fairen-Jimenez. The fascination around these monolithic materials stems from their unique features, including the synthesis of nanosized MOF particles during the sol-gel synthesis method, providing a way forward to unlock a major issue in the translation of MOFs to industry: their shaping into useful materials with high porosity while maintaining excellent packing and therefore high density. All these properties, on top of the inherent unique properties of MOFs (e.g. unprecedented high surface areas, catalytic single-site dispersion, and chemical tunability) make monolithic MOFs extremely promising for catalytic applications. Here, two monolithic frameworks, UiO-66 and UiO-67, are proposed in this study to be used for biomass conversion, i.e. glyoxal conversion to glycolic acid.
This project aims to solve two important challenges: a) how to control the structure of these monolithic frameworks by tuning the synthetic conditions; b) how are the chemo-structural properties of the monolithic frameworks interrelated with their catalytic performance. This is particularly important when considering that the synthetic parameters for the sol-gel monolithic synthesis are more complicated than conventional solvothermal reactions used for standard MOFs. In this case, our knowledge about controlling monolithic structures is even less developed than the knowledge we have on the synthesis of conventional materials.
To solve the above challenges, I propose to use design of experiments (DOE) followed by machine learning (ML) techniques to gain intuition into how to: 1) control the structure of monolithic frameworks, especially UiO-type frameworks, to gain optimal structural properties, and 2) gain control over the structural properties of monolithic MOFs and their functionality for glyoxal conversion to glycolic acid. Such knowledge will be further used to design and synthesize efficient catalysts for chemical reactions.
