Material Informatics for Discovery of Porous Materials
Porous materials contain complex networks of void channels and cages that are exploited in many industrial applications. The zeolite class of these materials is the most well-known as they have found wide use in industry since the late 1950s, with common applications as chemical catalysts and membranes for separations and water softeners (their value is estimated at $350 billion per year). There is increasing interest in utilizing zeolites as membranes or adsorbents for CO2 capture applications. In addition to zeolites, metal organic frameworks (MOFs) and their subfamily of zeolitic imidazolate frameworks (ZIFs) have recently generated interest for their potential use in gas separation or storage. A key requirement for the success of any nanoporous material is that the chemical composition and pore geometry and topology must be optimal at the given conditions for a particular application. However, finding the optimal material is an arduous task, since the number of possible pore topologies is extremely large. There are approximately 190 unique zeolite frameworks known to exist today in more than 1400 zeolite crystals of various chemical composition and different geometrical parameters. However, these experimentally known zeolites constitute only a very small fraction of more than 2.7 million structures that are feasible on theoretical grounds. Databases of similar or greater magnitude can be developed for other nanoporous materials such as MOFs or ZIFs. As a result, new automated computational and cheminformatic techniques need to be developed to characterize, categorize, and screen such large databases. Our Material Informatics team focuses on development of such techniques as well as work together with our collaborators (chemists, chemical engineers, mathematicians and computer scientists) to use these techniques to discover new materials with outstanding properties.
Algorithms and tools for high-throughput geometry-based analysis of crystalline porous materials
Identifying suitable materials for a given application can, in principle, be done by screening material databases. Such a screening requires automated high-throughput analysis tools that calculate structural properties for all materials contained in a database so they can be compared with search queries, grouped or classified. One important aspect of the structural analysis of materials such as zeolites and metal organic frameworks is the investigation of the geometrical parameters describing pores. We have been developing algorithms and tools to efficiently calculate some of these important parameters. Our tools are based on the Voronoi decomposition, which for a given arrangement of atoms in a periodic domain provides a graph representation of the void space. The resulting Voronoi network is analyzed to obtain the diameter of the largest included sphere and the largest free sphere, which are two geometrical parameters that are frequently used to describe pore geometry. Accessibility of nodes in the network is also determined for a given guest molecule and the resulting information is later used to retrieve dimensionality of channel systems as well as in Monte Carlo sampling of accessible surfaces and volumes. The presented algorithms are implemented in a software tool, Zeo++, which includes a modified version of the Voro++ library.
In Silico Screening of Carbon Capture Materials
One of the main bottlenecks to deploying large-scale carbon dioxide capture and storage (CCS) in power plants is the energy required to separate the CO2 from flue gas. For example, near-term CCS technology applied to coal-fired power plants is projected to reduce the net output of the plant by some 30% and to increase the cost of electricity by 60-80%. Developing capture materials and processes that reduce the parasitic energy imposed by CCS is therefore an important area of research. We have developed a computational approach to rank adsorbents for their performance in CCS. Using this analysis, we have screened hundreds of thousands of zeolite and ZIF structures and identified many different structures that have the potential to reduce the parasitic energy of CCS by 30-40% compared to near-term technologies.
Addressing challenges of identifying geometrically diverse sets of crystalline porous materials
Efficient handling and screening of large sets of materials require special structure representations and structural descriptors. We have developed a novel descriptor that captures shape and geometry characteristics of pores. Together with proposed similarity measures, it can be used to perform diversity selection on a set of porous materials. Our representations are histogram encodings of the probe-accessible fragment of the Voronoi network representing the void space of a material. They have proven their value in analysis of structures in the International Zeolite Association (IZA) database of zeolite frameworks and the Deem database of hypothetical zeolites, as well as zeolitic imidazolate frameworks constructed from IZA zeolite structures. The diverse structures retrieved by our method are complementary to those expected by emphasizing diversity in existing one-dimensional descriptors, e.g. surface area, and similar to those obtainable by a (subjective) manual selection based on materials' visual representations. Our technique provides means to reduce large sets of structures and enables the material researcher to focus efforts on maximally dissimilar structures. There techniques have been implemented in Zeo++ code.
Navigating molecular worms inside chemical labyrinths
Predicting whether a molecule can traverse chemical labyrinths of channels, tunnels, and buried cavities usually requires performing computationally intensive molecular dynamics simulations. Often one wants to screen molecules to identify ones that can pass through a given chemical labyrinth or screen chemical labyrinths to identify those that allow a given molecule to pass. Because it is impractical to test each molecule/labyrinth pair using computationally expensive methods, faster, approximate methods are used to prune possibilities, "triaging" the ability of a proposed molecule to pass through the given chemical labyrinth. Most pruning methods estimate chemical accessibility solely on geometry, treating atoms or groups of atoms as hard spheres with appropriate radii. We have developed an alternative approach in which we explore geometric configurations for a moving "molecular worm", which replaces spherical probes and is assembled from solid blocks connected by flexible links. The key is to extend the fast marching method, which is an ordered upwind one-pass Dijkstra-like method to compute optimal paths by efficiently solving an associated Eikonal equation for the cost function. First, we build a suitable cost function associated with each possible configuration, and second, we construct an algorithm that works in ensuing high-dimensional configuration space: at least seven dimensions are required to account for translational, rotational, and internal degrees of freedom. We demonstrated the algorithm to study shortest paths, compute accessible volume, and derive information on topology of the accessible part of a chemical labyrinth.
Chemoinformatics Projects - Discovery of Molecules with Interesting Properties
C3 approach for advanced molecular and material design
Typically a molecular/materials designer has a specific property in mind, such as the emission/absorption wavelength or the particular band structure, and searches for stable molecules/materials that would display the desirable value of the targeted property. My main goal is to develop algorithms and software tools that would facilitate combinatorial searches of this type based on the results of quantum mechanical electronic structure calculations. For example, the combinatorial-electronic structure approach could be used in the design of alloys. It could systematically screen various alloy compositions. The scientists would just have to decide on atom types included in the searches, and on ranges of concentrations. The software, with my approach implemented, would generate all possible alloys within the range of requested compositions, run required calculations, analyze the results, and finish up with a short list of alloy compositions that might have the requested property. In a similar manner the searches for novel electronic materials might be performed. For example, different compositions of materials might be screened to find ones with the requested band gap. The targeted property might be complex, e.g., a specific band gap combined with a specific band offset when interfaced, e.g., with silicon. The new hybrid methods may have an obvious application in the area of design and development of small-molecules like luminophores or indicators.
The hybrid quantum chemical/combinatorial approach has the potential to become a powerful tool in rational molecular/materials design. The approach involves three steps: (i) combinatorial generation of libraries of compounds, and (ii) screening of the libraries for the targeted property using electronic structure methods; (iii) analysis of generated data. The steps i-iii correspond to methodology employed, namely combinatorial, computational and chemoinformatics techniques, respectively. Therefore I proposed to name this hybrid approach as "Combinatorial*Computational*Chemoinformatics", or just abbreviated as C3 (or C-cube) approach.
Enumeration and Characterization of Congeners of Common Persistent Organic Pollutants
Congeners are molecules based on the same carbon skeleton but different by the number of substituents and/or a substitution pattern, eg. 1-chloronaphthalene, 1,4-dichloronaphthalene, 1,3,8-trichloronaphthalene etc. Various Persistent Organic Pollutants (POPs) exist in the environment as families of halogen substituted congeners and/or their hydroxyl and methoxy substituted derivatives. In this project we used C3 approach to generate and characterize libraries of congeners.
Visualization of molecular orbitals and the related electron densities
When plotting different orbitals with consistent contour values, one can create illusions about the relative extension of charge distributions. My study suggests that the comparison is not biased when plots reproduce the same fraction of the total charge. I have developed an algorithm and software that facilitate this type of visualization. The software allows superimposing molecules and associated orbitals, and selecting a particular part of the orbital limited by pre-defined planes.