Session P23: Computational Materials Discovery and Design - Defects and Interfaces

Computational design of inorganic-organic hybrid materials energy storage and conversion

Hybrid inorganic-organic materials are of interest for the design of new functional materials that combine the advantages of both organic and inorganic components to optimize properties and/or obtain new physical phenomena. In this talk, I will discuss our recent work using first-principles density functional theory to design nanostructured hybrid materials for energy storage and conversion applications. In particular, I will discuss the electronic, optical, thermal, and mechanical properties of a class of nanostructured hybrid materials based on layered transition metal phosphates, showing that these materials offer a highly tunable platform for the design of efficient, flexible photovoltaics and thermoelectrics. In addition to optimizing individual properties, I will also discuss exciting possibilities for using this platform for the design of materials with strong coupling between functionalities.   

Theoretical and Experimental Studies of Designed Molecular Interfaces for Improved Thermal Conductivity

Certain molecular structures such as carbon nanotubes (CNTs) can potentially improve the thermal conductivity of composite materials. However, their thermal boundary resistance is an obstacle to their effective implementation as a medium for heat flow. We are concerned with overcoming this Kapitza resistance with the aid of chemical functional groups. These functional groups will, in principal, eliminate phonon mismatch between our structures and their matrix. The result will maximize the transmission of thermal vibrations to and from their surrounding medium. We develop a method to predict the thermal properties of our functionalized materials through the calculation of vibrational normal modes and Green's functions. We show how the configuration of attached functional groups affect the samples' thermal conductivity ($ \kappa $) and attempt to find the arrangement in which $ \kappa $ is maximized. We will make explicit comparisons with thermal conductivity experiments done on nanocomposites of functionalized and pristine CNTs. We will discuss how the bonds connecting the functional groups to the CNT affects $ \kappa $. We compare these results to measurements on our particular synthesized materials and discuss how to better optimize their design.   

Modifying the Optoelectronic Properties of Rubrene by Strain

Rubrene crystals are promising organic electronic and optoelectronic materials due to their high charge carrier mobility. Recent studies have shown that the electronic properties of rubrene films can be tuned by substrate-induced strain, suggesting a new route towards the design of more efficient devices. Here, we present a first-principles density functional theory and many-body perturbation theory analysis of strain-induced changes to the mechanical, electronic, and optical properties of rubrene crystals. With an applied strain that is consistent with experiment, we predict changes of hole motilities in excellent agreement with electrical conductivity measurements. Furthermore, we predict that the optical absorption and nature of low-energy excitons within the crystal can be tuned by an applied strain as low as 1{\%}.   

First-principles studies of the electric-field effect on the band structure of trilayer graphenes

Electric-field effects on the electronic structure of trilayer graphene are investigated using the density functional theory in the generalized gradient approximation. Two different stacking orders, namely Bernal and rhombohedral, of trilayer graphene are considered. Our calculations reproduce the experimentally data on band gap opening in Bernal stacking and band overlap in rhombohedral trilayer graphene. In addition, we studied effects of charge doping using dual gate configurations. The size of band gap opening in Bernal trilayer graphene can be tuned by charge doping, and charge doping also causes an electron-hole asymmetry in the density of states. Furthermore, hole-doping can reopen a band gap in the band overlapping region of rhombohedral trilayer grapheme induced by electric fields, which contributes to an extra peak in the optical conductivity spectra.   

Design of transparent conductors and periodic two-dimensional electron gases without doping

The functionality of transparency plus conductivity plays an important role in renewable energy and information technologies, including applications such as solar cells, touch-screen sensors, and flat panel display. However, materials with such seemingly contraindicated properties are difficult to come by. The traditional strategy for designing bulk transparent conductors (TCs) starts from a wide-gap insulator and finds ways to make it conductive by extensive doping. We propose a different strategy [1] for TC design---starting with a metallic conductor and designing transparency by control of intrinsic interband transitions and intraband plasmonic frequency. We identified specific design principles for prototypical intrinsic TC classes and searched computationally for materials that satisfy them. The electron gases in the 3D intrinsic TCs demonstrate intriguing properties, such as periodic 2D electron gas regions with very high carrier density. We will discuss a more extended search of these functionalities, in parallel with stability and growability calculations. [1] X. Zhang, L. Zhang, J. D. Perkins, and A. Zunger, Phys. Rev. Lett. 115, 176602 (2015). Supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Grant No. DEFG02-13ER46959.   

Role of sp$^3$ Defect in Ordered Nanoporous Carbon

Schwarzite is considered an ideal model for nanoporous carbon and is energetically more stable than fullerene. However, carbon don't form well-ordered Schwarzite-type nanoporous material possibly due to kinetic arrests under pyrolytic conditions. We computationally discovered a new thermodynamically stable local defect in carbon sp$^2$ networks: an sp$^3$ carbon defect, which inspires new solutions to the problem. The defect is most stable in nanoporous carbon (i.e., networks with negative curvatures, known as Schwarzites) and its topological merit, carrying negative curvature, results in the design of new model structures of nanoporous materials (periodic, negatively-curved networks), and provides a handle of the negative curvature carrier in nanoporous carbon, and we propose a kinetics-dominated synthetic route to novel nanoporous carbon with long range order by controlling the sp$^3$ defect through sp$^3$ carbon atom injection or Si atom substitution, with the aid of first principle molecular dynamics simulation. Calculations also suggest the defect can be observed by Raman.   

Computational Discovery of a Novel Semiconductor: A Vacancy-Ordered Fe$_{\mathbf{1.5}}$TiSb Heusler Phase

Many full- and half-Heusler phase compounds are half-metallic ferromagnets, and are attractive for spintronic applications due to their relatively high Curie temperatures. However, while it is known that defects such as vacancies (on the X site of an X$_2$YZ Heusler phase) can lead to a loss of half-metallic character, their effect on the stability and order of these compounds has not been adequately explored. To address this shortcoming, we perform a binary cluster expansion (CE) of Fe and vacancies on the Fe sublattice of the Fe$_x$Vac$_{2-x}$TiSb Heusler compound. From our CE, we computationally predict the stability of a novel semiconductor phase with an interesting new structure type: $R3m$ spacegroup with composition Fe$_{1.5}$TiSb, i.e., between the full- and half-Heusler compositions. By comparing the electronic structure of all the competing structures at $x=1.5$, we find that the gap opened in the minority-spin channel due to vacancies strongly correlates with the stability of the structure. We study the effect of vacancies on the structural order in Fe$_{1.5}$TiSb by generating special quasi-random structures (SQSs) as approximations to the true disordered state, and find that the material undergoes an order-disorder transition at elevated temperatures of $\sim$1450~K.   

Doping and defects by design: Ga2O3

Density functional supercell calculations are widely employed to describe the defect physics in semiconductors and insulators. Due to a variety of challenges such as finite size effects for charged defects and the band gap error of DFT, results were often controversial in the past. With developments over the past decade, defect theory should hopefully be truly predictive, and be able to guide experimental efforts. The present work on n-type doping in Ga2O3 compares different potential doping routes via anion-site doping with F, and cation site doping with group IV elements (C, Si, Ge, Sn). The study addresses dopant solubility, electrical activity, and compensation by native defects, including non-equilibrium effects due to supersaturated dopant concentrations and the mechanism of dopant-defect pair formation.   

Understanding the Impact of Point Defects on the Optoelectronic Properties of Gallium Nitride from First-Principles

Gallium nitride (GaN) and related alloys form a class of wide bandgap semiconductors that have broad applications as components in optoelectronic devices; in particular, power electronics and blue and ultraviolet optical devices. Nitride films grow with high defect densities, and understanding the relationship between structural defects and optoelectronic function will be central to the design of new high-performance materials. Here, we take a first-principles density functional theory (DFT) and many-body perturbation theory (MBPT) approach to quantify the influence of defects on the electronic and optical properties of GaN. We predict, as expected, that introduction of a N or Ga vacancy results in several energetically favorable charged states within bulk GaN; these energetically favorable defects result in a significant modification of the quasiparticle and excitonic properties of GaN. We will discuss the implications of defect-induced-states for the electron transport and absorption properties of GaN.   

Prediction of direct band gap silicon superlattices with dipole-allowed optical transition

While cubic diamond silicon (c-Si) is an important element in electronic devices, it has poor optical properties owing to its indirect gap nature, thereby limiting its applications to optoelectronic devices. Here, we report Si superlattice structures which are computationally designed to possess direct band gaps and excellent optical properties. The computational approach adopts density functional calculations and conformational space annealing for global optimization. The Si superlattices, which consist of alternating stacks of Si(111) layers and a defective layer with Seiwatz chains, have either direct or quasi-direct band gaps depending on the details of attacking layers. The photovoltaic efficiencies are calculated by solving Bethe-Salpeter equation together with quasiparticle G0W0 calculations. The strong direct optical transition is attributed to the overlap of the valence and conduction band edge states in the interface region. Our Si superlattices exhibit high thermal stability, with the energies lower by an order of magnitude than those of the previously reported Si allotropes. We discuss a possible route to the synthesis of the superlattices through wafer bonding.   

From facets to facets: how does work function vary over a gold nanocluster?

Owing to their potential applications in catalysis, gold nanoclusters are a focus of intense research. The work function $\Phi $, which can be measured using photoemission spectroscopy is a key parameter used to characterize the catalytic performance of the cluster. $\Phi $ is determined by the difference between the electrostatic potential just outside the metal surface and the Fermi energy of the cluster. We use a relativistic version of the real space first-principles code PARSEC to compute the work function of gold nanoclusters with dimensions on the order of a nanometer, which is similar in size to those used in experiment. We illustrate how the work function depends on the surface orientation of the nanocluster facets and compare our results with available experimental data We acknowledge supports from SciDAC program, Department of Energy, Office of Science, Advanced Scientific Computing Research and Basic Energy Sciences grant DE-SC0008877 for work on algorithms. Two of us (JRC and JS-C) acknowledge support for the work on nanostructures from grant from the U.S. Department of Energy: DE-FG02-06ER46286.   

Gap engineering using Hellmann-Feynmann forces: method and applications

Materials with optimized band gap are needed in many specialized applications. In this talk, we demonstrate that Hellmann-Feynman forces associated with the gap states can be used to find atomic coordinates that yield desired electronic density of states. Using tight-binding models, we show that this approach may be used to arrive at electronically designed models of amorphous silicon and carbon. We provide a simple recipe to include {\it a priori} electronic information in the formation of computer models of materials, and prove that this information may have profound structural consequences. We'll briefly discuss implementation of the method in ab-initio molecular dynamics simulations and highlight the latest feats of the method ranging from modeling amorphous semi-conducting materials to understanding the structure and properties of memory materials.   

Bayesian cluster expansion with lattice parameter dependence for studying surface alloys

The Bayesian cluster expansion approach has proven to be an efficient method for predicting the structure and properties of materials with substitutional disorder. It is particularly effective for low-symmetry systems such as nanoparticles and surfaces. However for surfaces of solid solutions, the lattice parameter of the surface, and hence the interactions among near-surface atoms, varies with the composition of the underlying bulk material. We demonstrate that surfaces under a variety of strains can be used to train a single cluster expansion that predicts properties as a function of atomic order and surface strain. We have applied this method to study Ni-Pt alloys with up to a monolayer of adsorbed oxygen, an important class of catalysts for the oxygen reduction reaction. Through Monte Carlo simulations, we are able to determine how the structure and properties of these surfaces vary as a function of temperature, composition, chemical potential, and surface strain, enabling both the identification of thermodynamically stable surface structures and the rational design of Pt-Ni surfaces with high catalytic activity.