Ionic Memory and Computing

A 2026 reference scenario projects that data centers could consume nearly 12% of U.S. electricity by 2030. The human brain operates on roughly 20 watts. (Smith et al., 2026; Levy & Calvert, 2021).

This research examines how ionic materials process, retain, and transmit information, toward principles for energy-efficient computing.

Biological signaling, artificial neural networks, and ionic devices that reproduce aspects of neuronal function.
Biological signaling, artificial neural networks, and ionic devices that reproduce aspects of neuronal function.

Selective Ion Transport

In 2024, 91% of the world’s refined magnet rare earths came from a single country. Their chemical similarity makes separation difficult, limiting routes to more resilient supply chains. (IEA, 2026).

This research examines the molecular basis of selective transport, with the goal of separating valuable elements using fewer resources and generating less waste.

Lanmodulin and a potassium channel with enlarged ion-coordination sites beside a synthetic nanopore showing rare-earth-ion entry, binding, and release.
Lanmodulin and K⁺ channels illustrate biological ion coordination alongside a conceptual synthetic nanopore for rare-earth separation through entry, binding, and release. The Nd³⁺/Dy³⁺ contrast is illustrative and does not report measured separation performance.

Interfacial Molecular Recognition

A 2021 Lancet Commission estimated that 47% of the world’s population had little or no access to diagnostics. Continuous glucose monitors demonstrate the value of molecular measurements beyond the laboratory. (Fleming et al., 2021).

This research studies molecular recognition at interfaces to extend continuous sensing beyond glucose and make molecular measurements more accessible.

Target recognition by biological receptors and by receptors attached to sensing electrodes.
Target recognition by biological receptors and by receptors attached to sensing electrodes.

AI/ML Automation

In a survey of 1,972 scientists, respondents reported spending, on average, 30% of their working time developing scientific software and 40% using it (Hannay et al., 2009).

This research aims to reduce the hands-on workload of molecular simulation through AI-assisted automation, while keeping scientific judgment with the researcher.

Conceptual illustration linking researcher direction, computational automation, illustrative candidate screening, and experimental studies.
Computation-guided experimental discovery. Candidates and screening outcomes are illustrative, not reported results. The robotic arms represent software-based automation, not physical laboratory robots.

Funding & Support

Current
Previous
Current: ACCESS, Notre Dame, NCSA, TACC. Previous: NSF, Illinois, Berkeley, Kwanjeong, NIST, Colorado Boulder.