Search

Nuclear quantum effects (NQEs) are highly important for understanding a host of kinetic processes that occur with the participation of H (e.g., H adsorption, diffusion, permeation, and trapping in materials). In this paper, ab initio path integral molecular dynamics simulations were used to investigate NQEs on the lattice diffusion of H in common face-centered cubic (fcc) metals such as Al, Ag, and Cu over a wide temperature range of 75--1200 K (75--900 K for Al). We determined that the dependence of H diffusivities on temperature in Ag and Cu has a ``reversed-S'' shape on Arrhenius plots, as confirmed for fcc Pd in our recent study [H. Kimizuka et al., Phys. Rev. B 100, 024104 (2019)]. This result illustrates that the phenomenon is common in many fcc metals in which H atoms prefer to occupy octahedral sites. On the other hand, in the case of Al, in which H atoms prefer to occupy tetrahedral sites, the dependence of H diffusivities on temperature exhibits a familiar ``C'' shape. Such counterintuitive behavior is ascribed to differences in the dependence on temperature of the activation barriers for H migration between both types of fcc metals; this is due to the NQEs involving a competition between deceleration of H migration, which becomes effective at high temperatures because of zero-point vibrations, and acceleration of H migration, which becomes effective at low temperatures because of quantum tunneling. The dominance of the two mechanisms is determined by the coupling of the NQEs and the site preference of H depending on the metal. This finding has important implications for the interpretation of kinetic processes involving the crossover from classical to quantum behavior of H atoms jumping between different types of interstitial sites.

Motivated by the absence of experimental superconductivity in the metallic $\mathit{Pm}\overline{3}\mathit{n}$ phase of ${\mathrm{AlH}}_{3}$ despite the predictions, we reanalyze its vibrational and superconducting properties at pressures $P\ensuremath{\ge}99$ GPa making use of first-principles techniques. In our calculations based on the self-consistent harmonic approximation method that treats anharmonicity beyond perturbation theory, we predict a strong anharmonic correction to the phonon spectra and demonstrate that the superconducting critical temperatures predicted in previous calculations based on the harmonic approximation are strongly suppressed by anharmonicity. The electron-phonon coupling concentrates on the lowest-energy hydrogen-character optical modes at the X point of the Brillouin zone. As a consequence of the strong anharmonic enhancement of their frequency, the electron-phonon coupling is suppressed by at least 30%. The suppression in $\ensuremath{\lambda}$ makes ${T}_{c}$ smaller than 4.2 K above 120 GPa, which is well consistent with the experimental evidence. Our results underline that metal hydrides with hydrogen atoms in interstitial sites are subject to huge anharmonic effects.

Half-Heusler semiconductors satisfying 18-electron rule typically display promising characteristics for thermoelectric applications. A persistent inconsistency between the type of charge carriers in some of these alloys as obtained from experiment and theory however casts serious doubt on the computational prediction of new and efficient half-Heusler alloys. To gain insights into the origin of this disparity, we have investigated the effect of intrinsic point defects on the electronic structure of four frequently studied half-Heusler alloys of the form XYZ with Y being Ni or Co. Using state-of-the-art ab initio calculations, our study reveals that interstitial Ni and Co are energetically most stable point defects in these alloys. Remarkably, interstitial defect modifies the location of the Fermi level inside the band gap as well as the value of the band gap, thereby bringing in close agreement with the corresponding experimental result. This work thus highlights the decisive role played by interstitial defects in thermoelectric half-Heusler alloys, which may open a new avenue for deliberately utilizing these defects as a strategy for tailoring electronic structure and hence the corresponding thermoelectric properties.

The occurrence of the Kondo effect in the ultimate two-dimensional (2D) case, when the thickness of the conducting film is reduced down to the atomic-scale limit, was explored taking the Tl-adsorbed Au/Si(111)$\sqrt{3}\ifmmode\times\else\texttimes\fi{}\sqrt{3}$ reconstruction with embedded Cr atoms as a prototype system. The reconstruction is essentially an atomic-layer film, which resides on a semiconductor Si(111) substrate and demonstrates conductivity of a metallic type, at least down to 1.8 K. The tight binding of Cr atoms in the interstitial sites surrounded by six Si atoms prevents them from clustering. The Kondo effect in this system was investigated experimentally using low-temperature scanning tunneling spectroscopy and electronic transport measurements. Both techniques yield a Kondo temperature of $\ensuremath{\sim}70--80\phantom{\rule{0.28em}{0ex}}\mathrm{K}$. A substantial increase of the sample resistivity caused by adding Cr atoms is believed to be a specific feature of the extremely thin metallic films, where small amounts of magnetic impurities can effectively block off the narrow conduction channel.

We apply first-principles calculations to metastable B-doped Fe-Ni compounds to compute their magnetic properties. We focus on crystal structures with tetragonal and orthorhombic lattices. Boron atoms, doped at interstitial sites, help in stabilizing noncubic Fe-Ni structures. At the same time, the dopants improve the magnetic properties, such as magnetocrystalline anisotropy, compared to FeNi alloys in the tetragonal $L{1}_{0}$-ordered structures. The calculated magnetocrystalline anisotropy constants of $1--2 \mathrm{MJ}/{\mathrm{m}}^{3}$ along with sufficient magnetic polarization saturation $(\ensuremath{\ge}1 \mathrm{T})$ show the potential of our metastable Fe-Ni-B compounds for replacing rare-earth-based magnets in permanent-magnet applications.

A single-phase sample of $\mathrm{Ta}{\mathrm{H}}_{2.2(1)}$ with a hexagonal close-packed metal lattice (hcp; space group $P{6}_{3}/mmc$) was synthesized under a hydrogen pressure of 9 GPa and a temperature of 100 \ifmmode^\circ\else\textdegree\fi{}C; quenched to the liquid nitrogen temperature; recovered to ambient pressure and studied by neutron diffraction (ND) and inelastic neutron scattering (INS). The ND study showed that hydrogen atoms occupied one half of the tetrahedral (T) and all octahedral (O) interstitial sites in the hcp lattice of Ta atoms. The arrangement of the H atoms over the T sites was proven to be ordered, which lowered the symmetry of the full crystal structure of the dihydride to $P{6}_{3}mc$. Due to the resulting asymmetry in the local environment of the O sites, the H atoms were considerably displaced from the centers of these sites along the $z$ axis, away from the H atoms occupying the neighboring T sites. The INS study demonstrated that the potential wells for H atoms at both the T and O sites are highly anharmonic and anisotropic. The potential wells at the O sites are softer along the $z$ axis than in the $x, y$ plane, while the T sites show opposite anisotropy.

The aim of this work is to clarify the scintillation-yield enhancement in $\mathrm{Lu}\mathrm{YAG}:\mathrm{Pr}$ scintillators obtained by $\mathrm{Li}$ codoping via integrated study of the valence state of activators, the preferential site occupancy of $\mathrm{Li}$ codopants, and defect structures from experimental and theoretical insights. With $\mathrm{Li}$ codoping, the light yield and energy resolution of $10\ifmmode\times\else\texttimes\fi{}10\ifmmode\times\else\texttimes\fi{}10\phantom{\rule{0.1em}{0ex}}{\mathrm{mm}}^{3}$ $\mathrm{Lu}\mathrm{YAG}:\mathrm{Pr}$ samples are improved from 15 600 to 24 800 photons/MeV, and 5.3 to 4.3% at 662 keV, respectively. The optical absorption spectra indicate that $\mathrm{Li}$ codoping does not induce conversion of stable ${\mathrm{Pr}}^{3+}$ to ${\mathrm{Pr}}^{4+}$ in $\mathrm{Lu}\mathrm{YAG}:\mathrm{Pr}$ single crystals. Based on the formation energies of substitutional and interstitial $\mathrm{Li}$ sites using density-functional-theory (DFT) calculations and the ${}^{7}\mathrm{Li}$ nuclear magnetic resonance results, it is shown that the $\mathrm{Li}$ ions prefer to dominantly occupy the fourfold coordinated interstitial sites and fourfold coordinated $\mathrm{Al}$ sites. The systematic analysis of thermoluminescence glow curves, positron annihilation lifetime spectroscopies, and defect formation energies derived from DFT calculations reveals that the concentration of isolated $\mathrm{Lu}$ and $\mathrm{Al}$ vacancies as dominant acceptor defects is reduced by $\mathrm{Li}$ codoping, whilst the shallow ${\mathrm{Li}}_{i}$ interstitial defects and the deep ${V}_{O}$ oxygen vacancies are introduced simultaneously. We propose that the lowering of hole trapping at defects resulting from $\mathrm{Li}$ codoping contributes to the scintillation-yield enhancement.

The ionic materials ${\mathrm{Li}}_{4}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ and ${\mathrm{Na}}_{4}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ are both based on the same building blocks of the dimer ions (${\mathrm{P}}_{2}{\mathrm{S}}_{6}{)}^{4\ensuremath{-}}$. Motivated by new experimental structural and ion conductivity studies, we computationally examine this family of materials, finding ${\mathrm{Na}}_{4}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ and its modification ${\mathrm{Li}}_{2}{\mathrm{Na}}_{2}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ to be promising Na ion electrolytes. Using first-principles calculations based on density functional theory and density functional perturbation theory within the harmonic phonon approximation, we show that vibrational effects provide nontrivial contributions to the structural stabilization of these materials. Computed nonresonant Raman phonon spectra and temperature-dependent ionic conductivity for ${\mathrm{Na}}_{4}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ are both found to be in reasonable agreement with experiment. First-principles analysis of ionic conductivity in both ${\mathrm{Na}}_{4}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ and ${\mathrm{Li}}_{2}{\mathrm{Na}}_{2}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ indicates that Na ions move primarily within the interlayer region between the (${\mathrm{P}}_{2}{\mathrm{S}}_{6}{)}^{4\ensuremath{-}}$ layers, efficiently proceeding via direct or indirect hops between vacancy sites, with indirect processes involving intermediate interstitial sites.

Molybdenum contamination of silicon can have serious detrimental consequences for the efficiency of solar cells, raising durability concerns for novel solar cell designs that utilize $\mathrm{Mo}{\mathrm{O}}_{3}$ in contact with Si. Density functional theory simulations of Mo defects in Si revealed that Mo is preferentially accommodated in tetrahedrally coordinated interstitial sites and that the contamination may reach a sufficiently high concentration to cause a 20% relative solar cell efficiency degradation if processing steps are performed between 950 and 1300 K. The formation energy of the most energetically favored Mo defect in Si has a minimum value of 1.58 eV at the valence band maximum and a maximum of 2.10 eV at higher Fermi levels, indicating that higher Mo defect concentrations may occur in $p$-type Si than intrinsic or $n$-type Si. The diffusion processes for Mo in Si were investigated, and it was identified that interstitial diffusion dominates over a vacancy-mediated mechanism under all equilibrium conditions. Migration barriers were calculated to be 2.29 eV for charge neutral and 2.03 eV for charge +1 defects, occurring under $n$-type and $p$-type doping, respectively, indicating that Mo diffusion is faster in $p$-type Si, and hence potentially more effectively gettered than it would be in $n$-type Si.

We investigated the influence of oxygen over-stoichiometry on apical oxygen disorder and magnetic correlations in Nd2NiO4+d (d~0.11) in the temperature range of 2-300 K by means of synchrotron x-ray powder diffraction, neutron single crystal and powder diffraction studies, combined with macroscopic magnetic measurements. In the investigated temperature range, the compound crystalizes in a tetragonal commensurate structure with the P42/ncm space group with excess oxygen atoms occupy the 4b (3/4 1/4 1/4) interstitial sites, coordinated by four apical oxygen atoms. Large and anisotropic thermal displacement parameters are found for equatorial and apical oxygen atoms, which are strongly reduced on an absolute scale compared to the Nd2NiO4.23 phase. Maximum Entropy analysis of the neutron single crystal diffraction data uncovered anharmonic contributions to the displacement parameters of the apical oxygen atoms, toward the nearest vacant 4b interstitial site, related to the phonon assisted oxygen diffusion mechanism. Macroscopic magnetization measurements and neutron powder diffraction studies reveal long-range antiferromagnetic ordering of the Ni-sublattice at TN ~ 53 K with a weak ferromagnetic component along the c-axis, while the long-range magnetic ordering of the Nd-sublattice occurs below 10 K. Temperature dependent neutron diffraction patterns show the appearance of a commensurate magnetic order at TN with the propagation vector k = (100) and the emergence of an additional incommensurate phase below 30 K, while both phases coexist at 2 K. The commensurate magnetic structure is best described by the P42/nc`m` Shubnikov space group. Refined magnetic moments of the Ni and Nd-sites at 2 K are 1.144(76) muB and 1.632(52) muB respectively. A possible origin of the incommensurate phase is discussed and a tentative magnetic phase diagram is proposed.

Ytterbium dihydride (${\mathrm{YbH}}_{2}$) shows a well-known transition at $\ensuremath{\approx}16\phantom{\rule{0.16em}{0ex}}\mathrm{GPa}$ from a ${\mathrm{CaH}}_{2}$-type structure to a high-pressure (high-$P$) phase with Yb at hcp sites and unknown H positions. Here, we report its complete structure determination by neutron diffraction at 34 GPa. Hydrogen (deuterium) is located at $2a$ and $2d$ positions of space group $P{6}_{3}/mmc$, thus forming a high-symmetry ``collapsed'' close-packed lattice. The transition is sluggish and can be seen as a transfer of 1/2 of the hydrogen atoms from strongly corrugated H layers to interstitial sites of the Yb lattice. We demonstrate by first-principles calculations that the transition is related to a change from a completely filled $f$-electron configuration to a fractional $f$-hole ($\ensuremath{\approx}0.25\phantom{\rule{0.16em}{0ex}}\mathrm{h}$) occupation in the high-$P$ phase. The $f\ensuremath{\rightarrow}d$ charge transfer closes the gap at the transition and leads to a metallic ground state with a sizable electron-phonon interaction involving out-of-plane vibrational modes of interstitial hydrogen.

Passivation of barium disilicide ($\mathrm{BaS}{\mathrm{i}}_{2}$) films is very important for their use in solar cell applications. In this paper, we demonstrated the effect of hydrogen (H) passivation on both the photoresponsivity and minority-carrier lifetime of $\mathrm{BaS}{\mathrm{i}}_{2}$ epitaxial films grown by molecular beam epitaxy. First, we examined the growth conditions of a 3-nm-thick hydrogenated amorphous silicon (a-Si) capping layer formed on a 500-nm-thick $\mathrm{BaS}{\mathrm{i}}_{2}$ film and found that an H supply duration (${t}_{\mathrm{a}\text{\ensuremath{-}}\mathrm{Si}:\mathrm{H}}$) of 15 min at a substrate temperature of 180 \ifmmode^\circ\else\textdegree\fi{}C sizably enhanced the photoresponsivity of the $\mathrm{BaS}{\mathrm{i}}_{2}$ film. We next supplied atomic H to $\mathrm{BaS}{\mathrm{i}}_{2}$ epitaxial films at 580 \ifmmode^\circ\else\textdegree\fi{}C and changed supply duration (${t}_{\mathrm{BaSi};\mathrm{H}}$) in the range of 1--30 min, followed by capping with an a-Si layer. The photoresponsivity of the films changed considerably depending on ${t}_{\mathrm{BaSi};\mathrm{H}}$ and reached a maximum of 2.5 A/W at a wavelength of 800 nm for the sample passivated for ${t}_{\mathrm{BaSi};\mathrm{H}}=15$ min under a bias voltage of 0.3 V applied to the front-surface indium-tin-oxide electrode with respect to the back-surface aluminum electrode. This photoresponsivity is approximately one order of magnitude higher than the highest value previously reported for $\mathrm{BaS}{\mathrm{i}}_{2}$. Microwave photoconductivity decay measurements revealed that the minority-carrier lifetime of the $\mathrm{BaS}{\mathrm{i}}_{2}$ film with the highest photoresponsivity was 14 \ensuremath{\mu}s, equivalent to its bulk carrier lifetime ever reported. We performed theoretical analyses based on a rate equation including several recombination mechanisms and reproduced the experimentally obtained decay curves. We also calculated the total density of states of $\mathrm{BaS}{\mathrm{i}}_{2}$ by ab initio studies when one Si vacancy existed in a unit cell and one, two, and three H atoms occupied Si vacancy or interstitial sites. A Si vacancy caused a localized state with two energy bands to appear close to the middle of the band gap. In certain cases, H passivation of the Si dangling bonds can markedly decrease trap concentration. From both experimental and theoretical viewpoints, we conclude that an atomic H supply is beneficial for $\mathrm{BaS}{\mathrm{i}}_{2}$ solar cells.

Inelastic neutron scattering was performed for an iron-based superconductor ${\mathrm{LaFeAsO}}_{0.9}{\mathrm{D}}_{0.1}$, where most of D (deuterium) replaces oxygen, while a tiny amount goes into the interstitial sites. By first-principles calculation, we characterize the interstitial sites for D (and for H slightly mixed) with four equivalent potential minima. Below the superconducting transition temperature ${T}_{\mathrm{c}}=26\phantom{\rule{0.16em}{0ex}}\mathrm{K}$, excitations emerge in the range 5--15 meV, while they are absent in the reference system ${\mathrm{LaFeAsO}}_{0.9}{\mathrm{F}}_{0.1}$. The strong excitations at 14.5 and 11.1 meV broaden rapidly around 15 and 20 K, respectively, where each energy becomes comparable to twice the superconducting gap. The strong excitations are ascribed to a quantum rattling, or a band motion of hydrogen, which arises only if the number of potential minima is larger than two.

For environmental reasons there is a need for alternative Cd-free buffer materials in Cu(In,Ga)(S,${\mathrm{Se})}_{2}$ (CIGSSe) based thin film solar cells. In this context, $\ensuremath{\beta}\ensuremath{-}{\mathrm{In}}_{2}{\mathrm{S}}_{3}$ is one candidate material, whose optoelectronic properties can be affected by the presence of impurities. In this study, we investigate the impact of O and Cl impurities on the electronic and optical behavior of $\ensuremath{\beta}\ensuremath{-}{\mathrm{In}}_{2}{\mathrm{S}}_{3}$ by means of electronic structure calculations within density functional theory using hybrid functionals. We find that $\ensuremath{\beta}\ensuremath{-}{\mathrm{In}}_{2}{\mathrm{S}}_{3}$ is thermodynamically stable being in contact with both O and Cl reservoirs. Furthermore, we present evidence that O on interstitial sites (${\mathrm{O}}_{\mathrm{i}}$) and Cl on $8c$ In sites (${\mathrm{Cl}}_{\mathrm{In}}$) cause low-temperature persistent electron photoconductivity. At room temperature, defect levels associated with Cl on S sites (${\mathrm{Cl}}_{\mathrm{S}}, {\mathrm{Cl}}_{{\mathrm{S}}^{\ensuremath{'}}}$, and ${\mathrm{Cl}}_{{\mathrm{S}}^{\ensuremath{'}\ensuremath{'}}}$) get thermally ionized and release free electrons into the system. Thus, the $n$-type conductivity of the ${\mathrm{In}}_{2}{\mathrm{S}}_{3}$ buffer layer increases. O impurities on S sites, in contrast, are electrically inert. Hence, we conclude that intentional doping by Cl is a means to improve the properties of $\ensuremath{\beta}\ensuremath{-}{\mathrm{In}}_{2}{\mathrm{S}}_{3}$ serving as buffer material.

A multiscale approach including density functional theory (DFT) and atomistic kinetic Monte Carlo (AKMC) simulations is applied to investigate the diffusion of interstitial oxygen atoms in bcc Fe under the influence of substitutional foreign atoms (Al, Si, P, S, Ti, Cr, Mn, Ni, Y, Mo, and W). The substitutional atoms can be assumed to be immobile since their diffusion coefficient is much smaller than that of oxygen. First, jumps of oxygen in pure bcc Fe between first-, second-, and third-neighbor octahedral interstitial sites are investigated. It is found that the first-neighbor jump is most relevant, with the tetrahedral site as the saddle point. The second-neighbor jump consists of two consecutive first-neighbor jumps. The barrier for a direct third-neighbor jump is too high to be significant for the diffusion process. In the presence of substitutional atoms the most important migration paths are first-neighbor jumps between modified octahedral sites with modified tetrahedral sites as saddle points. Calculations show that Si, P, Ni, Mo, and W cause some modifications of the migration barriers of oxygen and their interaction with O is mainly repulsive. Al, Cr, and Mn have a significant influence on the barriers and they exhibit strong attractive interactions with O. The most important modification of the barriers is found for S, Ti, and Y where deep attractive states exist. Based on the migration energies obtained by DFT, AKMC simulations on a rigid lattice are employed to determine the diffusion coefficient of oxygen in a dilute iron alloy containing the different substitutional atoms. It is found that Si, P, Ni, Mo, and W have almost no influence on the diffusivity of O. The presence of Al, Cr, Mn, S, Ti, and Y causes a significant reduction of the mobility of oxygen. In these cases the temperature dependence of the oxygen diffusion coefficient shows considerable deviations from an Arrhenius law. These phenomena are discussed in detail by considering the occupation time for the different states. The present findings on the strong dependence on the kind of substitutional atoms change the picture of oxygen diffusion in dilute iron alloys substantially.

Understanding the response of functional oxides to extrinsic ion insertion is important for technological applications including electrochemical energy storage and conversion, corrosion, and electronic materials in neuromorphic computing devices. Decoupling the complicated chemical and mechanical effects of ion insertion is difficult experimentally. In this work, we assessed the effect of lithium incorporation in zirconium oxide as a model system, by performing first-principles based calculations. The chemical effect of lithium is to change the equilibria of charged defects. Lithium exists in $\mathrm{Zr}{\mathrm{O}}_{2}$ as a positively charged interstitial defect, and raises the concentration of free electrons, negatively charged oxygen interstitials, and zirconium vacancies. As a result, oxygen diffusion becomes faster by five orders of magnitude, and the total electronic conduction increases by up to five orders of magnitude in the low oxygen partial pressure regime. In the context of Zr metal oxidation, this effect accelerates oxide growth kinetics. In the context of electronic materials, it has implications for resistance modulations via ion incorporation. The mechanical effect of lithium is in changing the volume and equilibrium phase of the oxide. Lithium interstitials together with zirconium vacancies shrink the volume of the oxide matrix, release the compressive stress that is needed for stabilizing the tetragonal phase $\mathrm{Zr}{\mathrm{O}}_{2}$ at low temperature, and promote tetragonal-to-monoclinic phase transformation. By identifying these factors, we are able to mechanistically interpret experimental results in the literature for zirconium alloy corrosion in different alkali-metal hydroxide solutions. These results provide a mechanistic and quantitative understanding of lithium-accelerated corrosion of zirconium alloy, as well as, and more broadly, show the importance of considering coupled electro-chemo-mechanical effects of cation insertion in functional oxides.

A distinguished structural characteristic of martensite in Fe-C steels is its tetragonality originating from carbon atoms occupying only one set of the three available octahedral interstitial sites in the body-centered-cubic (bcc) Fe lattice. Such a body-centered-tetragonal (bct) structure is believed to be thermodynamically stable because of elastic interactions between the interstitial carbon atoms. For such phase stability, however, there has been a lack of direct experimental evidence despite extensive studies of phase transformations in steels over one century. In this Rapid Communication, we report that the martensite formed in a high carbon Fe-8Ni-1.26C (wt%) steel at room temperature induced by applied stress/strain has actually a bcc rather than a bct crystal structure. This finding not only challenges the existing theories on the stability of bcc vs bct martensite in high carbon steels, but also provides insights into the mechanism for martensitic transformation in ferrous alloys.

In this paper we present a method to determine the occupation of Mn ions in epitaxial (In,Mn)As nanostructures by resonant x-ray diffraction near the Mn $K$-absorption edge, exploiting the dependency of the intensity of both (200) superstructure and the (400) fundamental reflections. The concentrations of Mn atoms on substitutional In sites, as well as on In- and As-interstitial sites, were unambiguously determined. A threshold concentration for the interstitial sites, which are occupied first for low nominal Mn deposition content, was found. Calculations using density-functional theory indicate that a higher occupancy of such sites can be favorable with respect to the occupation of substitutional sites, depending on surface potentials and growth kinetic effects.

Perovskite oxide interfaces have attracted tremendous research interest for their fundamental physics and promising all-oxide electronic applications. Here, based on first-principles calculations, we propose a surface La interstitial promoted interface insulator-metal transition in $\mathrm{LaAl}{\mathrm{O}}_{3}/\mathrm{SrTi}{\mathrm{O}}_{3}$ (110). Compared with surface oxygen vacancies, which play a determining role on the insulator-metal transition of $\mathrm{LaAl}{\mathrm{O}}_{3}/\mathrm{SrTi}{\mathrm{O}}_{3}$ (001) interfaces, we find that surface La interstitials can be more experimentally realistic and accessible for manipulation and more stable in an ambient atmospheric environment. Interestingly, these surface La interstitials also induce significant spin-splitting states with a Ti ${d}_{yz}/{d}_{xz}$ character at a conducting $\mathrm{LaAl}{\mathrm{O}}_{3}/\mathrm{SrTi}{\mathrm{O}}_{3}$ (110) interface. On the other hand, for insulating $\mathrm{LaAl}{\mathrm{O}}_{3}/\mathrm{SrTi}{\mathrm{O}}_{3}$ (110) (4 unit cells $\mathrm{LaAl}{\mathrm{O}}_{3}$ thickness), a distortion between La (Al) and O atoms is found at the $\mathrm{LaAl}{\mathrm{O}}_{3}$ side, partially compensating the polarization divergence. Our results reveal the origin of the metal-insulator transition in $\mathrm{LaAl}{\mathrm{O}}_{3}/\mathrm{SrTi}{\mathrm{O}}_{3}$ (110) heterostructures, and also shed light on the manipulation of the superior properties of $\mathrm{LaAl}{\mathrm{O}}_{3}/\mathrm{SrTi}{\mathrm{O}}_{3}$ (110) for different possibilities in electronic and magnetic applications.

The influence of the partial Er for Y substitution on the crystal structure and magnetic properties of $\mathrm{YF}{\mathrm{e}}_{2}{\mathrm{D}}_{4.2}$ has been investigated by high field magnetization and neutron diffraction experiments. ${\mathrm{Y}}_{0.7}\mathrm{E}{\mathrm{r}}_{0.3}\mathrm{F}{\mathrm{e}}_{2}{\mathrm{D}}_{4.2}$ compound crystallizes in the same monoclinic structure as $\mathrm{YF}{\mathrm{e}}_{2}{\mathrm{D}}_{4.2}$ described in $Pc\phantom{\rule{0.28em}{0ex}}({P}_{1}{c}_{1})$ space group with D atoms located in 18 different tetrahedral interstitial sites. A cell volume contraction of 0.6% is observed upon Er substitution, inducing large modification of the magnetic properties. Electronic effect of D insertion as well as lowering of crystal symmetry are important factors determining the magnetic properties of Fe sublattice, which evolves towards more delocalized behavior and modifying the Er-Fe exchange interactions. In the ground state, the Er and Fe moments are arranged ferrimagnetically within the plane perpendicular to the monoclinic $b$ axis and with average moments ${m}_{\mathrm{Er}}=6.4(3){\ensuremath{\mu}}_{B}{\mathrm{Er}}^{\ensuremath{-}1}$ and ${\mathrm{m}}_{\mathrm{Fe}}=2.0(1){\ensuremath{\mu}}_{B}{\mathrm{Fe}}^{\ensuremath{-}1}$ at 10 K. Upon heating, ${m}_{\mathrm{Er}}$ decreases progressively until ${T}_{\mathrm{Er}}=55\phantom{\rule{0.16em}{0ex}}\mathrm{K}$. Between 55 K and 75 K, the Fe sublattice undergoes a first-order ferromagnetic-antiferromagnetic (FM-AFM) transition with a cell volume contraction due to the itinerant metamagnetic behavior of one Fe site. In the AFM structure, ${m}_{\mathrm{Fe}}$ decreases until the N\'eel temperature ${T}_{\mathrm{N}}=125\phantom{\rule{0.16em}{0ex}}\mathrm{K}$. At high field, two different types of field induced transitions are observed. The Er moments become parallel to the Fe one and saturates to the $\mathrm{E}{\mathrm{r}}^{3+}$ free ion value, leading to an unusual field induced FM arrangement at a transition field ${B}_{\mathrm{Trans}}$ of only 78 kG below 30 K. Then above ${{T}_{\mathrm{M}}}_{0}=66\phantom{\rule{0.16em}{0ex}}\mathrm{K}$, an AFM-FM transition of the Fe sublattice, accompanied by a cell volume increase is observed. ${B}_{\mathrm{Trans}}$ increases linearly versus temperature and with a larger $d{B}_{\mathrm{Trans}}/dT$ slope than for $\mathrm{YF}{\mathrm{e}}_{2}{\mathrm{D}}_{4.2}$. This has been explained by the additional contribution of Er induced moments above ${B}_{\mathrm{Trans}}$.

Recent high-pressure studies have uncovered an alternative class of materials, insulating electride phases created by compression of simple metals. These exotic insulating phases develop an unusual electronic structure: the valence electrons move away from the nuclei and condense at interstitial sites, thereby acquiring the role of atomic anions or even molecules. We show that they are also topological phases as they exhibit a wide diversity of metallic surface states (SSs) that are controlled by the bulk electronic structure. The electronic reconstruction occurs that involves charge transfer between the surfaces of opposite polarity making both of them metallic, resembling the appearance of the two-dimensional gas at the renowned $\mathrm{SrTi}{\mathrm{O}}_{3}/\mathrm{LaAl}{\mathrm{O}}_{3}$ interface. Remarkably, these materials thus embody seemingly disparate physical concepts---chemical electron localization, topological control of bulk-surface conductivity, and the two-dimensional electron gas. Such metallic SSs could be probed by direct electrical resistance or by standard photoemission measurements on recovery to ambient conditions.

The light elemental solutes B, C, N, and O can penetrate the surface of Mg alloys and diffuse during heat treatment or high temperature application, forming undesirable compounds. We investigate the diffusion of these solutes by determining their stable interstitial sites and the inter-penetrating network formed by these sites. We use density functional theory (DFT) to calculate the site energies, migration barriers, and attempt frequencies for these networks to inform our analytical model for bulk diffusion. Due to the nature of the networks, O diffuses isotropically, while B, C, and N diffuse anisotropically. We compute the elastodiffusion tensor which quantifies changes in diffusivity due to small strains that perturb the diffusion network geometry and the migration barriers. The DFT-computed elastic dipole tensor which quantifies the change in site energies and migration barriers due to small strains is used as an input to determine the elastodiffusion tensor. We employ the elastodiffusion tensor to determine the effect of thermal strains on interstitial diffusion and find that B, C, and N diffusivity increases on crystal expansion, while O diffusivity decreases. From the elastodiffusion and compliance tensors we calculate the activation volume of diffusion and find that it is positive and anisotropic for B, C and N diffusion, whereas it is negative and isotropic for O diffusion.

The role of oxygen doping in CdTe is addressed by first-principles calculations. Formation energies, charge transition levels, and quasiparticle defect states are calculated within the $\text{DFT}+GW$ formalism. The formation of a new defect is identified, the $({\text{O}}_{\text{Te}}{\text{-Te}}_{\text{Cd}})$ complex. This complex is energetically favored over both isovalent $({\mathrm{O}}_{\text{Te}})$ and interstitial oxygen $({\mathrm{O}}_{\text{i}})$, in the Te-rich limit. We find that the incorporation of oxygen passivates the harmful deep energy levels associated with $({\mathrm{Te}}_{\text{Cd}})$, suggesting an improvement in the efficiency of CdTe based solar cells. Substitutional $({\text{O}}_{\text{Cd}})$ is only stable in the neutral charge state and undergoes a Jahn--Teller distortion. We also investigate the diffusion profiles of interstitial oxygen and find a low-energy diffusion barrier of only 0.14 eV between two structurally distinct interstitial sites.

The strongly correlated spin-electron system on a diamond chain containing localized Ising spins on its nodal lattice sites and mobile electrons on its interstitial sites is exactly solved in a magnetic field using the transfer-matrix method. We have investigated in detail all available ground states, the magnetization processes, the spin-spin correlation functions around an elementary plaquette, fermionic quantum concurrence and spin frustration. It is shown that the fermionic entanglement between mobile electrons hopping on interstitial sites and the kinetically-induced spin frustration are closely related yet independent phenomena. In the ground state, quantum entanglement only appears within a frustrated unsaturated paramagnetic phase, while thermal fluctuations can promote some degree of quantum entanglement above the non-frustrated ground states with saturated paramagnetic or classical ferrimagnetic spin arrangements., 11 pages, 10 figures

The Mn dopant distribution in ultrathin (20 nm) highly doped (nominal $x$ = 0.20) ${Ga}_{1\ensuremath{-}x}$Mn${}_{x}$As epitaxial films with critical temperatures close to 175 K and magnetization of 100 emu/cm${}^{3}$ is analyzed by Rutherford backscattering spectrometry (RBS) in a random and channeling configuration. We could quantify the total concentration and the respective fraction of substitutional, interstitial, and random site Mn ions in as-grown and annealed samples. The measured total Mn concentration is $x$ = 0.23. In the as-grown state 30% of the Mn dopant is located on interstitial sites. Thermal annealings at 180 \ifmmode^\circ\else\textdegree\fi{}C for several hours monotonically reduce the interstitial Mn fraction to 11%. Simultaneously the fraction of randomly located Mn is increased by the same amount. The substitutional Mn concentration is stable under these annealing conditions. The effective Mn concentration could be increased to $x$ = 0.13. However, the critical temperature does not increase proportionally with the magnetization. A comparison of the magnetization values demonstrates that the interstitial Mn ions are already incorporated during the growth in the form of Mn${}_{\mathrm{Ga}}$-Mn${}_{i}$ clusters.

A new compound with the composition of ${\mathrm{Sr}}_{3}{\mathrm{Ru}}_{2}{\mathrm{O}}_{7}{\mathrm{F}}_{2}$ has been obtained by fluorination of the double layered Ruddlesden--Popper phase, ${\mathrm{Sr}}_{3}{\mathrm{Ru}}_{2}{\mathrm{O}}_{7}.$ Neutron powder diffraction (NPD) reveals that due to a similar rotation of ${\mathrm{RuO}}_{6}$ octahedra, the structure of ${\mathrm{Sr}}_{3}{\mathrm{Ru}}_{2}{\mathrm{O}}_{7}{\mathrm{F}}_{2}$ can be described with the same space group $\mathrm{Pban}$ as ${\mathrm{Sr}}_{3}{\mathrm{Ru}}_{2}{\mathrm{O}}_{7}$ but with a large expansion in the c axis and slight contraction in the $\mathrm{ab}$ plane $[a=5.4231(4) \AA{},$ $b=5.4168(4) \AA{},$ and $c=24.2505(8) \AA{},$ 290 K]. The fluorinated ruthenate retains the layer structure: F occupies the interstitial sites in the rock salt layer leaving the double RuO layer intact. NPD at 2 K shows that ${\mathrm{Sr}}_{3}{\mathrm{Ru}}_{2}{\mathrm{O}}_{7}{\mathrm{F}}_{2}$ is antiferromagnetic, with a G-type ordering of the Ru spins and an ordered moment of ${\ensuremath{\mu}}_{z}=1{\ensuremath{\mu}}_{B}/\mathrm{Ru}$ along the c direction. Magnetization measurements implies a small ferromagnetic moment below 185 K, which is attributed to canting of the Ru spins from the c axis.

The structural arrangements and optic-vibrational densities of states of H and D in various fcc-based, mixed-isotope, rare-earth hydrides as measured by neutron powder diffraction and neutron vibrational spectroscopy clearly indicate preferential isotopic occupation of H in the octahedral interstices and D in the tetrahedral interstices. Although lessened by entropic contributions at high temperature, the degree of isotopic separation increases with decreasing temperature, presumably driven by the differences in the zero-point energies of H and D in the different interstitial sites. Decreasing the temperature below \ensuremath{\approx}150 K leads to a ``freezing in'' of the H and D lattice configuration prevalent near this temperature, kinetically precluding the complete isotopic site separation expected at temperatures approaching 0 K.

We investigate the structural properties, formation energies, and electronic structure of oxygen impurities in GaAs using first-principles total-energy calculations. Five charge states of oxygen occupying an arsenic site (OAs) and various interstitial sites (O i) were considered. For the OAs defect in negative charge states we find off-center configurations withC2v symmetry as reported experimentally. Our results for the formation energies reveal a negative-U behavior for the OAs defect, in which the paramagnetic 22 charge states is never stable. For the Oi defect, we find three equilibrium configurations for the O atom, which are present in all the charge states investigated. The stable configuration for the neutral defect shows the O atom between an As-Ga pair forming the As-O-Ga structure. However, for the negative charge states, the stable configuration shows the O atom exactly at the tetrahedral interstitial site, bonding with four gallium first neighbors. Further, we find that the 12 charge state is never stable, suggesting that the interstitial defect also exhibits a negative- U behavior. Based on our results, we suggest that the A, B8, and B bands of the local-vibrational-mode absorption spectrum of oxygen in GaAs are due to the off-center OAs defect. Also we show that this spectrum cannot be associated with an interstitial-oxygen configuration as previously proposed.

Raman scattering from an as-grown or annealed AlAs carbon δ-doping superlattice reveals lines at 1752 and 1856cm-1: the latter line shows the weaker intensity but has a resonant enhancement for incident light with an energy of 3 eV. These lines are comparable with those assigned to vibrational modes of two directly bonded dicarbon centers in GaAs [J. Wagner et al., Phys. Rev. Lett. 78, 74 (1997)]. First principles calculations are carried out to determine the structure and vibrational modes of dicarbon C-C defects located at various substitutional and interstitial sites in both AlAs and GaAs. The frequency of the C-C stretch mode is sensitive to the charge state and orientation and errors are not expected to exceed 10%. The dicarbon complex centered at an arsenic site is a deep donor and in its positively charged state is found to have axes aligned close to either 〈110〉 or 〈111〉 directions. The calculated frequencies and energies for the two orientations are essentially the same, so that these two structures offer an explanation for the observation of the two dicarbon Raman modes. An alternative assignment of one of the two observed modes to a different defect, such as an interstitial complex or neutral substitutional dimers, are considered but are ruled out as being incompatible with the experimental observations.

Titanium/zirconium-based quasicrystals and their related crystal approximants have been identified as potential new materials for hydrogen storage applications. To better understand the local chemistry and atomic ordering in these phases, preferential interstitial sites for hydrogen/deuterium were determined for $\ensuremath{\alpha}(\mathrm{TiCrSiO}).$ This is a bcc 1/1 crystal approximant to the icosahedral quasicrystal phase that contains a two-shell, Mackay-icosahedral cluster of atoms at each bcc site. It absorbs hydrogen or deuterium, without formation of other hydride phases, to a maximum hydrogen to metal atom ratio $(H/M)$ of 0.26. For fully deuterated samples, both tetrahedral and octahedral interstitial sites are occupied with fractions of 0.14 and 0.12, respectively. Here, the hydrogen/deuterium sites are determined from a Rietveld analysis of x-ray and neutron powder diffraction data taken from samples of $\ensuremath{\alpha}(\mathrm{TiCrSiO})$ loaded with deuterium. Only the octahedral sites are occupied in the partially deuterated samples $(D/M=0.11).$ A decrease in the oxygen concentration below the stoichiometric value for $\ensuremath{\alpha}(\mathrm{TiCrSiO})$ leads to an increase in the total amount of hydrogen that can be absorbed, suggesting that these interstitial atoms are competing for the same octahedral interstitial sites.

By combining vapor quenching technique at low temperature and in situ M\"ossbauer spectroscopy, we have studied the site occupation and magnetism of Fe in Pb and Yb films, as well as their stability with temperature. ${}^{57}\mathrm{Fe}$ M\"ossbauer spectra at 7 K, for as-prepared Pb--0.5 at. % Fe and Yb--0.5 at. % Fe films, have two paramagnetic components: a singlet, which is attributed to Fe at interstitial sites, and a doublet, assumed to be due to Fe atoms at ``substitutional'' sites, with at least one Fe near neighbor. M\"ossbauer measurements under an external magnetic field have been performed in as-prepared films and revealed that Fe atoms, associated with the doublet, have a magnetic moment in both systems, while those Fe atoms located at interstitial sites have magnetic moments only in the case of Yb films. These results confirm theoretical calculations predicting magnetic moments for interstitial Fe atoms in divalent hosts (Yb) and negligible probability of magnetic Fe atoms in tetravalent hosts (Pb).

A number of vacancy and interstitial defect migration mechanisms are characterized for crystalline silicon using supercells containing 64 and 216 atoms and a tight-binding approach. We investigate various defect configurations corresponding to minima and the pathways that connect them. A modified eigenvector-following approach is used to locate true transition states. We exploit the fact that only one Hessian eigenvector is needed to define the uphill search direction and use conjugate gradient minimization in the tangent space to produce a hybrid algorithm. Two implementations of this approach are considered, the first where second derivatives are available but full diagonalization of the Hessian would be the most time-consuming step, and the second where only first derivatives of the energy are known.

The atomic and the electronic structures of the \ensuremath{\sum}5 (210) [001] tilt grain boundary in ${\mathrm{Ni}}_{3}\mathrm{Al},$ with and without a hydrogen impurity, have been calculated using the full potential linearized-augmented plane-wave method. The strain field normal to the boundary plane and the excess grain boundary volume are calculated and compared with the results obtained using the embedded-atom method (EAM). The interlayer strain normal to the grain boundary oscillates with increasing distance from the grain boundary. The bonding charge distributions suggest that bonding in the boundary region is different from that in the bulk. Total-energy calculations show that the hydrogen impurity prefers to occupy interstitial sites on the Ni-rich grain boundary plane. Hydrogen is found to reduce the bonding charge across the boundary plane. The grain boundary energy and the Griffith cohesive energy for both the ``clean'' and H-segregated grain boundary are calculated and compared with the available EAM results. The hydrogen impurity is found to increase the grain boundary energy and reduce the Griffith cohesive energy of the boundary, which indicates that hydrogen is an embrittler of the grain boundary.

The vibrational modes of ${\mathrm{H}}_{2}$ molecules in Si are found using a first-principles method and compared with recent experimental investigations. The isolated molecule is found to lie at a ${T}_{d}$ interstitial site, oriented along [011] and is infrared active. The rotational barrier is at least 0.17 eV. The molecular frequency is a sensitive function of cage size and increases to lie close to the gas value for cages about 50% larger than the ${T}_{d}$ site. It is suggested that Raman-active modes around 4158 ${\mathrm{cm}}^{\ensuremath{-}1}$ are due to molecules within voids.

Hydrogen atoms have been produced at 77 K in single crystals of potassium dihydrogen phosphate (${\mathrm{KH}}_{2}{\mathrm{PO}}_{4}$ or KDP) using either 60-kV x rays or the fourth harmonic (266 nm) of a pulsed Nd:YAG (yttrium aluminum garnet) laser. The electron-paramagnetic-resonance spectrum from these hydrogen atoms exhibits a characteristic 500-G hyperfine splitting and is easily saturated with microwave power. These atoms occupy interstitial sites and are electron traps. The corresponding hole trap is the well-known $({\mathrm{HPO}}_{4}{)}^{\ensuremath{-}}$ hole center identified by its 31-G phosphorus hyperfine splitting. Both the hydrogen atoms and the hole centers thermally decay in the temperature range between 80 and 200 K. The observed displacement of protons by 266-nm photons provides direct evidence in support of the proton-transport mechanism recently proposed by Davis, Hughes, and Lee [Chem. Phys. Lett. 207, 540 (1993)] to explain laser-induced transient optical absorption at room temperature in KDP.

We have used x-ray scattering to measure the structure of the passive oxide film formed at high anodic potentials on Fe(110) and Fe(001). The crystalline film has a small crystallite size ( $\ensuremath{\approx}50\AA{}$) and is oriented with the substrate. The film structure is based on ${\mathrm{Fe}}_{3}{\mathrm{O}}_{4}$, but with cation vacancies on octahedral and tetrahedral sites (80% and 66% occupancies, respectively) and with cations occupying octahedral interstitial sites (12% occupancy). These results resolve the long-standing controversy surrounding the film structure and provide a basis for understanding and modeling film properties important for corrosion resistance.

The local order around carbon atoms in epitaxial ${\mathrm{Si}}_{1\ensuremath{-}y}{\mathrm{C}}_{y}$ alloys (with $y$ around 1%) grown by Si molecular-beam epitaxy and thermal decomposition of ${\mathrm{C}}_{2}{\mathrm{H}}_{4}$ on Si(001) or Si(111) has been studied as a function of growth temperature. To this end, information from local probes such as x-ray photoelectron spectroscopy C $1s$ binding energies, experimental and simulated C $1s$ core-level x-ray photoelectron diffraction (XPD) distributions, and Raman spectroscopy are compared. Between the growth conditions leading to a solid solution of substitutional C at lower temperature and those leading to SiC precipitation at higher temperature, original metastable and ordered C-rich alloy phases appear, which may be coherently embedded in Si without degrading crystal quality. The results for such phases probed by the two latter techniques are indicative of a crystalline order implying concentrated third-nearest-neighbor (NN) carbon pairs. The comparison of experimental XPD distributions recorded in different azimuthal planes with simulated ones with C either in substitutional or in interstitial sites is in favor of substitution with a local contraction of the first-neighbor Si-C bond length between 10% and 20%. If we admit that the surface ordering of the C atoms in the Si(001) surface layers is the extension of such particular bulk arrangements in third NN C pairs we are able to explain an experimentally observed $c\ensuremath{-}(4\ifmmode\times\else\texttimes\fi{}4)$ low-energy electron diffraction pattern.

When heating over 300--800 K in a calorimeter, ${\mathrm{Nd}}_{2}{\mathrm{Fe}}_{14}{\mathrm{BH}}_{x},$ $xl~5,$ microcrystals desorb the H atoms in six irreversible endothermic signals. They appear at temperatures ${T}_{p}\ensuremath{\sim}340,$ 439 (doubly degenerate), 567, 605, and 653 K at the heating rate $\ensuremath{\beta}=15\mathrm{K}/\mathrm{m}\mathrm{i}\mathrm{n},$ following the modified Kissinger relation $\mathrm{ln}{(T}_{p}/\ensuremath{\beta}{)=(E}_{a}{/R)T}_{p}^{\ensuremath{-}1}+\mathrm{const},$ with $R$ the gas constant. The 340 signal has an extremely weak intensity at a slow heating, $\ensuremath{\beta}l~40\mathrm{K}/\mathrm{m}\mathrm{i}\mathrm{n},$ because the H atoms in the involved interstitial site in the sample slowly tunnel to neighboring higher-transition-energy sites. The thermogram at 439 K is the most prominent. It contains two signals which could be resolved by selective isothermal desorptions. These six different identified thermal signals are assigned to the desorptions of the H atoms from six specific $4c,$ $16k2,$ $16k1,$ $4e,$ $8j2,$ and $8j1$ crystallographic interstitial sites (between the Fe atoms) in the sample, characterized by six different activation energies ${E}_{a}$ between 48 and 123 kJ/mol, taking into account their H occupancies ${n}_{i}(\mathrm{H}).$ The distribution of partial enthalpies in the thermal signals (at $\ensuremath{\beta}=15\mathrm{K}/\mathrm{m}\mathrm{i}\mathrm{n}$) determines ${n}_{i}(\mathrm{H})\ensuremath{\cong}0,$ 6, 5, 2, 5, and 2 H atoms in the respective sites per ${\mathrm{Nd}}_{2}{\mathrm{Fe}}_{14}{\mathrm{BH}}_{x},$ $x\ensuremath{\sim}5,$ crystal unit cell with a total of $4x=20\mathrm{H}$ atoms. It is found that a significant portion of the thermally excited H atoms in these sites, in the process to the desorption, redistributes over neighboring sites of modified energies to keep the H atoms at high temperatures. The redistribution reaction is exothermic. It has been observed separately in $8j2$ and $8j1$ sites at 616 and 684 K (at $\ensuremath{\beta}=50\mathrm{K}/\mathrm{m}\mathrm{i}\mathrm{n}$) in a partially H-desorbed sample in the lower-transition-energy sites by heating it at 583 K. A local redistribution of the interatomic distances and/or the electronic charges occurs within the lattice following the thermal desorption of the H atoms. It results in a monotonically decreasing exothermic (structural relaxation) signal of the rate of the change of the enthalpy, $(\ensuremath{\partial}H/\ensuremath{\partial}{t)}_{T},$ with time $t,$ after the desorption at $t=0,$ following the primary endothermic desorption signal. The results are discussed with simulations of related processes.

We have used dielectric spectroscopy to measure the frequency and temperature dependence of the low-frequency (0.5 mHz--100 kHz) complex dielectric function \ensuremath{\varepsilon}(\ensuremath{\omega}) of thin ${\mathrm{C}}_{60}$ films. A small charge transfer occurs between the ${\mathrm{C}}_{60}$ and ${\mathrm{O}}_{2}$ molecules which occupy interstitial spaces of solid ${\mathrm{C}}_{60}$ exposed to oxygen (or ambient air). Due to the large size of the ${\mathrm{C}}_{60}$ molecules, this small charge transfer creates large dipole moments, which in turn are coupled to the applied ac electric field via a diffusion-controlled relaxation mechanism. This gives rise to a significant increase in the permittivity ${\mathrm{\ensuremath{\varepsilon}}}_{1}$ accompanied by a broad dielectric loss peak ${\mathrm{\ensuremath{\varepsilon}}}_{2}$ observed at \ensuremath{\sim}10--100 Hz. With increasing oxygenation, the interstitial sites become nearly fully occupied, interstitial hopping is inhibited, and the loss peaks, together with the enhanced polarization, disappear. Both tracer and chemical diffusion coefficients for the ${\mathrm{C}}_{60}$/${\mathrm{O}}_{2}$ system have been obtained purely from dielectric spectroscopy measurements.

A quantitative model of phosphorous-diffusion gettering in silicon is presented, which combines the effects of segregation and self-interstitial injection on the distribution of dissolved metallic impurities. The model describes metal diffusion both in the bulk and in the highly phosphorous-doped layer and makes it possible to include phosphorous-diffusion models. By analyzing an approximate solution for the quasi-steady-state metal distribution, we show that for impurities like gold and platinum self-interstitial injection enhances the gettering efficiency compared to pure segregation. We apply the results to phosphorous-diffusion gettering of gold and demonstrate that all relevant features of recently measured gold distributions can be interpreted consistently. For 3d metals, which are predominantly dissolved on interstitial sites in intrinsic silicon, the model allows us to include the formation of precipitates resulting from self-interstitial injection as proposed earlier. {copyright} {ital 1997} {ital The American Physical Society}

Computer simulation techniques are used to investigate the defect and dopant properties of the oxyfluoride superconductor, Sr{sub 2}CuO{sub 2}F{sub 2+{delta}}. The observed La{sub 2}CuO{sub 4}-type structure is correctly reproduced by our interatomic potential model. Defect calculations find that the most favorable fluorine interstitial position is near the ideal site and is associated with significant local relaxation, in accord with diffraction studies. A variety of redox reactions relevant to high-T{sub c} behavior have been examined. Our results confirm that oxidation, involving fluorine incorporation at interstitial sites, is the most favorable process to introduce the holes necessary for superconductivity. We also consider the energetics of a range of dopant substitutions (alkali and alkaline-earth ions) at the Sr site. {copyright} {ital 1997} {ital The American Physical Society}

experiments indicate that part of the doped TM ions (TM =Fe, Co, and Ni) substitute for Ir, and the rest intercalate into the octahedral interstitial sites located in between IrTe2 layers. Due to the lattice mismatch between MnTe2 and IrTe2, Mn has limited solubility in IrTe2 lattice. The trigonal structure is stable in the whole temperature range 1.80 K T 300 K for all doped compositions. No long range magnetic order or superconductivity was observed in any doped compositions above 1.80 K. A spin glass behavior below 10 K was observed in Fe-doped IrTe2 from the temperature dependence of magnetization, electrical resistivity, and specic heat. The low temperature specic heat data suggest the electron density of states is enhanced in Feand Co-doped compositions but reduced in Ni-doped IrTe2. With the 3d transition metal doping the trigonal a-lattice parameters increases but the c-lattice parameter decreases. Detailed analysis of the single crystal x-ray diraction data shows that interlayer Te-Te distance increases despite a reduced c-lattice. The importance of the Te-Te, Te-Ir, and Ir-Ir bonding is discussed.

The dynamic responses of a rhodium catalyst and a graphene sheet are investigated upon random excitation with 80 kV electrons. An extraordinary electron microscope stability and resolution allow studying temporary atom displacements from their equilibrium lattice sites into metastable sites across projected distances as short as 60 pm. In the rhodium catalyst, directed and reversible atom displacements emerge from excitations into metastable interstitial sites and surface states that can be explained by single atom trajectories. Calculated energy barriers of 0.13 eV and 1.05 eV allow capturing single atom trapping events at video rates that are stabilized by the Rh [110] surface corrugation. Molecular dynamics simulations reveal that randomly delivered electrons can also reversibly enhance the $s{p}^{3}$ and the $s{p}^{1}$ characters of the $s{p}^{2}$-bonded carbon atoms in graphene. The underlying collective atom motion can dynamically stabilize characteristic atom displacements that are unpredictable by single atom trajectories. We detect three specific displacements and use two of them to propose a path for the irreversible phase transformation of a graphene nanoribbon into carbene. Collectively stabilized atom displacements greatly exceed the thermal vibration amplitudes described by Debye-Waller factors and their measured dose rate dependence is attributed to tunable phonon contributions to the internal energy of the systems. Our experiments suggest operating electron microscopes with beam currents as small as zepto-amperes/nm${}^{2}$ in a weak-excitation approach to improve on sample integrity and allow for time-resolved studies of conformational object changes that probe for functional behavior of catalytic surfaces or molecules.

While the bulk, stoichiometric Bi${}_{2}$Te${}_{3}$ single crystals often exhibit $p$-type metallic electrical conduction due to the Bi${}_{\mathrm{Te}}$-type antisite defects, doping by thallium (Bi${}_{2\ensuremath{-}x}$Tl${}_{x}$Te${}_{3}$, $x=0\ensuremath{-}0.30$) progressively changes the electrical conduction of single crystals from $p$ type (0 \ensuremath{\le} $x$ \ensuremath{\le} 0.08) to $n$ type (0.12 \ensuremath{\le} $x$ \ensuremath{\le} 0.30). This is observed via measurements of both the Seebeck coefficient and the Hall effect performed in the crystallographic (0001) plane in the temperature range of 2--300 K. Since any kind of substitution of Tl on the Bi or Te sublattices would result in an enhancement of the density of holes rather than its decrease, and because simple incorporation of Tl at interstitial sites or in the van der Waals gap is unlikely as it would increase the lattice parameters which is not observed in experiments, incorporation of Tl likely proceeds via the formation of TlBiTe${}_{2}$ fragments coexisting with the quintuple layer structure of Bi${}_{2}$Te${}_{3}$. At low levels of Tl, 0 \ensuremath{\le} $x$ \ensuremath{\le} 0.05, the temperature-dependent in-plane ($I\ensuremath{\perp}c$) electrical resistivity maintains its metallic character as the density of holes decreases. Heavier Tl content with 0.08 \ensuremath{\le} $x$ \ensuremath{\le} 0.12 drives the electrical resistivity into a prominent nonmetallic regime displaying characteristic metal-insulator transitions upon cooling to below \ensuremath{\sim}100 K. At the highest concentrations of Tl, 0.20 \ensuremath{\le} $x$ \ensuremath{\le} 0.30, the samples revert back into the metallic state with low resistivity. Thermal conductivity measurements of Bi${}_{2}$Te${}_{3}$ single crystals containing Tl, as examined by the Debye-Callaway phonon conductivity model, reveal a generally stronger point-defect scattering of phonons with the increasing Tl content. The systematic evolution of transport properties suggests that the Fermi level of Bi${}_{2}$Te${}_{3,}$ which initially lies in the valence band (for $x$ = 0), gradually shifts toward the top of the valence band (for 0.01 \ensuremath{\le} $x$ \ensuremath{\le} 0.05), then moves into the band gap (for 0.08 \ensuremath{\le} $x$ \ensuremath{\le} 0.12), and eventually intersects the conduction band (for 0.20 \ensuremath{\le} $x$ \ensuremath{\le} 0.30).

The electronic properties of Cu0.07Bi2Se3 have been investigated using Shubnikov-de Haas and optical reflectance measurements. Quantum oscillations reveal a bulk, three-dimensional Fermi surface with anisotropy k(F)(c)/k(F)(ab) approximate to 2 and a modest increase in free-carrier concentration and in scattering rate with respect to the undoped Bi2Se3, also confirmed by reflectivity data. The effective mass is almost identical to that of Bi2Se3. Optical conductivity reveals a strong enhancement of the bound impurity bands with Cu addition, suggesting that a significant number of Cu atoms enter the interstitial sites between Bi and Se layers or may even substitute for Bi. This conclusion is also supported by x-ray diffraction measurements, where a significant increase of microstrain was found in Cu0.07Bi2Se3, compared to Bi2Se3.

Combined density functional theory and empirical-potential calculations are performed to investigate the lowest-energy sites and migration mechanisms of He in various $\ensuremath{\alpha}$-Fe grain boundaries (GBs). Before the defect calculations, we show that structural optimizations, including simulated annealing and atom removal, are crucial for locating the stable GB structure in a given temperature regime. Then, the He formation energies for all the substitutional and interstitial sites in two different GBs are evaluated, showing a strong He segregation tendency. At variance with the bulk Fe case, the formation energy of an interstitial He is either lower than or similar to that of a substitutional He in the GBs. Finally, both static and dynamic barriers for interstitial He diffusion in the GBs are determined. Although the diffusion details and precise paths are GB dependent, some common features are identified: (1) The He atom always remains confined to the GB region while diffusing; (2) the He diffusion is highly anisotropic along the GBs; (3) the GB diffusion of an interstitial He atom is found to be always slower than its bulk diffusion, but it can still be faster than the bulk diffusion of a substitutional He.

A combined experimental and theoretical study on the incorporation of Mn in GaAs has been presented. We have successfully controlled the location of Mn atoms at GaAs(001) surfaces by changing the surface atomic geometry. While Mn atoms prefer to substitute Ga sites at a subsurface layer under the As-rich conditions, the incorporation into interstitial sites becomes more favorable as the surface As coverage is decreased. The present results provide a mechanism for the enhanced incorporation of substitutional Mn atoms in GaMnAs under low-temperature (i.e., As-rich) growth conditions.

Boron is a common alloying element in modern steels with a significant influence on the mechanical properties already at concentrations of only a few parts per million. The effect of boron depends on its distribution in the microstructure. Here, we characterize the elemental factors that determine the boron distribution in $\ensuremath{\alpha}$-iron by density functional theory calculations. Boron as point defect has been considered in substitutional and interstitial sites. The calculated migration barriers for the substitutional and interstitial mechanisms show the first nearest-neighbor hops being preferred over second nearest-neighbor hops. A dissociative mechanism shows boron migrating via an interstitial mechanism to be likely trapped by vacancies. In order to characterize the interaction with other point defects, we determined the distance-dependent interaction energy of a boron defect with a vacancy, a second boron, and with hydrogen, carbon, nitrogen, oxygen, aluminum, silicon, phosphorus, and sulfur atoms. We find that substitutional boron binds strongly to interstitial point defects with dumbbell formation and weaker to substitutional point defects. Interstitial boron tends to repel substitutional and interstitial point defects. We find a similarity of substitutional boron and vacancies regarding their influence on elastic properties and their interaction with point defects in $\ensuremath{\alpha}$-iron.

The growth, microstructure, and electrochemical oxygen intercalation of c-axis ${\mathrm{La}}_{2}$${\mathrm{CuO}}_{4+\mathrm{\ensuremath{\delta}}}$ thin films on substrates with different lattice mismatch [${\mathrm{SrTiO}}_{3}$ (001) and ${\mathrm{SrLaAlO}}_{4}$ (001) substrates] are compared. Except for the absence of planar defects in the latter case, the microstructural properties of both film types are very similar. For films on ${\mathrm{SrTiO}}_{3}$, oxygen can be intercalated electrochemically into the grown c-axis thin film with a high diffusion coefficient (${\mathit{D}}_{\mathrm{ox}}$=${10}^{\mathrm{\ensuremath{-}}13}$--${10}^{\mathrm{\ensuremath{-}}14}$ ${\mathrm{cm}}^{2}$/s), and subsequently additional (00l) reflections, l=2n+1, are observed by x-ray diffraction measurements. A double transition with a resistivity drop at \ensuremath{\simeq}55-58 K, suggesting a more strongly oxidized phase, and a zero-resistance state at \ensuremath{\simeq}42 K are found. For films on ${\mathrm{SrLaAlO}}_{4}$, the value of ${\mathit{T}}_{\mathit{c}}$ could not be raised further, as the films decompose during the anodic polarization. This comparison reveals the role of planar defects, and we propose an electrochemical oxidation mechanism that occurs in two steps: First oxygen is transported into the film by intercalation into the planar defects, and then a slower oxygen diffusion into the interstitial sites occurs along the ab planes. \textcopyright{} 1996 The American Physical Society.