The reaction pathway for shallow arsenic doping of silicon by methylarsenic acid molecules directly grafted on oxide-free, H-terminated Si(111) surfaces is unraveled combining Infrared absorption spectroscopy, X-ray Photoelectron Spectroscopy, Low Energy Ion Scattering and ab initio Molecular Dynamics simulations. The overall driving force is identified as a thermodynamic instability of As+5 in contact with silicon, which initiates a self-decomposition of chemisorbed methylarsenic molecules at ∼600 K. As the temperature is increased, the As-C bond breaks -- the weakest link of the adsorbed molecule -- with release of the organic ligand and a rearrangement from a monodentate to a bidentate bonding configuration. In this process, oxygen atoms evolve by partial desorption as H2O and partial incorporation into the surface Si atom backbonds. At ∼1050 K, diffusion of As into the sub-surface region of silicon is observed. There is no evidence for As desorption and no remaining C contamination.

Using first-principles calculations, we investigate lithium vacancy and interstitial defects in lithium phosphate (γ-Li3PO4) and in its interface with metallic Li. We find that γ-Li3PO4 is good electronic insulator with a wide band gap of 6 eV. The calculated formation energies of Li vacancies are higher than those of Li interstitials, which indicate that the ionic conductivity is determined by the migration of Li interstitial defects in bulk electrolyte. The Li vacancy-interstitial pair defect formation energy in the Li/γ-Li3PO4 interface is comparable to the sum of Li vacancy defect at the electrode and Li ion interstitial defect in the electrolyte. Our calculation indicates that the low ionic conductivity of Li/electrolyte interface is associated with the high Li ion defect formation energy. Our study provides some useful insights on Li defect formation and migration mechanisms at the electrode-electrolyte interface and, hence, a research direction for designing future Li-ion batteries.

We investigate the electronic structure of interstitial Li and Li vacancy in Li10GeP2S12 by first principles calculations. We find that the Li vacancy and interstitial Li+ ion do not introduce states in the band gap hence they do not deteriorate the electronic properties of Li10GeP2S12. The energy barrier for Li interstitial diffusion in Li10GeP2S12 is estimated to be 1.4 eV, which is much larger than that of the Li vacancy in Li10GeP2S12. This fact suggests that the ion conductivity arises from the migration of Li vacancy.

We investigate the electronic structure of interstitial Li and Li vacancy in Li7P3S11 by first principles calculations. We find that Li7P3S11 is a good insulator with a wide band gap of 3.5 eV. We find that the Li vacancy and interstitial Li+ ion do not introduce states in the band gap hence they do not deteriorate the electronic properties of Li7P3S11. The calculated formation energies of Li vacancies are much larger than those of Li interstitials, indicating that the ion conductivity may arise from the migration of interstitial Li.

Bimetallic Pd-Au nanoparticles have received much attention due to their potential applications in catalysis. We have developed a Pd-Au alloy potential based on Chen-Mobius lattice inversion method and applied it to the investigation of the melting of Pd-Au binary nanoparticles via molecular dynamics simulations. Our simulation results show the particle size dependence of the melting point and an enrichment of Au atoms to the surface near melting temperature.

We investigate the impact of various dopants (Na, Ag, Cd, Zn, Al, Ga, In, Tl, Ge, and Sn) on the electronic structure of Mg2Si by first principles calculations using a hybrid functional that does not need a band gap correction. We find that for Na and Ge in Mg2Si, the impurity-induced states do not affect the density of states at both edges of the valence band and the conduction band. Ag- and Sn affect slightly the density of states at the valence band edge, while Cd and Zn affect slightly the density of state at the conduction band edge. Al and In could modify significantly the density of states at the conduction band edge. Ga introduces states just at the bottom of the conduction band. Tl introduces states in the band gap. This study provides useful information on optimizing the thermoelectric efficiency of Mg2Si.

Metal-graphene contact is of critical significance in graphene-based nanoelectronics. There are two possible metal-graphene contact geometries: side-contact and end-contact. In this paper, we apply first-principles calculations to study metal-graphene end-contact for these three commonly used electrode metals (Ni, Pd and Ti) and find that they have distinctive stable end-contact geometries with graphene. Transport properties of these metal-graphene-metal (M-G-M) end-contact structures are investigated by density functional theory non-equilibrium Green’s function (DFT-NEGF) algorithm. The Transmission as a function of chemical potential (E-EF) shows asymmetric curves relative to the Fermi level. Transmission curves of Ni-G-Ni and Ti-G-Ti contact structures indicate that bulk graphene sheet is n-doped by Ni and Ti electrodes, but that of Pd-G-Pd shows p-doping of graphene by Pd electrode. The contact behaviors of these electrodes are consistent with experimental observations.

A study of the impact of surface preparation and post-deposition annealing on contact resistivity for sputtered Ni and Co contacts to thin film Bi2Te3 is presented. The contact resistance values obtained using the transfer length method (TLM) for Ni is compared to Co as a potential contact metal to Bi2Te3. Post-deposition annealing at 100°C on samples that were sputter cleaned reduces the contact resistivity to < 10-7 Ω-cm2 for both Ni and Co contacts to Bi2Te3. Co provided similar contact resistance values as Ni, but had better adhesion and less diffusion into the thermoelectric (TE) material, making it a suitable candidate for contact metallization to Bi2Te3 based devices.

The impact of interfacial oxygen content on the band offsets of GaAs:HfO2 interfaces was investigated using the density functional theory (DFT) method. Reference potential method was used to determine the band offsets. Moreover, GW correction was utilized to find more accurate value of the valence band edge of HfO2 and hence obtain more accurate band offsets. With gradually decreasing the interfacial O content from 100% to 30% (by changing O chemical potential corresponding to varying the growth condition), the valence band offset increases from 1.06 to 3.34 eV. It is found that this increase of the valence band offsets is inversely proportional to the charge loss of interfacial Ga atoms. Specifically, less charge loss of interfacial Ga induces less charge transfer from GaAs to HfO2 side. Consequently, the less charge loss of interfacial Ga essentially leads to an increase of the valence band offsets.

We investigated the impact of doping group III elements (Al, Ga, In and Tl) on the electronic structure of PbTe by first principles calculations. The impurity-induced defect level changes with respect to the charge state of the impurity. We find that among the four elements, Tl is the best candidate for the enhancement of thermoelectric efficiency, consistent with the experimental data.

The atomic and electronic structures of pyramidal model STM tips of transition metals (W, Rh, Pd, Ir and Pt) were investigated using density functional theory (DFT) method. The calculated density of states show that d electrons of the apex atoms in the M4 (M = W, Rh, Pd, Ir, Pt) model tips behave differently near the Fermi level, with the dz2 state being dominant only for W tip. The electronic structures of pyramid structures of W and Pd single-atom tips with larger sizes are studied and compared. The density of states of Pd apex atom and W apex atom show different occupation of d-bands leading to asymmetric density of states for Pd tip. The asymmetric tunneling currents measured by W and Pt-Ir STM tips are explained by the calculated electronic structures of W and Pd model tips.

We investigate the band offsets and stability for Ni/Bi2Te3 and Co/Bi2Te3 interfaces by first principles calculations. It is found that the surface termination strongly affects the band offsets. Ni and Co are found to form Ohmic contacts to Bi2Te3. The interface formation energies for Co/Bi2Te3 interfaces are much lower than those of Ni/Bi2Te3 interfaces. Our calculations are consistent with the experimental data.

Alloying has been one of the strategies to develop alternatives to Pt based CO oxidation catalyst. PdAu bimetallic alloy has recently been shown to have better reactivity and thermal stability toward CO oxidation for diesel engine applications as compared to pure metal catalysts. The key factor for low temperature light off in diesel engine catalysis is reactivity of alloy catalysts under CO environment, which in turn depends on the alloy surface composition and morphology. We explored the segregation processes in bimetallic Pd-Au alloy using first-principles calculations, assisted by a Monte-Carlo (MC) scheme that combines an improved Embedded Atom Method (EAM) and an atomistic treatment for adsorbed CO molecules for searching low energy states. Our simulation results show that PdAu surface changes from Au-rich to Pd-rich with increase in CO coverage up to 0.75 ML, beyond which additional CO adsorption is no longer favorable. A quantitative relationship between CO coverage and Pd concentrations on the surface is also revealed.

We investigated the HfO2:GaAs interface electronic structure and interface passivation by first principles calculations. The HfO2:GaAs interface of HfO2 terminated with four O atoms and GaAs terminated two Ga atoms is found to be the most energetically favorable. It is found that the interface states mainly arise from the interfacial charge mismatch, more specifically from the electron loss of interfacial As. Si or Ge as an interfacial passivating layer helps to maintain the charge of interfacial As and hence reduce the interface states.

We investigate the surface kinetics of Pt using the extended embedded-atom method, an extension of the embedded-atom method with additional degrees of freedom to include the nonbulk data from lower-coordinated systems as well as the bulk properties. The surface energies of the clean Pt (111) and Pt (100) surfaces are found to be 0.13 eV and 0.147 eV respectively, in excellent agreement with experiment. The Pt on Pt (111) adatom diffusion barrier is found to be 0.38 eV and predicted to be strongly strain-dependent, indicating that, in the compressive domain, adatoms are unstable and the diffusion barrier is lower; the nucleation occurs in the tensile domain. In addition, the dissociation barrier from the dimer configuration is found to be 0.82 eV. Therefore, we expect that atoms, once coalesced, are unlikely to dissociate into single adatoms. This essentially tells that by changing the applied strain, we can control the patterning of nanostructures on the metal surface.

First-principles calculations are performed for high-K gate dielectric materials using model bulk and interface systems. Detailed electronic structures and atomic configurations are investigated for transition metal (Hf and Zr) oxide, metal doped silicate bulk system and a model Si-silicate interface system. Pseudo polymorphs of metal oxides are investigated to elucidate the underlying driving mechanisms in microscopic configurations of metal oxides and silicates in amorphous structures. We studied energetics and electronic structure of metal oxide pseudo morph with varying oxygen coordination. Dielectric constants of metal oxide and silicate materials are also investigated using the density functional perturbation theory method implemented in the ABINIT code. Electronic and dielectric properties of silica interface layers between high-κ dielectric and Si substrate are investigated leading to a confirmation that 1 nm is the physical limit of gate oxide thickness. Furthermore silica interface layer is found to have small dielectric constant of 3.4∼3.9.

First-principles calculations are performed for high-K gate dielectric materials using model bulk and interface systems. Detailed electronic structures and atomic configurations are investigated for transition metal (Hf and Zr) oxide, metal doped silicate bulk system and a model Si-silicate interface system. Pseudo polymorphs of metal oxides are investigated to elucidate the underlying driving mechanisms in microscopic configurations of metal oxides and silicates in amorphous structures. We studied energetics and electronic structure of metal oxide pseudo morph with varying oxygen coordination. Dielectric constants of metal oxide and silicate materials are also investigated using the density functional perturbation theory method implemented in the ABINIT code. Electronic and dielectric properties of silica interface layers between high-κ dielectric and Si substrate are investigated leading to a confirmation that 1 nm is the physical limit of gate oxide thickness. Furthermore silica interface layer is found to have small dielectric constant of 3.4∼3.9.

The external chemical reactivity of graphene sheet, fullerenes and carbon nanotubes has been investigated. The total reaction energy is analyzed with several contributing terms and formulated as a function of the pyramidal angles of C atoms. We have determined the parameters for the formulae from ab initio simulation of graphene. We have applied them to predict hydrogenation energy of several nanotubes and C60, and demonstrated that the predicted total reaction energies are very close to the results of total energy pseudo-potential density functional theory calculations. This analysis can be used to predict the reaction energy and local bonding configuration of a reactant with diverse fullerenes and nanotubes within 0.1 eV accuracy.

Two different models for solid-state quantum bits have been investigated. Both are based on the nuclear spin of doped atoms in endo-fullerenes or bucky-onions. 1H or 31P have been tested as suitable dopant atoms because they have half nuclear spins. The thermal stability and electronic properties of the dopant atoms and the encapsulating cages have been examined with ab-initio pseudo potential density functional methods, and the results show that both models are suitable for single qubit applications.