Sustainability
of chemical process requires more energy-efficient processes, utilization of
renewable energy such as solar and wind to drive reactions and separations,
better catalysts to improve activity and selectivity and thus to reduce
separation cost and energy demand, new technologies that are more efficient,
and our ability to tap into underutilized and renewable resources, such as
offshore and stranded gas, biogas, and biomass. The distributed nature of many
underutilized and renewable resources and the low energy density begs for
distributed manufacturing, which can be achieved with modular systems and
process intensification, such as plants on wheels. The design of such systems
needs much more intimate process integration with high fidelity models. A cross
cutting need in all of these systems is the need for better materials, whether
catalysts, adsorbents, battery materials, or electrocatalysts, to improve
performance, reduce cost, catalyst stability, and robustness. Over the past two
decades, multiscale modeling has advanced tremendously, and several algorithms
currently exist. Yet, our ability to apply first principles modeling to process
design is seriously limited due to multiple challenges. In this talk, we will
outline these challenges and introduce computational methods to overcome of
them. Specifically, we will discuss how to handle complex reaction networks
with first principles accuracy but at a very low computational cost, how to
estimate and reduce errors in multiscale models, how to determine the active
site of a catalyst, and how to predict novel combinations of active sites to drive
activity and selectivity. The concepts of small data, correlations in energies
and entropies, correlative uncertainty quantification, machine learning for
catalysis, atomistic optimization for improved activity and stability, will be
discussed. These concepts will be illustrated with examples focusing on ammonia
decomposition chemistry and electrocatalysis.
https://mediaspace.gatech.edu/media/vlachos.mpg/1_0epoh4le
- Tags
-