Chemical, mechanical, electronic, and thermal properties of materials all change at the nanoscale. Entropic fluctuations become more important and materials confined in 1- or 2- dimensions (tubes, wires, sheets) behave differently. Capturing these effects in real devices and applications requires a range of modeling approaches, from hard theory (DFT and kinetics), to soft theory (continuum, statistical mechanics and molecular dynamics), and up through systems engineering approaches.

Applications of include biomedical sensors (nanotube-based optical sensors) and energy (CO2 to fuels, fuel cells, thermal catalysis).

Machine-learning approaches to accelerate materials discovery

Density functional theory (DFT) is the tool of choice for studying reactions on chemical surfaces. However, it is prohibitively expensive to use full-accuracy DFT to identify catalyst surface with optimal properties due to the sheer number of possible materials, facets, and active sites. The search process can be greatly accelerated by using flexible surrogate models that learn from previous DFT calculations to predict interesting materials. This process can be accelerated with deep-learning methods from the computer science community. Prof. Ulissi contributes to the AMP software package and is exploring larger models and networks for accelerated learning and uncertainty quantification using Google’s tensorflow library. Applications include both thermal and electrochemical catalysis.

High-throughput calculations for materials informatics

Developing new structure-property relationships or machine learning regression methods requires large self-consistent datasets that are rare in literature. We use and develop modern workflow tools like fireworks and luigi to organize, distribute, and analyze many thousands of DFT calculations. These methods drive down the cost of investigating new chemistries and materials while simultaneously making the results accessible and searchable. This approach also allow for on-line material optimization or active learning approaches to be applied and tested quickly.

Bayesian methods for reaction mechanism reduction

Identifying the most important reaction mechanisms in hydrocarbon reaction networks remains a challenge for the computational chemistry community. Treating all possible intermediates with full-accuracy DFT is impossible, but these tools can be focused on the most interesting reactions. Bayesian techniques allow uncertainty to be propagated through these networks and through multiple levels of approximation and the most important reaction steps to be identified probabilistically. Current applications are large hydrocarbon reaction networks in thermal catalysis.

Nanoscale selectivity

Interfacial selectivity is central to catalysis and biomedical imaging. Nanoparticles with desirable properties (optical, electronic, catalytic) are often available, but face challenges in complex chemical environments with a variety of substrates. Restricting access to these surfaces using soft functionalizations (polymers, DNA, surfactants) has been successful in developing selective carbon-nanotube sensors, but designing the selectivity a priori remains a challenge. Prof. Ulissi uses molecular simulations and thermodynamic models to study interfacial packing and selectivity at these interfaces.