431472 Accelerated Computational Discovery of Materials for Production, Storage, and Efficient Use of Energy

Sunday, November 8, 2015
Exhibit Hall 1 (Salt Palace Convention Center)
Diego A. Gomez Gualdron, Department of Chemical and Biological Engineering, Northwestern University, Evanston, IL

One of the greatest challenges that mankind is facing is to continue the development of technologies that improve our quality of life without depleting our energy resources.  At the center of addressing this challenge is the need to develop new materials that allow us to produce, store, and efficiently use energy. The advent of nanotechnology has given us unprecedented access to the atomistic structure of matter and enabled the rational design of materials by precisely controlling their nanostructure. My primary research interest is to develop high-throughput screening strategies to use with quantum and classical mechanics molecular simulations and other computational methods to (help) design materials that can play pivotal roles to address energy-related challenges in fields such as transportation and chemical industry.

At its core, my research aims to elucidate the links between properties, structure, performance, and processing of materials, with the goal of using this knowledge to accelerate the discovery of new materials.  In this poster, I will provide an overview of relevant research carried out during my doctoral and postdoctoral work. During my doctoral studies, I used molecular simulations (density functional theory (DFT) and reactive molecular dynamics (RMD) to elucidate how the processing or catalyzed synthesis of single-walled carbon nanotubes (SWCNTs) could be modified to produce desired structures of these materials selectively.  Selective synthesis is perhaps the “holy grail” of the nanotube synthesis field, and it would eliminate current bottlenecks for the exploitation of nanotube-based technologies arising from the well-established structural dependence of nanotube properties, and costly and energy-intensive nanotube separations methods. I investigated the feasibility of using the catalyst nanoparticle structure to act as a template to the nascent nanotube, and thus provided useful synthesis guidelines to experts in nanotube synthesis.

During my postdoctoral work, I have sought to design metal-organic frameworks (MOFs) for applications such as natural gas storage, hydrogen storage, carbon capture, and selective catalysis. MOFs are novel, remarkable, crystalline materials whose nanopore structure can be finely tuned for a given application by appropriate selection of essentially limitless possible combinations of inorganic “nodes” and organic “linkers.” Thus the challenge is to efficiently identify the best possible material and elucidate the link between structure and performance, as well as performance boundaries. In addressing these challenges, my work has involved i) developing computational methods to automatedly generate thousands of MOFs structures for subsequent high throughput screening of their properties, and ii) developing simulation strategies that can facilitate the efficient screening of these materials.

In the first case, I have helped developed an automated algorithm used to generate thousands of MOFs using a topologically guided strategy. The goal was to establish limits (through data mining, analysis and visualization) and topologically-dependence performance of this class of materials for natural gas and hydrogen storage applications, and identify the best materials within these boundaries. Through close collaboration with synthetic chemists, some of the best MOFs identified via simulation have been synthesized, and their promising adsorption characteristics have been experimentally confirmed.

In the second case, I have contributed to the development of efficient MOF screening strategies based on genetic algorithms, which has been used as an alternative to the “brute force” screening methods.  As a proof-of-concept, this strategy has been used to identify high-performance MOFs for pre-combustion carbon capture from a database of over hundred thousand materials, and it has demonstrated a reduction of computational expenses used during screening of several orders of magnitude.  Also, I have developed a simulation strategy to investigate the design of hybrid nanoparticle at MOFs (NP@MOF) catalysts for regioselective oxidation of n-butane to n-butanol. Mimicking the working mechanism of enzymes, the nanopores of the MOF are expected to provide control on how the molecule “attacks” the surface of the nanoparticle, thus potentially engendering selective catalysis which is highly desirable for energy-efficient chemical synthesis (separation processes typically account for 50-60%of the energy cost to operate a chemical plant).  A major challenge in modeling reactions at the NP/MOF interface includes the large number of atoms typically needed to construct a chemically meaningful model of the system.  In principle, this problem could be solved using QM/MM methods, but unknowns about the exact structure of the NP/MOF interface structure, among other factors, also hinder the efficient construction of appropriate simulation models. To overcome this problem, I implemented a conceptually simple approach that can be used to efficiently explore how changes in the NP composition affect the energetics of a given reaction under sterically constrained conditions, as would occur at a NP/MOF interface.  In this approach, two size- and shape-tunable rings composed of helium atoms placed on top of the catalyst surface mimic the steric constraints that would arise from a given MOF pore. This model can be implemented with standard DFT programs, and it has been used to investigate the oxidation of butane to butanol on palladium and palladium-alloy catalysts.

Building on my expertise in materials science, computational methods, catalysis, and data mining, my future research plans include the efficient computational design and discovery of catalytic interfaces and MOFs for storage, production, and efficient use of energy, as well as addressing the emerging challenge of polymorphism control on MOFs.

Doctoral Advisor:  Professor Perla B. Balbuena, Texas A&M University, College Station TX

Postdoctoral Advisor: Professor Randall Q. Snurr, Northwestern University, Evanston IL

Extended Abstract: File Uploaded