CBE’s Peng Bai and Wei Fan Receive Two NSF Awards Totaling $1,191,033

Content

Associate Professor Peng Bai and Professor Wei Fan of the UMass Amherst Chemical and Bimolecular Engineering (CBE) Department are awarded two collaborative grants from the National Science Foundation (NSF). One is a three-year, $691,033 award to support the research on “Physics-enabled Deep Learning for Adsorptive Separations of Aqueous Mixtures.” The second is a two-year, $500,000 grant that will fund the project titled “ACED: 3D ConvNets for Discovery of Nanoporous Materials.” Professor Subhransu Maji of the College of Information & Computer Sciences is the co-principal investigator on both projects. 

The team of chemical engineers and computer scientists aims to develop new tools to discover nanoporous adsorbents for complex chemical mixtures by integrating experiments, computer modeling, and machine learning.

Image

Peng Bai

As Bai and his collaborators explain the background of the first award, “Many industrial processes require separating a chemical mixture into its components. For example, crude oil is separated into various fractions to give gasoline, lubricants, asphalt, etc. Alcohol and water are separated to produce spirits and bioethanol for fuel. Such separation processes are highly energy intensive. Energy efficiency may be improved by using adsorptive- or membrane-based separations using nanoporous materials, which are materials with pores that are a few nanometers wide. But mixtures confined within nanopores behave differently from unconfined mixtures.”

To design better nanoporous adsorbents, as the researchers say, “The project will identify general principles governing the behavior of mixtures within confined environments through a combination of experiments, computer modeling, and machine-learning predictions. The results will help improve the energy efficiency of chemical-separation processes.”

This project will initially focus on the non-ideal adsorption in zeolites, a group of microporous, crystalline, aluminosilicate minerals commonly used as commercial adsorbents and catalysts. To accomplish this objective, the research team will analyze aqueous-polar mixtures confined within zeolite nanopores using the framework of the real-adsorbed solution theory.

However, according to Bai and his team, “The performance of these materials ultimately relies upon precisely matching the nanocavity shape to the target application, similar to matching a lock to a key. Identifying optimal structures among millions of possibilities through traditional experiments and physics-based simulations can be impractical or prohibitively labor- and resource-extensive.” 

This use case inspired their second NSF project to develop new machine-learning tools capable of rapidly predicting structure-performance relationships for nanoporous materials with a focus on complex molecules.

The project is funded through the interdisciplinary program “Accelerating Computing-Enabled Scientific Discovery” (ACED), which “seeks to harness computing to accelerate scientific discovery, while driving new computing advancements.” 

Image

Wei Fan

In this spirit, the research team will develop scalable, group-equivariant, three-dimensional, convolutional-neural networks (ConvNets) to exploit the symmetry and invariance of the underlying materials structures, investigate unsupervised learning and multi-tasking to obtain transferable representations, and integrate ConvNets with graph representations to enable zero-shot learning for arbitrary host-guest systems.

The researchers explain that “These computational developments may translate to other extended-materials systems and accelerate the discovery of nanoporous materials for a diverse array of emerging applications for clean energy and sustainability, including gas storage for clean-fuel vehicles, membrane separations, solid-state batteries, and plastic-waste upcycling.” 

These two projects will contribute to developing the next generation of engineers equipped with expertise in computationally assisted materials-discovery and sustainable-economic systems.

Bai is also the principal investigator on a recent two-year, $500,000 Young Faculty Award funded by the Defense Advanced Research Projects Agency to build a machine-learning framework that can rapidly identify the optimal inorganic porous material for a target application and predict if such a structure is feasible to create in reality.