Our Research In silico discovery of metal-organic frameworks for selective ion separation

Machine learning models will learn which properties of metal-organic frameworks make them stable

A diagram showing the multi-faceted approach to discovering stable metal-organic frameworks suitable for water purification

A multi-faceted approach to discovering stable metal-organic frameworks suitable for water purification

Machine learning models will learn which properties of metal-organic frameworks make them stable

A multi-faceted approach to discovering stable metal-organic frameworks suitable for water purification

Principal Investigator

Heather J. Kulik

Lammot du Pont Professor
Department of Chemical Engineering
Department of Chemistry

Heather Kulik is the Lammot du Pont Professor in the Departments of Chemical Engineering and Chemistry at MIT. She earned a bachelor’s degree in chemical engineering from The Cooper Union in 2004 and a PhD from MIT in 2009, followed by postdoctoral training at Lawrence Livermore National Laboratory and Stanford University. Since joining the MIT faculty in 2013, she has received numerous honors, including the DARPA Young Faculty Award, NSF CAREER Award, Sloan Fellowship, and the Presidential Early Career Award for Scientists and Engineers. Since 2025, she has served as an Associate Editor of the Journal of the American Chemical Society.

Photo Credit: Gretchen Ertl

Challenge:

How can we identify optimal materials for water purification that have precision pores for separation and good stability in operating conditions?

Research Strategy

Leverage and extend existing models of stability for metal-organic frameworks (MOFs) to predict suitable MOFs for ion separation
Design a new dataset of "ultrastable" candidate MOF materials
Screen "ultrastable" MOFs with molecular simulation to identify optimal materials for water purification

Project description

The treatment and reuse of water mandate the ability to purify this precious resource by selectively removing small ions from solution. Nature has achieved this exquisite selectivity, but manmade materials lag behind despite the primary importance of selective ion separation in scalable materials for water purification and desalination. Metal-organic frameworks (MOFs) represent promising materials for this task because their pores can be tailored to have precise shapes and chemical makeup for selective ion affinity. Nevertheless, the combinatorial space of all conceivable MOFs is vast, and few MOFs have been assessed for their properties relevant to water purification. Many MOFs will break down at reasonable temperatures or otherwise lack stability, including in water.

To unlock the potential of MOFs for water purification, virtual high-throughput screening (VHTS) accelerated by machine learning (ML) models and molecular simulation can accelerate discovery of MOFs that do not have these limitations. Specifically, these VHTS strategies will use ML to leverage existing knowledge about what dictates MOF stability. This research project will develop novel computational strategies to identify optimal MOF materials for water purification by curating and searching a wide space of new "ultrastable" MOF structures.