Session L20: Focus Session: Computational Design of New Materials -- Nanostructure Design

Global structure optimization of complex systems

Finding the global minimum structure of complex condensed matter systems is a long-standing challenge in computational physics. Genetic algorithm, basin-hopping or simulated annealing are popular methods to explore the configurational space. Here we applied these methods to periodical systems such as crystalline solids and alloys, metal on semiconductor surfaces and interface structures. Novel structures can be revealed in complex and multicomponent systems. The efficiency of different search strategies is also discussed.   

Molecular Dynamic Studies on Controlled Synthesis of Carbon Nanotube Junctions

We report molecular dynamic studies on synthesis of Carbon Nanotube (CNT) junctions with various terminals based on a self-assembling process from two tailored graphene nanoribbons (GNRs) either with perfect or irregular tailoring. Based on the recent experimental advancement on tailoring GNRs, our studies provide an approach to controlled synthesize CNT junctions with desired functionality. Quantum transport simulations are carried out to confirm the functionality of the synthesized CNT junctions.   

Quantum Boundary Effect on Electronic Structures of Nanomaterials

Nanosciences are, to a large extent, related to the quantum size effect. Nanomaterials with similar characteristic sizes, however, do not necessarily exhibit similar quantum effects, e.g., the shape of the nanomaterials is known to affect their electronic properties also. Other than the size and shape, the boundary of a nanostructure, from which the quantum confinement arises, is another key factor that affects the expressions of the quantum effects. Here, we discuss a quantum boundary effect on the electronic structures of nanomaterials. We argue that different boundary atomic structures of nanomaterials of the same characteristic size and shape could be energetically degenerate, but give distinct electronic properties, e.g., band gaps. Such an effect could pose a challenge on preparing nanomaterials with expected electronic properties. Our discussion will be based on first-principles electronic structure calculations.   

Design of Nanostructured Materials for Electronic, Thermoelectric, and Optoelectronic Applications

There is a large interest in tailoring semiconductor nanostructures to display desired electronic, thermal, or optical properties. Nanostructure dimensions are typically between a few and a few hundred nanometers, so they are too large to treat using atomistic methods, yet too small for continuum techniques. Their characteristics are strongly influenced by the properties of the interfaces, such as roughness, surface defects, or adsorbed charges. A major challenge in predicting the properties of nanostructures lies precisely in capturing the complex interplay between the confined particle states and the surface condition. I will review techniques typically used to analyze and predict the electronic, thermal, and optoelectronic properties of semiconductor nanostructures, with particular focus on the verstatility that the ensemble Monte Carlo technique offers in simulating these different transport phenomena. In particular, I will present our results on electronic and thermal transport in nanowires, based on the coupled electronic and thermal ensemble Monte Carlo simulation with confined electron and phonon dispersions. We will take a close look into boundary scattering of electrons and phonons, and features such as phonon localization, and discuss where atomistic simulations naturally come to play to aid in the desription of interfaces. We will then look into the design on strain superlattices for thermoelectric applications and the design of nanowire interfaces for tailoring thermal conduction. We will also examine how efficient transport simulation aids in the design of quantum cascade lasers. Multivalley ensemble Monte Carlo simulation, combined with k.p bandstructrue calculations and the dielectric continuum model, captures the transport of heat and charge in midinfrared quantum cascade lasers, and helps pinpoint the flaws of a laser design and directions for performance improvement through minimized leakage. We also discuss some promising new avenues, such as the simulaiton of high-frequency and transient phenomena in nanostructures using a combination of full electrodynamics together with the transport of charge and heat, from low- temperature ballistic to room-temperature diffusive transport regimes.   

Tailoring the Electronic Structure by Alloying: Ag$_{n}$Cu$_{34-n}$ Nanoparticle Family

We report results of first principles calculations of the electronic structure of the Ag$_{n}$Cu$_{34-n }$bi-metallic nanoparticle family where n=0, 1, {\ldots}, 34. We find that alloying of the pure Ag cluster with a few Cu atoms displays substantial changes in the electronic structure but the reverse is not the case when few Ag atoms are substituted in pure Cu clusters. We find that local environment control the length and the strength of the Cu-Ag bonds. The Cu atoms, which form the core display shortened bond length and present above 1 eV shift in their d-band center. The HOMO-LUMO gap for the set of nanoalloys falls in three regions: 0.19 eV to 0.31 eV, 0.40 eV to 0.57 eV, and 0.73 eV to 0.88 eV. For several nanoparticles slight change in composition may thus lead to a change of about 600 meV in the gap. The highest gap is found for the most symmetric nanoparticle (Ag$_{17}$Cu$_{17})$, while the lowest is for Ag$_{29}$Cu$_{5}$. We present a systematic analysis of the changes in Cu-Ag, Ag-Ag and Cu-Cu bond lengths and hybridization with composition to understand their effect on the HOMO-LUMO gap and other characteristics of the nanoparticles.   

\textit{Ab initio} calculations of pre-exponential factors for the diffusion of CO on Ag(001): importance of the full phonon dispersion.

Knowledge of factors that determine the diffusivity of CO on catalytic surfaces is of utmost importance for understanding why some surfaces render higher reaction rates (e.g. CO oxidation) than others. A rigorous calculation of the diffusion prefactor ($\nu _{0})$ is not only the first approximation one should target to analyze diffusivity but also a prerequisite to adequately take into account anharmonic processes since calculation of both $\nu _{0}$ and anharmonic processes rates require the full dispersion of all phonons in the system as input information. In this work, we obtain from first principles calculations the diffusion path of CO on Ag(001), the corresponding energy barrier, and diffusion prefactor $\nu _{0}$. The latter has been calculated via the total phonon density of states in the surface Brillouin zone (SBZ) from density functional perturbation theory calculations. On comparing the results for $\nu _{0}$ obtained using the full phonon dispersion curve with those confined only to the adsorbate modes on a frozen substrate point to significant differences and to the importance of having knowledge of the full vibrational dynamics of the system. Work supported in part by NSF under grant CHE-0741423.   

Towards Nano-Materials with Precise Control over Properties via Cluster-Assemblies

One pathway towards nanomaterials with controllable band gaps is to assemble solids where atomic clusters serve as building blocks, because clusters' electronic structures vary with size, composition, and the charged state. To study the role of architecture in cluster assemblies, we synthesized multiple architectures of As$_{7}$$^{3-}$ clusters through controlling the counter-cations. Optical measurements revealed that the band gaps vary from 1.1-2.1 eV, even though the assemblies are constructed from identical cluster building blocks. First principles theoretical studies reveal that the variation is a result of altering the LUMO levels by changing the counter-cations. Additional variation in the gap is found by covalently linking the clusters with species of varying electronegativity to alter the degree of charge transfer. The findings offer a novel protocol for synthesis of nanoassemblies with tunable electronic properties.   

A mechanistic density functional study of partial oxidation of propene on supported subnanometer silver clusters

Density functional theory (DFT) calculations were carried out to investigate the fundamental chemical mechanisms underlying the selective oxidation of propene (CH$_{3}$--CH=CH$_{2}$) to propylene oxide (CH$_{3}$--CH--CH$_{2}$O) or acrolein (CH$_{2}$=CH--CHO). Silver which is a known catalyst for these reactions, is modeled as a silver trimer supported on $\theta$-Al$_{2}$O$_{3}$ surface to investigate reaction mechanisms. Each reaction step starts with O$_{2}$ dissociation on interface of Ag$_{3}$ and alumina with an activation barrier that was found to be significantly smaller than what was reported on an Ag(111) surface. Propylene oxide formation was achieved through oxametallocyle formation with a much small barrier of 0.12 eV compared to 0.70 eV on Ag(111). We will also show how acrolein formation can be triggered by the abstraction of first H atom with essentially no apparent barrier on a Ag$_{3}$ cluster. Finally, we rationalize our results in comparison to experiments that show an enormous increase in reactivity for propylene epoxidation on subnanometer silver cluster. We will also discuss the preference of Ag trimers over tetramers for O$_{2}$ dissociation and how the microscopic understanding of such information can help better design new catalysts.   

Molecular dynamics study of phase transition in substrate-supported nanoparticles of bcc metals

The phase transition between liquid and solid phases of substrate-supported nanoparticles of bcc metals with size ranging from 2000 to 31200 atoms was investigated using a molecular dynamics simulation. The effect of the interaction energy between the nanoparticle and the substrate on the contact angle, melting point and nucleation temperature was focused on. Unidirectional solidification and inward melting after surface melting were observed in the substrate-supported nanoparticles during cooling and heating, respectively. The depression of the melting point from the bulk melting point was proportional to the inverse of the effective radius of the substrate-supported nanoparticles [1]. The gradient of proportionality increased with decreasing contact angle and deviated from that of freestanding nanoparticles [2]. On the other hand, the undercooling temperature for solidification decreased with decreasing contact angle, which agrees with the theory of heterogeneous nucleation. [1] Y. Shibuta, T. Suzuki, Phys. Chem. Chem. Phys., 2010, DOI:10.1039/b919869e. [2] Y. Shibuta, T. Suzuki, J. Chem. Phys. 129 (2008) 144102.   

Enhanced Static Approximations for Efficient Calculation of Electronic Excitation Energies

The GW method for the calculation of excited-state energies has been proven accurate for many physical systems. However, the unfavorable scaling with the size of the system has severely limited its applications. In particular, quantitative prediction of frontier electronic energy levels, crucial in the design of complex and nanostructured materials, has not been feasible. While the static COHSEX approximation significantly simplifies the calculations, the quantitative results are substantially less accurate. Analysis shows that much of the error in the COHSEX approximation traces to the overestimation of short wavelength contributions to the Coulomb-hole term of the self-energy. We have implemented modified static methods that compensate for most of this error. Results will be shown for a series of material applications.   

From Microstructures to Predict Properties of Materials

Understanding the precise and fundamental manner in which materials structures (nanostructures or microstructures) and their evolution influences properties and service lifetimes of advanced materials profoundly impacts material design and today materials design plays an increasingly important r\^{o}le in many engineering applications. Linking structures to properties and predicting properties of materials is fundamental step for materials design. First, a framework of applications of multiscale modeling to property prediction of advanced materials will be briefly presented. As an example, a methodology will be shown to link micro-scale to the continuum scale, integrating microstructure modeling with the large Thermo-Calc$^{\mbox{{\textregistered}} }$ database. This paradigm was successfully applied to the case of Fe-12Ni-6Mn maraging steel. Next, methodology for integrating first-principle calculation into simulations of microstructure evolution will be reviewed. Our methods are sufficiently reliable to permit control and fabrication of quantum-dots structures, nanocrystals, and particle-reinforced nanocomposites, as well as assist in the predictive behavior of macro-scale colloids, aerosols, and other soft matter systems.   

First-Principles Studies of the Electronic and Vibrational Properties of \textit{sila-}Diamondoids

Diamondoids are cage-like, stable, saturated hydrocarbons, which possess a~rigid carbon-framework that is superimposable upon the diamond crystal structure. These carbon-based diamondoids could be important building blocks in a variety of applications in biochemistry and nanotechnology. While the chemistry of silicon is not as robust as that of carbon, silicon-based analogues of diamondoids such as \textit{sila}-adamantane (Si$_{10}$H$_{16})$ do exist and there is reason to believe that larger sila-diamondoids might either be synthesized in the lab one day or observed experimentally in chemical vapor deposition (CVD) experiments. In this work we show that density-functional theory (DFT) can accurately compute the electronic, structural, and vibrational properties of a variety of lower and medium-order \textit{sila-}diamondoids. We believe that DFT is an important tool that will assist experimentalists in identifying more complicated \textit{sila}-diamondoids that either may be synthesized or already may exist as important intermediates in CVD experiments to grow silicon surfaces.