Session H23: Computational Materials Discovery and Design - Materials for Catalysis

First-principles data-driven discovery of transition metal oxides for artificial photosynthesis

We develop a first-principles data-driven approach for rapid identification of transition metal oxide (TMO) light absorbers and photocatalysts for artificial photosynthesis using the Materials Project. Initially focusing on Cr, V, and Mn-based ternary TMOs in the database, we design a broadly-applicable multiple-layer screening workflow automating density functional theory (DFT) and hybrid functional calculations of bulk and surface electronic and magnetic structures. We further assess the electrochemical stability of TMOs in aqueous environments from computed Pourbaix diagrams. Several promising earth-abundant low band-gap TMO compounds with desirable band edge energies and electrochemical stability are identified by our computational efforts and then synergistically evaluated using high-throughput synthesis and photoelectrochemical screening techniques by our experimental collaborators at Caltech. Our joint theory-experiment effort has successfully identified new earth-abundant copper and manganese vanadate complex oxides that meet highly demanding requirements for photoanodes, substantially expanding the known space of such materials. By integrating theory and experiment, we validate our approach and develop important new insights into structure-property relationships for TMOs for oxygen evolution photocatalysts, paving the way for use of first-principles data-driven techniques in future applications. This work is supported by the Materials Project Predictive Modeling Center and the Joint Center for Artificial Photosynthesis through the U.S. Department of Energy, Office of Basic Energy Sciences, Materials Sciences and Engineering Division, under Contract No. DE-AC02-05CH11231. Computational resources also provided by the Department of Energy through the National Energy Supercomputing Center.   

Computational Nano-materials Design for Spinodal Nanotechnology as a New Class of Bottom-up Nanotechnology

Based on the spinodal nano-decomposition (SND) of dilute magnetic semiconductors (DMS) [1,2], we generalized the SND to the application of catalysis [3,4] and photovoltaic solar-cells [5], where nano-scale particle formation in catalysis and and nano-scale separation of electrons and holes are essential in order to enhance the efficiency. First, we summarize the shape control (Konbu- {\&} Dairiseki-Phases) and dimensionality dependence of crystal growth condition on SND in DMS. Second, we discuss the application of SND for the formation of nano-particles and the self-regeneration in three-way catalysis for automotive emission control by Perovskite La(Fe,Pd or Rh)O$_{3}$. Third, we propose (i) self-regeneration mechanism and (ii) self-organized nano-structures by SND in chalcopyrite Cu(In,Ga)Se$_{2}$, Kesterite Cu$_{2}$ZnSnSe$_{4}$, and Perovskite CsSnI$_{3}$ for the low-cost, environment-friendly and high-efficiency photovoltaic solar cells using first-principles calculations. [1] K. Sato et al., Rev. Mod. Phys., 82, 1633 (2010). [2] T. Dietl, et al., Rev. Mod. Phys., (2015) in press. [3] H. Kizaki et al., Chem. Phys. Lett. 579, 85 (2013). [4] I. Hamada et al., J. Am. Chem. Soc. 133, 18506 (2011). [5] Y. Tani et al., Appl. Phys. Express, 3, 101201 (2010).   

Simulation of Photo-isomerization of Functionalized Azobenzene Derivatives

Photo-isomerization is the process of changing the isomer ($cis$, $trans$) of a molecule using light. In azobenzene this process can be utilized in a Metal Organic Framework (MOF) for adsorption of CO$_{2}$. MOFs are created by two major components, metal ions, and organic molecules which are called linkers. The metal ions and linkers can be coordinated in a way that they form a porous material. In the $cis$ isomer of azobenzene, the MOF’s pore is available to be filled by CO$_{2}$, but in the $trans$ isomer the pore is filled with a benzene ring. The change from $cis$ to $trans$ will evacuate the pore if CO$_{2}$ is present. The important considerations in using azobenzene photo-isomerization as a photo-switch in MOFs are, the quantum yield of the process, and the wavelength of the light which triggers photo-isomerization. By substitution of the functional groups of azobenzene and using the fewest switches surface-hopping algorithm in FIREBALL to simulate the photo-isomerization process we can tune the properties of the molecule as we desire and predict the best substitution sites for azobenzene functional groups. We studied the effects of functionalizing the molecule with OH, CH$_{3}$, NH$_{2}$, NO$_{2}$ and COOH on isomerization quantum yield.   

Computational Design of Metal--Organic Frameworks with High Methane Deliverable Capacity

Metal--organic frameworks (MOFs) are a rapidly emerging class of nanoporous materials with largely tunable chemistry and diverse applications in gas storage, gas purification, catalysis, etc. Intensive efforts are being made to develop new MOFs with desirable properties both experimentally and computationally in the past decades. To guide experimental synthesis with limited throughput, we develop a computational methodology to explore MOFs with high methane deliverable capacity. This \textit{de novo} design procedure applies known chemical reactions, considers synthesizability and geometric requirements of organic linkers, and evolves a population of MOFs with desirable property efficiently. We identify about 500 MOFs with higher deliverable capacity than MOF-5 in 10 networks. We also investigate the relationship between deliverable capacity and internal surface area of MOFs. This methodology can be extended to MOFs with multiple types of linkers and multiple SBUs.   

Computational design of materials for solar hydrogen generation

Photocatalysis has a great potential for the production of hydrogen from aquerous solution under solar light [1]. In this talk, two different approaches toward the computational materials desing for solar hydrogen generation will be presented. Tin (Sn), which has two major oxidation states, Sn$^{\mathrm{2+}}$ and Sn$^{\mathrm{4+}}$, is abundant on the earth's crust. Recently, visible-light responsive photocatalytc H$_{\mathrm{2}}$ evolution reaction was identified over a mixed valence tin oxide Sn$_{\mathrm{3}}$O$_{\mathrm{4}}$ [2]. We have carried out crystal structure prediction for mixed valence tin oxides in different atomic compositions under ambient pressure condition using advanced computational methods based on the evolutionary crystal-structure search and density-functional theory. The predicted novel crystal structures realize the desirable band gaps and band edge positions for H$_{\mathrm{2}}$ evolution under visible light irradiation. It is concluded that multivalent tin oxides have a great potential as an abundant, cheap and environmentally-benign solar-energy conversion photofunctional materials [3]. Transition metal doping is effective for sensitizing SrTiO$_{\mathrm{3}}$ under visible light. We have theoretically investigated the roles of the doped Cr in STO based on hybrid density-functional calculations [4]. Cr atoms are preferably substituting for Ti under any equilibrium growth conditions. The lower oxidation state Cr$^{\mathrm{3+}}$, which is stabilized under an n-type condition of STO, is found to be advantageous for the photocatalytic performance. It is firther predicted that lanthanum is the best codopant for stabilizing the favorable oxidation state, Cr$^{\mathrm{3+}}$. The prediction was validated by our experiments that La and Cr co-doped STO shows the best performance among examined samples [5]. This work was supported by the Japan Science and Technology Agency (JST) Precursory Research for Embryonic Science and Technology (PRESTO) and International Research Fellow program of Japan Society for the Promotion of Science (JSPS) through project P14207. [1] H. Tong, S. Ouyang, Y. Bi, N. Umezawa, M. Oshikiri, J. Ye, \textit{Adv. Mater.} \textbf{24}, 229 (2012). [2] Maidhily Manikandan, Toyokazu Tanabe, Peng Li, Shigenori Ueda, Gubbala V. Ramesh, Rajesh Kodiyath, Junjie Wang, Toru Hara, Arivuoli Dakshanamoorthy, Shinsuke Ishihara, Katsuhiko Ariga, Jinhua Ye, Naoto Umezawa, and Hideki Abe, ``Photocatalytic Water Splitting under Visible Light by Mixed-Valence Sn$_{\mathrm{3}}$O$_{\mathrm{4}}$'' \textit{ACS Applied Materials {\&} Interfaces}, \textbf{6}, 3790 (2014). [3] Junjie Wang, Naoto Umezawa*, and Hideo Hosono, ``Mixed Valence Tin Oxides as Novel van der Waals Materials: Theoretical Predictions and Potential Applications'' \textit{Adv. Energy Mater. }2015$, $DOI: 10.1002/aenm.201501190 [4] P. Reunchan, N. Umezawa, S. Ouyang, J. Ye, \textit{Phys. Chem. Chem. Phys.} \textbf{14}, 1876 (2012). [5] P. Reunchan, S. Ouyang, N. Umezawa, H. Xu, Y. Zhang, and J. Ye, \textit{Journal of Materials Chemistry A}, \textbf{1}, 4221 (2013).   

Understanding and controlling the water stability of MOF-74

Metal organic framework (MOF) materials in general, and MOF-74 in particular, have promising properties for many technologically important processes. However, their instability under humid conditions severely restricts their practical use. In this work, we show that this instability and the accompanying reduction of the CO$_2$ uptake capacity of MOF-74 under humid conditions originate in the dissociation of water molecules at the metal centers. In turn, the reduction in the CO$_2$ uptake capacity occurs as the metal centers are occupied by the remaining OH groups after the water molecules dissociate. It follows that the reduction in CO$_2$ uptake depends on the catalytic activity of MOF-74 towards the water dissociation reaction H$_2$O~$\rightarrow$~OH+H. On the other hand, we show that---while the water molecules themselves do only have a negligible effect on the crystal structure of MOF-74---the OH and H products of the dissociation reaction can significantly weaken the MOF framework and lead to the observed crystal structure breakdown. With this knowledge, we can now propose a way to suppress this particular reaction and therefore control the stability of the system under humid conditions.   

High-throughput discovery of electrochemically stable photocatalysts for oxygen evolution.

Widespread use of artificial photosynthesis hinges upon development of photocatalysts and light absorbers with excellent electrochemical stability in aqueous solution. The poor stability of most semiconductors in the highly oxidizing environment of a solar fuels photoanode has been a key factor limiting the use of many candidates light absorbers. We assess the stability of candidate transition metal oxides (TMOs) in alkaline aqueous environments from calculated Pourbaix diagrams. Our analysis reveals interesting trends in the electrochemical stability of TMOs containing elements which have not traditionally been explored for photocatalysts. Utilizing the Pourbaix diagram analysis as the first screen-layer in a high-throughput workflow that incorporates automating density functional theory and hybrid functional calculations, we screen for electrochemically stable TMO compounds with low band gaps and optimal band edge energies. Applying our new data-driven approach, we successfully identify several new TMOs with promising band gaps and edges that are predicted to resist corrosion under aqueous conditions relevant to solar water splitting. Materials synthesis and electrochemical measurements confirm several predictions and demonstrate the utility of computational screening for identifying new solar fuels materials.   

Rational Co-Design of Polymer Dielectrics for Energy Storage

While intuition-driven experiments and serendipity have guided traditional materials discovery, computational strategies have become increasingly important and a powerful complement to experiments in modern day materials research. With the example of polymer dielectrics for electrostatic energy storage applications, we demonstrate how a rational co-design approach---involving synergies between high-throughput computational screening and experimental synthesis and testing---can be harnessed for quick and efficient discovery. We highlight recent co-design efforts that can potentially lead to replacement of present-day ``standard'' polymer dielectrics (such as biaxially oriented polypropylene) not only by new organic polymer candidates within known generic polymer subclasses (e.g., polyurea, polythiourea, polyimide), but also by organometallic polymers, a hitherto untapped but promising chemical subspace. We also discuss the utilization of vast computational data (generated in the aforementioned process) towards the development of statistical learning models for relevant properties of dielectric polymers, which can further accelerate the guidance to experiments and thus the successful discovery of new materials.   

\textbf{Simulated synthesis of lithium manganese oxide nanostructures and their characterisation.}

Simulated amorphisation recrystallization methods, are now routinely used to generate models of various nano-architectures for metal oxides with complex microstructural details [1]. Nano-architectures, i.e. nano- sphere, sheet, porous and bulk, associated with the Li-Mn-O ternary were synthesised from amorphous spinel nanosphere. The resulting crystallised nanostructures are characterised from visual images, radial distribution functions, XRDs and simulated microstructures. An analysis of microstructures and simulated X-ray diffractions reveals the presence of the layered Li$_{\mathrm{2}}$MnO$_{\mathrm{3}}$ and spinel LiMn$_{\mathrm{2}}$O$_{\mathrm{4}}$ together with a wide variety of defects, including grain boundaries and ion vacancies. [1] T.X.T. Sayle, R.R. Maphanga, P. E. Ngoepe and D. C. Sayle, J. Am. Chem. Soc., \textbf{131 }(2009), 6161-6173   

Study of Hydrogen Adsorption in Titanium, Nickel and Pd Cluster Supported on Graphene Monovacancies

A detailed description of the atomic structure and the energetics of H$_{\mathrm{2\thinspace }}$adsorption on Ti$_{\mathrm{4,}}$ Ni$_{\mathrm{4}}$, and Pd$_{\mathrm{4}}$ clusters on graphene monovacancies is presented. The large binding energy of that clusters on vacancies is a result of strong hybridization between the unsaturated carbon. We found that the binding energy of a single H$_{\mathrm{2}}$ is strongly dependent on the specific cluster. In particular, the H$_{\mathrm{2\thinspace }}$bond cleavage is favored by titanium clusters. On the other hand, the Ni and Pd clusters favours the formation of Kubas complexes. The analysis of the adsorption energies and H$_{\mathrm{2\thinspace }}$average bond lengths suggests that supported Ti$_{\mathrm{4\thinspace }}$cluster is a potential hydrogen storage candidate, being able to hold up to 6 molecules covalently with moderate average binding energy within the optimal range for an efficient cyclic adsorption/desorption process at room temperature and moderate pressures. These results give physical insigths about how the interaction between graphene monovacancies and metal clusters can be used to enhanced the hydrogen uptake, towards of design of new graphene-based material for hydrogen storage.   

Using Cluster Expansions to Model Diffusion in Pt-Ni Nanoparticles

Pt-Ni alloys have been shown to have excellent catalytic activity for the oxygen reduction reaction. However in practice these particles suffer from Ni dissolution, degrading the performance of the catalyst over time. The exact mechanism by which this occurs is unknown and difficult to determine experimentally. Using density functional theory, we have calculated the energies of formation of several dozen PtxNix-1 nanoparticles with included vacancies and used this data to parametrize a cluster expansion. Because the activation energies for diffusion are included in the model, we are able to use the cluster expansion to rapidly predict the transition rates in a Pt-Ni nanoparticle as a function of particle shape, size, and local atomic order. By using the cluster expansion model in a kinetic Monte Carlo algorithm we are able to model the diffusion of Pt and Ni and provide insights into the dissolution of Ni from Pt-Ni nanoparticles.