2018 AIChE Annual Meeting
(6bg) Catalysis Informatics: Accelerating Search and Discovery of New Catalysts
Author
I am interested in the development of tools for catalysis informatics. More specifically, this includes methods for improving the rate at which we search through possible catalytic systems of interest using database mining, machine learning, and high-throughput screening techniques.
During my PhD, I studied computational catalysis at Carnegie Mellon University. I worked with Professor John Kitchin to develop a new kind of atomic potential constructed from neural networks. These potentials have become increasingly popular in the last decade with the advent of fingerprinting schemes, proposed by Behler and Parrinello, which allow them to be trained to a wide variety of atomic input structures. The methods are now capable of reproducing potential energy surfaces (PES) of a multitude of chemical systems with levels of accuracy unrivaled by their physical potential counterparts1,2. They are also significantly faster to evaluate than a full quantum chemistry calculation resulting in the ability to study much larger unit cells and more complex chemistry.
For the last year, I have been working as a post-doc at Stanford working with Thomas Bligaard and the catalytic methods development team at SUNCAT. Here I am developing the tools needed to realize high-throughput techniques for the field of computational catalysis. To accomplish this goal, there is a need to enumerate the space of all possible catalytic surfaces of interest. In cheminformatics this is performed through using graph theory tools which allows molecules to be considered as component vertices (the atoms) connected by bonds (edges). With this basic precept, a great number of molecular structures of interest can be represented computationally; a critical first step to any computational high-throughput implementation.
For catalysis, there is a bulk with and surface to be considered along with the adsorbed molecule and producing a unique descriptor is not trivial. In the field of materials informatics, systems are represented in 3D periodically bound boxes, so graphical methodologies are not commonly used. My recent work has focused on beginning from bulk structures and generating graphs from these periodic systems. From a bulk structure, any slab of interest in heterogeneous catalysis can be generated. Finally, unique adsorption sites can be identified through symmetries and used to add molecules to the surface in a systematic way, resulting in a method for producing a unique connectivity matrix for slabs with adsorbates; the basis for a complete catalysis informatics approach. These methodologies are available through the open source Python package, CatKit, for which I am the lead developer.
As available computational resources continue to increase, the rate of data generation in the field of catalysis is quickly beginning to outstrip the ability of those in the field to interpret it all. To improve the ability of all those in the field of catalysis to continue making rapid scientific progress, there is a need to develop better databases which are capable of storing and searching the information; similar to the way AFLOW, Materials Project, OQMD, and others have done for the field of material science. For catalysts, this is significantly more challenging due to the even larger number of possible surfaces and adsorption configurations of interest. Even with a robust database structure capable of storing a breadth of relevant information, tools capable of intelligent and sophisticated queries to the database are also needed. Machine-learning not only benefits from such large repositories of information, it can also essential in making more intelligent selections from a large number of interesting structures to find new and interesting results. These tools will prove to be fundamental for high-throughput research.
As an instructor, I teach: thermodynamics, kinetics, molecular simulation (and underlying quantum theory), reaction engineering, functional coding languages, and general chemical engineering. Chemical engineers fulfill a diverse range of demand in industry positions today. As modern tools become increasingly computationally intensive, there is an ever growing need to shift the emphasis of education to computational tools and techniques. As such, I strongly support the integration of coding into the classroom. I am also interested in perusing educational research to do with code based instruction which I describe in greater detail below.
I have completed instruction on educational techniques for engineering in an evidence-based teaching course specific to STEM fields. I am a strong proponent of the flipped classroom as a pedagogical teaching model. This model requires students to study lecture material before the class and implement it on assignments given during class. Among other benefits, such a system allows the instructor to provide immediate feedback to students which is shown to improve student retention rates and longevity of student knowledge. Further details of my pedagogical teaching strategies and achievements can be found on my website.
I would like to study the potential benefits of teaching code-based chemical engineering tools in an immersive environment. The concept is similar to that of language immersion, in which various subject material is taught through the medium of a second language. Language immersion is known to dramatically improve student comprehension of the second language in technical communication with no discernible negative impact on primary language skills. Teaching students to solve chemical engineering problems can similarly be framed in the medium of a coding language, such as Python or MatLab.
Relevant Work
1. Jacob R. Boes, Mitchell C. Groenenboom, John A. Keith, and John R. Kitchin. Neural network and ReaxFF comparison for Au properties. International Journal of Quantum Chemistry, 116(13):979-987, 2016. DOI
2. Jacob R. Boes and John R. Kitchin. Modeling Segregation on AuPd(111) Surfaces with Density Functional Theory and Monte Carlo Simulations. The Journal of Physical Chemistry C, 121(6):3479-3487, 2017. DOI
Phone: (607) 342-1846
Email: jacobboes@gmail.com | jrboes@stanford