Session J36: Focus Session: New Energy III

Tailoring Surface Chemical Properties Using Electronic Structure Theory

Electronic structure methods based on density functional theory have reached a level of sophistication where they can be used to describe complete catalytic reactions on transition metal surfaces. This opens the possibility that computational methods can be used to tailor surfaces with desired chemical properties. Recent progress in this direction for transition metal catalysts will be discussed. A series of concepts will be introduced to describe and understand trends in reactivity from one metal surface to the next. It is shown how these concepts can be used to identify the factors determining the catalytic activity of a given transition metal surface, and how this can form the basis for screening of a large number of metals and alloys for catalytic properties.   

Ab initio reaction pathways for dissociative adsorption of dioxygen on Al (111)

We investigate the interaction of dioxygen with a clean aluminum (111) surface. The theoretical description of this fundamental process is challenging due to the discrete, abrupt charge transfer (CT) from the metal surface to the molecule. Indeed, experimental investigations suggest a sizeable activation barrier not accounted for by a conventional DFT description, due to exchange-correlation functional errors. We adopt a different approach, embedding a small (~12 atoms) aluminum cluster in a DFT-derived potential simulating the remainder of the Al surface. The interaction between this embedded cluster and an approaching $O_2$ molecule is treated using high-level correlated wave function methods (CASSCF, CASPT2) that allow for a correct description of the CT process involved. We map out the potential energy surface (PES) as a function of dioxygen bond length, orientation, and position. In agreement with experiment, we find an activation barrier of $\sim$500 meV, which corresponds primarily to the cost to induce CT. Additionally, the PES is consistent with oxygen abstraction as the dominant process in the case of incident perpendicular orientation, confirming the mechanism proposed to explain the surprisingly large fraction of single oxygen atoms found in STM measurements.   

Alkane and CO$_{2}$ Activation by Modified Oxide Catalysts

I will present computational and experimental results concerning the effect of cation or anion doping on the catalytic activity of oxides. We are mainly interested in the activation of alkanes or of CO$_{2}$ and their conversion to useful products.   

First-principles calculations of Fischer-Tropsch processes catalyzed by nitrogenase enzymes

The nitrogenase enzyme system of the bacteria $Azotobacter vinelandii$, which is used in nature to catalyze ammonia synthesis, has been found recently to catalyze the efficient conversion of carbon monoxide (CO) into hydrocarbons under ambient temperature and pressure [1]. These findings indicate that nitrogenase enzymes could inspire more efficient catalysts for electrochemical CO and CO$_2$ reduction to liquid fuels. The nitrogenase variants, in which vanadium substitutes the molybdenum in the active site of the enzyme, show distinct features in their reaction pathways to hydrocarbon production. To compare and contrast the catalytic properties of these nitrogenase enzymes, we perform first-principles calculations to map out the reaction pathways for both nitrogen fixation and for the reduction of CO to higher-order hydrocarbons. We discuss the trends and differences between the two enzymes and detail the relevant chemical species and rate-limiting steps involved in the reactions. By utilizing this information, we predict the electrochemical conditions necessary for the catalytic reduction of CO into fuels by the nitrogenase active sites, analogous to a Fischer-Tropsch process requiring less extreme conditions. \\[4pt] [1] Y. Hu, C.C. Lee, M.W. Ribbe, Science {\bf 333}, 753 (2011)   

Low-temperature dehydrogenation catalysis of isophorone (C$_9$H$_{14}$O) on Pd(111)

Selective hydrogenation and dehydrogenation of hydrocarbons is one of the most extensively used processes in catalysis. In praticular, it is important to design catalysis which can occur at or below room temperature. Here, we investigate the adsorption and dehydrogenation of a model organic/metal catalyst: isophorone (C$_9$H$_{14}$O) on the Pd(111) surface. Density-functional theory (DFT) calculations including van der Waals (vdW) interactions are carried out to elucidate the adsorbate geometry and reaction dynamics. The vdW interactions dramatically modify the potential-energy surface and enhance the binding energy by about 1 eV. The combination of our theoretical results with the analysis of infrared (IR) spectra and temperature-programmed desorption (TPD) experiments leads us to propose a dehydrogenation pathway from the weakly chemisorbed reactant (C$_9$H$_{14}$O) to the strongly chemisorbed product (C$_9$H$_{10}$O), which occurs after four C-H bond cleavages. The obtained low dehydrogenation temperature (130-150 K) from PBE+vdW barriers, including zero-point energy vibrations, is consistent with the TPD analysis.   

Interfacial binding in catalytic and photovoltaic materials, and the energetics of elementary surface processes during catalytic fuels production and utilization

The fundamental understanding of reactivity trends in organic and inorganic chemistry has led to spectacular scientific advances over the past 50 years in synthetic chemistry. At the basis of this understanding is knowledge of the strengths of the relevant chemical bonds being broken and formed. In this regard, surface chemistry is several decades behind. Yet surface chemical reactivity dictates our choices of materials for energy technology, including catalysts for clean fuels production and utilization, fuel cell electrodes, photocatalysts and photovoltaics. This past decade has seen important advances in our ability to measure surface chemical bond energies and to use them to make predictions of relevance to energy technology. We will review those advances here, with a focus on: (1) the interfacial bonding strength between the metal and the support material in metal nanoparticle catalysts, with emphasis on the roles of particle size and the support in catalyst reactivity and stability, (2) the interfacial reactions that occur between the metal electrode and the semiconducting polymer in organic photovoltaic (OPV) devices, and the heats of those reactions, and (3) the adsorption energies of small molecules on Pt(111), with emphasis on intermediates in catalytic steam reforming, oxidation and dehydrogenation, which reveal a strong correlation between the strength with which adsorbates bond to the Pt(111) surface and their bond energy to the H atom in gas-phase molecules. These measurements provide important benchmarks for comparisons with new computational methods designed to improved energy accuracy beyond standard periodic DFT.   

Ab initio modeling of EXAFS spectra of nanoparticles

The structure of nanoparticles in the 1-3 nm range is often investigated using extended x-ray adsorption fine structure (EXAFS). Structural information for nanoparticles is typically determined from the fits to EXAFS data using the bulk structure for theoretical calculations. The average coordination numbers, interatomic distances, and their mean squared disorder can be found. The applicability of this procedure is less obvious when the particles are strongly disordered, i.e., when there are significant structural differences between the nanoparticles and the corresponding bulk structure. We use molecular dynamics (MD) simulations of 2 nm nanoparticles to construct EXAFS spectra ab initio, with the atom-by-atom approach, averaged over the time of the MD run. We obtain that the analysis of this data done by conventional procedures reveals significant differences between the actual (time- and configuration- average) structure of the simulated nanoparticles and what was determined from data analysis using a bulk reference. We demonstrate that EXAFS spectra calculated for a variety of theoretical models of nanoparticles can be directly compared to the experiment and thus used to determine the best-fit structure.   

Predicting the Catalytic Reactions using Random Phase Approximation

Density functional theory has became the workhorse for simulations of catalytic reactions and computational design of novel catalysis. The generally applied semi-local exchange-correlation functionals have successfully predicted catalytic reaction trends over a variety of surfaces. However, in order to achieve quantitative predictions of reaction rates for molecule-surface systems, in particular where there is weak Van der Waals interactions or strong correlation, it is of vital importance to include non-local correlation effects. The use of random phase approximation (RPA) to construct the correlation energy, combined with the exact, self-interaction free exchange energy, offers a non-empirical way for accurately describe the adsorption energies [1] and dispersion forces [2]. We have recently implemented RPA in the GPAW code [3-4], an electronic structure package using projector augmented wave method and real space grids. In this talk I will present our initial results comparing RPA and generalized gradient functionals for the activation energies and reaction energies for transition metal or metal oxide surfaces. \\[4pt] [1] L. Schimka, et.al, Nature Mat. 9, 741 (2010) [2] T. Olsen, et.al, Phys. Rev. Lett. 107, 156401 (2011) [3] J. Yan, et.al, Phys. Rev. B 83, 245122 (2011). [4] J. Yan, et.al, Phys. Rev. Lett. 106, 146803 (2011)   

In situ XAS of Pt monolayer model fuel cell catalysts: balance of na-nostructure and bimetallic interactions

The mechanism of the electrochemical oxygen reduction reaction (ORR) has been well understood based on DFT calculations, but there has been a lack of supporting experimental data, due to the difficulties of probing the electrocatalyst surface in situ. Our new approach using Pt monolayer model catalysts provides true surface sensitivity for - originally bulk sensitive - x-ray absorption spectroscopy (XAS) and, owing to the high resolution of the Bragg analyzer at SSRL beamline 6-2, allows for in situ detection of chemisorbed O and OH, whose stability can be used as a descriptor in predicting the activity of new ORR catalyst materials. Our ability to control the growth mode in the Pt/Rh(111) model system allows us to generate Pt nanostructures with highly different O affinities from identical starting materials.