PhD Defence by Abhimanyu Pudi
“An Integrated Multiscale Modeling Framework for Multiphase Reactive Extraction”
Many chemical processes that drive our economy and lifestyle are a result of interactions between complex underlying physicochemical phenomena, especially in the case of intensified processes such as multiphase reactive extraction (MRE) that is a combination of two fundamental process blocks of reaction and separation. Historically, this complexity has often led to countless hours of painstaking experimental work to develop new products and processes or to optimize existing ones, which takes a lot of time and money while still not arriving at the best or the most suitable answer. The rise in empirical or statistical models has eased some of this burden for scientists and engineers in both academia and industry, but these models have limited ability to extrapolate to new systems.
On the other hand, models based on theory can emulate and analyze any molecular system and are only limited by the availability of computational resources. They are universally applicable, unlike empirical or statistical models, and can always provide optimal or nearoptimal solutions, unlike experiments. This is not to say that theoretical models will make experiments obsolete. Science proceeds by both experiment and theory. Utilizing the synergy between experiments and theoretical models provides a promising pathway to a detailed and adequate description of any multiphase reactive system.
As of now, there is no such generally applicable model for MRE, particularly if catalysts are used for reactions and/or interfacial mass transfer. This is the problem we addressed here. How? Systematic modeling approaches have been in the mainstream of process systems engineering for over two decades now. Similarly, molecular modeling has helped computational chemists gain fundamental insights on (especially, catalytic) reactions that has led to the development of many new catalysts and/or reactions over the years. However, these two computational domains have largely remained disjunct. Due to the inherent multiscale nature of multiphase catalysis, we hypothesized that an integrated modeling framework that can bridge observations and theory both within and across the scales of computational chemistry and process systems engineering can lead to the sought-after universal model for these processes.
In this work, we applied a bottom-up approach to design such an integrated multiscale framework that is universally applicable to any physically imaginable chemical system, limited only by the available computational resources. As part of the framework, quantum chemical and statistical thermodynamic methods are used to estimate the necessary phenomenological properties of a system that are then used in mathematical equations describing the macroscopic behavior of a multiphase reactive extraction system. The applicability of the framework is demonstrated in four case studies with different levels of complexity and different objectives. The results of this work showcase that the multiscale framework fills a critical gap in the development of multiphase catalysis.
On the other hand, models based on theory can emulate and analyze any molecular system and are only limited by the availability of computational resources. They are universally applicable, unlike empirical or statistical models, and can always provide optimal or nearoptimal solutions, unlike experiments. This is not to say that theoretical models will make experiments obsolete. Science proceeds by both experiment and theory. Utilizing the synergy between experiments and theoretical models provides a promising pathway to a detailed and adequate description of any multiphase reactive system.
As of now, there is no such generally applicable model for MRE, particularly if catalysts are used for reactions and/or interfacial mass transfer. This is the problem we addressed here. How? Systematic modeling approaches have been in the mainstream of process systems engineering for over two decades now. Similarly, molecular modeling has helped computational chemists gain fundamental insights on (especially, catalytic) reactions that has led to the development of many new catalysts and/or reactions over the years. However, these two computational domains have largely remained disjunct. Due to the inherent multiscale nature of multiphase catalysis, we hypothesized that an integrated modeling framework that can bridge observations and theory both within and across the scales of computational chemistry and process systems engineering can lead to the sought-after universal model for these processes.
In this work, we applied a bottom-up approach to design such an integrated multiscale framework that is universally applicable to any physically imaginable chemical system, limited only by the available computational resources. As part of the framework, quantum chemical and statistical thermodynamic methods are used to estimate the necessary phenomenological properties of a system that are then used in mathematical equations describing the macroscopic behavior of a multiphase reactive extraction system. The applicability of the framework is demonstrated in four case studies with different levels of complexity and different objectives. The results of this work showcase that the multiscale framework fills a critical gap in the development of multiphase catalysis.