The substitutional boron-vacancy BsV complex in silicon is investigated using the local density functional theory. These theoretical results give an explanation of the experimentally reported, well established metastability of the boron-related defect observed in p-type silicon irradiated at low temperature and of the two hole transitions that are observed to be associated with one of the configurations of the metastable defect. BsV is found to have several stable configurations, depending on charge state. In the positive charge state the second nearest neighbor configuration with C1 symmetry is almost degenerate with the second nearest neighbor configuration that has C1h symmetry since the bond reconstruction is weakened by the removal of electrons from the center. A third nearest neighbor configuration of BsV has the lowest energy in the negative charge state. An assignment of the three energy levels associated with BsV is made. The experimentally observed Ev+0.31 eV and Ev+0.37 eV levels are related to the donor levels of second nearest neighbor BsV with C1 and C1h symmetry respectively. The observed Ev+0.11 eV level is assigned to the vertical donor level of the third nearest neighbor configuration. The boron-divacancy complex BsV2 is also studied and is found to be stable with a binding energy between V2 and Bs of around 0.2 eV. Its energy levels lie close to those of the V2. However, the defect is likely to be an important defect only in heavily doped material.
The electrical activity of Cs-H defects in Si has been investigated in a combined modeling and experimental study. High-resolution Laplace capacitance spectroscopy with the uniaxial stress technique has been used to measure the stress-energy tensor and the results are compared with theoretical modeling. At low temperatures, implanted H is trapped as a negative-U center with a donor level in the upper half of the gap. However, at higher temperatures, H migrates closer to the carbon impurity and the donor level falls, crossing the gap. At the same time, an acceptor level is introduced into the upper gap making the defect a positive-U center.
A combination of normal-incidence x-ray standing-wave (NIXSW) spectroscopy, x-ray photoelectron spectroscopy (XPS), scanning tunneling microscopy (STM), and density-functional theory (DFT) has been used to investigate the interaction of a number of phthalocyanine molecules (specifically, SnPc, PbPc, and CoPc) with the Ag(111) surface. The metal-surface distances predicted by the DFT calculations for SnPc/Ag(111) (2.48Å) and CoPc/Ag(111) (2.88Å) are in good agreement with our NIXSW experimental results for these systems (2.31±0.09 and 2.90±0.05Å, respectively). Good agreement is also found between calculated partial density-of-states plots and STM images of CoPc on Ag(111). Although the DFT and Pb4f NIXSW results for the Pb-Ag(111) distance are similarly in apparently good agreement, the Pb4f core-level data suggest that a chemical reaction between PbPc and Ag(111) occurs due to the annealing procedure used in our experiments and that the similarity of the DFT and Pb4f NIXSW values for the Pb-Ag(111) distance is likely to be fortuitous. We interpret the Pb4f XPS data as indicating that the Pb atom can detach from the PbPc molecule when it is adsorbed in the "Pb-down" position, leading to the formation of a Pb-Ag alloy and the concomitant reduction in Pb from a Pb2 + state (in bulklike films of PbPc) to Pb0. In contrast to SnPc, neither PbPc nor CoPc forms a well-ordered monolayer on Ag(111) via the deposition and annealing procedures we have used. Our DFT calculations show that each of the phthalocyanine molecules donate charge to the silver surface, and that back donation from Ag to the metal atom (Co, Sn, or Pb) is only significant for CoPc
30° and 90° Shockley partial dislocations lying in {111} and basal planes of cubic and hexagonal silicon carbide, respectively, are investigated theoretically. Density-functional-based tight-binding total-energy calculations are used to determine the core structure and energetics of the dislocations. In a second step their electronic structure is investigated using a pseudopotential method with a Gaussian basis set. Finally, the thermal activation barriers to glide motion of 30° and 90° Shockley partials are calculated in terms of a process involving the formation and migration of kinks along the dislocation line. The mechanism for enhanced dislocation movement observed under current injection conditions in bipolar silicon carbide devices is discussed.
Silicon and germanium single crystals are implanted with protons. The infrared-absorption spectra of the samples contain sharp absorption lines due to the excitation of hydrogen-related local vibrational modes. The lines at 743.1, 748.0, 1986.5, and 1989.4 cm-1 in silicon and at 700.3, 705.5, 1881.8, and 1883.5 cm-1 in germanium originate from the same defect in the two materials. Measurements on samples coimplanted with protons and deuterons show that the defect contains two equivalent hydrogen atoms. Uniaxial stress measurements are carried out and a detailed analysis of the results is presented. It is shown that the defect has monoclinic-II symmetry, and the orientations of the Si-H and Ge-H bonds of the defect are determined. Ab initio local-density-functional theory is used to calculate the structure and local vibrational modes of the self-interstitial binding one and two hydrogen atoms in silicon and germanium together with the structure of the self-interstitial itself. The observed properties of the defect are in excellent agreement with those calculated for the self-interstitial binding two hydrogen atoms.
We use first-principles models to demonstrate how an organic oxidizing agent F(4)-TCNQ (7,7,8,8-tetracyano-2,3,5,6-tetrafluoroquinodimethane) modifies the electronic structure of silicon nanocrystals, suggesting it may enhance p-type carrier density and mobility. The proximity of the lowest unoccupied level of F(4)-TCNQ to the highest occupied level of the Si nanocrystals leads to the formation of an empty hybrid state overlapping both the nanocrystal and molecule, reducing the excitation energy to similar to 0.8-1 eV in vacuum. Hence, it is suggested that F(4)-TCNQ can serve both as a surface oxidant and as a mediator for hole hopping between adjacent nanocrystals in p-type doped silicon nanocrystal networks
The properties of the cation vacancy and the Te antisite, two dominant defects in CdTe and Cd1-xZnxTe alloys grown in Te-rich conditions, are examined using first-principles calculations. First, the structure, electronic levels, and migration paths of V-Cd and Te-Cd in CdTe are studied in detail. Additionally, we analyze the evolution of the stability and electronic properties in Cd1-xZnxTe alloys, taking into account both the role of alloying in the position of the ionization levels and its effects on the equilibrium concentration of those two defects. It is shown that the formation of cation vacancies becomes progressively more favorable as x increases, whereas Te antisites become less stable, backing the trend towards p-type conductivity in dilute Cd1-xZnxTe.
Silicon nanocrystals with diameters between 1 and 3 nm and surfaces passivated by chlorine or a mixture of chlorine and hydrogen were modeled using density functional theory, and their properties compared with those of fully hydrogenated nanocrystals. It is found that fully and partially chlorinated nanocrystals are stable, and have higher electron affinity, higher ionization energy, and lower optical absorption energy threshold. As the hydrogenated silicon nanocrystals, chlorinated silicon nanocrystals doped with phosphorus or boron require a high activation energy to transfer an electron or hole, respectively, to undoped silicon nanocrystals. The electronic levels of surface dangling bonds are similar for both types of surface passivation, although in the chlorinated silicon nanocrystals some fall outside the narrower energy gap.
We carry out a comprehensive density-functional study of the vacancy-oxygen (VO) center in germanium using large H-terminated Ge clusters. The importance of a nonlinear core correction to account for the involvement of the 3d electrons in Ge-O bonds is discussed. We calculate the electrical levels and the vibrational modes of VO0, VO-, and VO= finding close agreement with experiment. We also explore the reorientation, migration, and dissociation mechanisms of neutral and negatively charged VO and compare the calculated energy barriers with experimental data. We conclude that the defect is likely to anneal through both mechanisms.
The piezospectroscopic properties of the VOH defect in Si are found using stress Laplace deep level transient spectroscopy (DLTS) and are compared with local density-functional calculations of (i) the acceptor level and its shift under stress, and (ii) the alignment of the neutral center under stress. The theory is able to account for two acceptor levels observed for 〈100〉, 〈111〉, and 〈110〉 stress even though additional splitting is expected for a defect with static C1h symmetry. This is related to (i) a rapid reorientation of the H atom within the defect at temperatures at which the DLTS experiments are carried out, and (ii) the small effect of stress on two orientations of the defect under 〈110〉 stress. The theory is also able to give a quantitative account of the alignment of the center. The effect of stress on the reorientation barrier of the defect is also investigated. The reorientation barrier of the defect in its positive charge state is found theoretically to be very small, consistent with the lack of any splitting in the donor level under stress.
Ab initio density-functional calculations using Gaussian orbitals are carried out on large Si and Ge supercells containing oxygen defects. The formation energies, local vibrational modes, and diffusion or reorientation energies of Oi, O2i, VO, VOH, and VO2 are investigated. The piezospectroscopic tensors for Oi, VO, and VO2 are also evaluated. The vibrational modes of Oi in Si are consistent with the view that the defect has effective D3d symmetry at low hydrostatic pressures but adopts a buckled structure for large pressures. The anomalous temperature dependence of the modes of O2i is attributed to an increased buckling of Si-O-Si when the lattice contracts. The diffusion energy of the dimer is around 0.8 eV lower than that of Oi in Si and 0.6 eV in Ge. The dimer is stable against VO or VO2 formation and the latter defect has modes close to the reported 894-cm-1 band. The reorientation energies for O and H in VO and VOH defects are found to be a few tenths of an eV and are greater when the defect has trapped an electron.
The interstitial carbon-oxygen defect is a prominent defect formed in e-irradiated Cz-Si containing carbon. Previous stress alignment investigations have shown that the oxygen atom weakly perturb the carbon interstitial but the lack of a high-frequency oxygen mode has been taken to imply that the oxygen atom is severely affected and becomes overcoordinated. Local vibrational mode spectroscopy and ab initio modeling are used to investigate the defect. We find new modes whose oxygen isotopic shifts give further evidence for oxygen overcoordination. Moreover, we find that the calculated stress-energy tensor and energy levels are in good agreement with experimental values. The complexes formed by adding both single (CiOiH) and a pair of H atoms (CiOiH2), as well as the addition of a second oxygen atom, are considered theoretically. It is shown that the first is bistable with a shallow donor and deep acceptor level, while the second is passive. The properties of CiOiH and CiO2iH are strikingly similar to the first two members of a family of shallow thermal donors that contain hydrogen.
We present a comprehensive spin-density functional modeling study of the structural and electronic properties of donor-vacancy complexes (PV, AsV, SbV, and BiV) in Ge crystals. Special attention is paid to spurious results which are related to the choice of the boundary conditions (supercell-cluster approach), the resulting band-gap width, and the choice of the points to sample the Brillouin zone. The underestimated energy gap, resulting from the periodic conditions together with the local-density approximation to the exchange-correlation energy, leads to defect-related gap states that are strongly coupled to crystalline states within the center of the zone. This is shown to produce a strong effect even on relative energies. Our results indicate that in all E centers the donor atom occupies a nearly substitutional site, as opposed to the split-vacancy form adopted by the SnV complex in Si. The E centers can occur in four charge states, from positive to double negative, and produce occupancy levels at E(0/+)=Ev+0.1 eV, E(-/0)=Ev+0.3 eV, and E(=/-)=Ec-0.3 eV.
The trivacancy (V3) in silicon has been recently shown to be a bistable center in the neutral charge state, with a fourfold-coordinated configuration, V3[FFC], lower in energy than the (110) planar one [ V. P. Markevich et al. Phys. Rev. B 80 235207 (2009)]. Transformations of the V3 defect between different configurations, its diffusion, and disappearance upon isochronal and isothermal annealing of electron-irradiated Si:O crystals are reported from joint deep level transient spectroscopy measurements and first-principles density-functional calculations. Activation energies and respective mechanisms for V3 transformation from the (110) planar configuration to the fourfold-coordinated structure have been determined. The annealing studies demonstrate that V3 is mobile in Si at T>200 ∘C and in oxygen-rich material can be trapped by interstitial oxygen atoms so resulting in the appearance of V3O complexes. The calculations suggest that V3 motion takes place via consecutive FFC/planar transformation steps. The activation energy for the long-range diffusion of the V3 center has been derived and agrees with atomic motion barrier from the calculations
First-principles density functional calculations are used to investigate antisite pairs in 4H-SiC. We show that they are likely to be formed in close proximity under ionizing conditions, and they possess a donor level and thermal stability consistent with the series of 40 photoluminescent lines called the alphabet lines. Moreover, the gap vibrational mode of the silicon antisite defect is close to a phonon replica of the b1 line and possesses a weak isotopic shift with 13C in agreement with observation.
The DI center is a prominent defect which is detected in as-grown or irradiated SiC. It is unusual in that its intensity grows with heat treatments and survives anneals of 1700 °C. It has been assigned recently to either a close-by antisite pair or to the close-by antisite pair adjacent to a carbon antisite. We show here using local density functional calculations that these defects are not stable enough to account for DI. Instead, we assign DI to an isolated Si antisite and the four forms of the close-by antisite pair in 4H-SiC to the a, b, c, and d members of the alphabet series. The assignments allow us to understand the concentration of DI following growth, the recombination enhanced destruction of these alphabet defects and the annealing behavior of the remaining members of the series.
Local-density-functional methods are used to examine the behavior of the oxygen defect, gallium vacancy, and related defect complexes trapped at threading-edge dislocations in GaN. These defects are found to be particularly stable at the core of the dislocation where oxygen sits twofold coordinated in a bridge position. VGa-ON is found to be a deep double acceptor, VGa-(ON)2 is a deep single acceptor, and VGa-(ON)3 at the dislocation core is electrically inactive. We suggest that the first two defects are responsible for a deep acceptor level associated with the midgap yellow luminescence band.
We present ab initio calculations of the magnetic moments and magnetic anisotropy energies of small FeCo clusters of varying composition on top of a Cu(100) substrate. Three different cluster layouts have been considered, namely, 2×2, 3×3, and crosslike pentamer clusters. The ratio of Co atoms with respect to the total number in a chosen cluster ("concentration") was varied and all possible arrangements of the atomic species were taken into account. Calculations have been performed fully relativistic using the embedded-cluster technique in conjunction with the screened Korringa-Kohn- Rostoker method and the magnetocrystalline anisotropy energy (MAE) has been evaluated by means of the magnetic force theorem. A central result of the investigations is that the size of the magnetic moments of the individual Fe and Co atoms and their contributions to the anisotropy energy depend on the position they occupy in a particular cluster and on the type and the number of nearest neighbors. The MAE for the 2×2 and 3×3 clusters varies with respect to the concentration of Co atoms in the same manner as the corresponding monolayer case, whereas the pentamer clusters show a slightly different behavior. Furthermore, for the clusters with an easy axis along a direction in the surface plane, the MAE shows a significant angular dependence.
The fully relativistic spin-polarized screened Korringa-Kohn-Rostoker method is used to evaluate the electronic and magnetic structure as well as the optical conductivity of (ComIrm)n superstructures on Ir(111). By mapping the microscopic optical conductivity tensor onto the macroscopic permittivity tensor and by using the so-called 2×2 matrix technique the surface reflectivity matrices for these systems are then calculated, from which in turn the Kerr rotation and ellipticity angles can be determined. It is found (i) that when varying at a given value of m the number of repetitions n, these angles are linearly proportional to the total magnetic moment, and (ii) that at a frequency of about 3.8 eV the Kerr rotation angles have the largest value, the corresponding maximum being mainly caused by the Ir spacer layers. The optical properties of the free surface of Ir(111), which is considered to check the applied theoretical schemes, turn out to be in good agreement with existing experimental data.
In using the fully relativistic versions of the embedded cluster and screened Korringa-Kohn-Rostoker methods for semi-infinite systems the magnetic properties of single adatoms of Fe and Co on Ir(111) and Pt(111) are studied. It is found that for Pt(111) Fe and Co adatoms are strongly perpendicularly oriented, while on Ir(111) the orientation of the magnetization is only out of plane for a Co adatom; for an Fe adatom it is in plane. For comparison, the so-called band energy parts of the anisotropy energy of a single layer of Fe and Co on these two substrates are also shown. The obtained results are compared to recent experimental studies using, e.g., the spin-polarized STM technique
Ab initio cluster and supercell methods are used to investigate the local geometry and optical properties of hydrogen defects in diamond. For an isolated impurity, the bond-centered site is found to be lowest in energy, and to possess both donor and acceptor levels. The neutral defect possesses a single local mode with a very small infrared effective charge, but the effective charge for the negative charge state is much larger. H+ is calculated to be very mobile with a low activation barrier. Hydrogen dimers are stable as H2* defects, which are also found to be almost IR inactive. The complex between B and H is investigated and the activation energy for the reaction B-H→B-+H+ found to be around 1.8 eV in agreement with experiment. We also investigate complexes of hydrogen with phosphorus and nitrogen. The binding energy of H with P is too low to lead to a significant codoping effect. A hydrogen-related vibrational mode of the N-H defect, and its isotopic shifts, are close to the commonly observed 3107-cm-1 line, and we tentatively assign this center to the defect. Hydrogen is strongly bound to dislocations which, together with H2*, may form part of the hydrogen accumulation layer detected in some plasma studies.
The structure and properties of the {001} planar platelet in diamond are investigated using ab initio theory. We find that a carbonaceous model, based on a layer of self-interstitials, satisfies the requirements of transmission electron microscopy, infrared absorption data, and energetic considerations. The energetics of self-interstitial production during nitrogen aggregation are considered. It is found that the growth mechanism of the platelet involves a thermally activated release of vacancies from platelets. The role of vacant sites and platelet nitrogen are also investigated and it is shown that these defects embedded within the platelet could account for the observed optical activity.
First-principles methods are used to investigate the self-interstitial and its aggregates in diamond. The experimental assignment of the spin-1 R2 EPR center to the single interstitial has been questioned because of the small fine-structure term observed. We calculate the spin-spin interaction tensor for the three interstitial defects I1〈001〉, I2NN, and I3 and compare with the experimental D tensors. The results give support for the assignments of the single and di-interstitials to microscopic models and allow us to conclusively identify a recently observed EPR center, O3, with I3. This identification, in turn, suggests a low-energy structure for I4 and a generic model for an extended defect called the platelet. We also determine the optical properties of I1〈001〉 as well as its piezospectroscopic or stress tensor and find these to be in agreement with experiment. Several multi-interstitial defects are found to possess different structural forms which may coexist. We propose that a different form of the charged I2 defect gives rise to the 3H optical peak. Several structures of the platelet are considered, and we find that the lowest-energy model is consistent with microscopic and infrared studies.
Nitrogen impurities form complexes with native defects such as vacancies and self-interstitials in silicon which are stable to high temperatures. These complexes can then suppress the formation of large vacancy and self-interstitial clusters. However, there is little known about their properties. We use first-principles density-functional theory to the determine the local vibrational modes, electrical levels and stability of a range of nitrogen-interstitial and vacancy complexes. Tentative assignments of the ABC photoluminescence line and the trigonal SL6 EPR center are made to substitutional-nitrogen pair and the substitutional-nitrogen-vacancy complex.
The linear diffusion equation is proposed to provide a macroscopic description of ionic mobility in nanostructures. This approach has been demonstrated to account for diffusion processes in nanostructured titania-based films. The formulation of a classical diffusion inverse problem and the experimental determination of concentration profiles by Rutherford backscattering spectrometry were used for the purpose. The model has allowed the measurement of the diffusion coefficient of W and Mo impurities in titania. © 2005 The American Physical Society.
Local vibrational modes of carbon impurities in relaxed Si1-xGex have been studied with infrared absorption spectroscopy in the composition range 0.05≤x≤0.50. Carbon modes with frequencies in the range 512-600 cm-1 are observed in 13C+-implanted Si1-xGex after annealing at 550°C. Measurements on samples coimplanted with 12C+ and 13C+ show that these modes originate from defects containing a single carbon atom and from the variation of the mode frequencies with composition x, the modes are assigned to substitutional carbon in Si1-xGex. Based on the frequencies obtained from a simple vibrational model, the observed modes are assigned to specific combinations of the four Si and Ge neighbors to the carbon. The intensities of the modes indicate that the combination of the four neighbors deviates from a random distribution. Ab initio local-density-functional cluster theory has been applied to calculate the structure and the local mode frequencies of substitutional carbon with n Ge and 4-n Si neighbors in a Si and a Ge cluster. The calculated frequencies are ∼9% higher than those observed, but the ordering and the splitting of the mode frequencies agree with our assignments.
Local vibrational modes of a weakly bound carbon-hydrogen complex in silicon have been identified with infrared-absorption spectroscopy. After implantation of protons at ∼20 K and subsequent annealing at 180 K, two carbon modes at 596 and 661 cm-1, and one hydrogen mode at 1885 cm-1 are observed. The three modes originate from the same complex, which is identified as bond-centered hydrogen in the vicinity of a nearby substitutional carbon atom. Ab initio theory has been applied to calculate the structure and local modes of carbon-hydrogen complexes with hydrogen located at the first, second, and third nearest bond-center site to substitutional carbon. The results support our assignment.
Measurements of the absorption spectra of brown natural type IIa diamond as well as brown nitrogen-doped CVD diamond are reported. These are largely featureless and increase almost monotonically from about 1-5.5 eV. It is argued that the brown coloration is due to an extended defect and not to a point defect. First principles modeling studies demonstrate that the spectra could be attributed to vacancy disks lying on {111} planes. Such disks are unstable above about 200 vacancies and should relax to dislocation loops in natural diamond. Hydrogen is shown to passivate the optical activity of the disks.
The ring hexavacancy (V6) has been found by previous theoretical modeling to be a particularly stable defect, but it has not been identified with any observed center to date. Here, we use ab initio calculations to derive the structure and properties of two forms of V6H2 and identify these defects with the trigonal optical centers B41 and B711, which are known to contain two hydrogen atoms in equivalent and inequivalent sites, respectively. It follows from the calculations that V6 should also be optically active and we identify it with the B804 (J line) center. This allows us to place the acceptor level of V6 at Ec-0.04 eV.
The vibrational properties of interstitial silane (SiH4)i and silyl (SiH3)i molecules in crystalline silicon are calculated using a first-principle, cluster-based, spin-polarized local-density method. The Si-H stretch modes are found to be redshifted by ∼300 cm-1 from those of the isolated molecule, which lie around 2200 cm-1. These results refute recent suggestions that modes observed around 2200 cm-1, and previously assigned to hydrogenated vacancy defects, are due to these interstitial molecules.
The vibrational modes of H2 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 Td 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 Td site. It is suggested that Raman-active modes around 4158 cm-1 are due to molecules within voids.
Nanostructured materials-the subject of much of contemporary materials research-are defined by internal interfaces, the nature of which is largely unknown. Yet, the interfaces determine the properties of nanocomposites and nanolaminates. An example is nanocomposites with extreme hardness70-90 GPa, which is of the order of, or higher than, diamond. The Ti-Si-N system, in particular, is attracting attention for the synthesis of such superhard materials. In this case, the nanocomposite structure consists of TiN nanocrystallites encapsulated in a fully percolated SiNx "tissue phase" (1 to 2 monolayers thick) that is assumed to be amorphous. Here, we show that the interfacial tissue phase can be crystalline, and even epitaxial with complex surface reconstructions. Using in situ structural analyses combined with ab initio calculations, we find that SiNx layers grow epitaxially, giving rise to strong interfacial bonding, on both TiN(001) and TiN(111) surfaces. In addition, TiN overlayers grow epitaxially on SiNx/TiN(001) bilayers in nanolaminate structures. These results provide insight into the development of design rules for new nanostructured materials.
We report on a first-principles study of all the structurally different stacking faults that can be introduced by glide along the (0001) basal plane in 3C-, 4H-, and 6H-SiC based on the local-density approximation within the density-functional theory. Our band-structure calculations have revealed that both types of stacking faults in 4H-SiC and two of the three different types of stacking faults in 6H-SiC give rise to quasi-two-dimensional energy band states in the band gap at around 0.2 eV below the lowest conduction band, thus being electrically active in n-type material. Although stacking faults, unlike point defects and surfaces, are not associated with broken or chemically perturbed bonds, we find a strong localization, within roughly 10-15 Å perpendicular to the stacking fault plane, of the stacking fault gap state wave functions. We find that this quantum-well-like feature of certain stacking faults in SiC can be understood in terms of the large conduction-band offsets between the cubic and hexagonal polytypes. Recent experimental results give qualitative support to our results.
The main purpose of this article is to determine the two-dimensional effective mass tensors of electrons confined in thin 3C wells in hexagonal SiC, which is a first step in the understanding of in-plane electron motion in the novel quantum structures. We have performed ab initio band structure calculations, based on the density functional theory in the local density approximation, for single and multiple stacking faults leading to thin 3C-like regions in 4H- and 6H-SiC and deduced electron effective masses for two-dimensional electron gases around the cubic inclusions. We have found that electrons confined in the thin 3C-like layers have clearly heavier effective masses than in the perfect bulk 3C-SiC single crystal.
First-principles calculations of twin boundaries in 3C-SiC, Si, and diamond are performed, based on the density-functional theory in the local density approximation. We have investigated the formation energies and electronic properties of isolated and interacting twin boundaries. It is found that in 3C-SiC, interacting twin boundaries which are separated by more than two Si-C bilayers are actually energetically more favorable, implying a relatively frequent appearance of these defects. The effect of the spontaneous polarization associated with the hexagonal symmetry around twin boundaries is also studied, and we have observed that the wave functions belonging to the conduction- and valence-band edge states in 3C-SiC tend to be localized almost exclusively on different sides of the faulted layers, while there is no such feature in Si or diamond.
Experimental data indicate that boron diffuses very differently in Ge than in Si. To examine the kinetics of boron diffusion, density functional calculations were performed on a variety of boron diffusion mechanisms, including interstitial and vacancy-mediated paths, as well as a correlated exchange mechanism. It was found that although vacancy and correlated exchange mechanisms possess high diffusion barriers comparable with experiment, the barrier for interstitial-mediated diffusion lies around 3.8 eV and is similar to those found for boron diffusion in Si. This estimate is well below the experimental activation energy. The difference is attributed to the failure of the theory to include the effect of electronic excitations.
Ab initio calculations were performed to study phosphorus diffusion in germanium through vacancy and interstitial-mediated mechanisms as well as a correlated exchange mechanism without interaction with a mediating defect. It was found that the most favorable diffusion mechanism is sensitive to the position of the Fermi level within the band gap. For material with a midgap Fermi level, the neutral or singly positive phosphorus interstitial is the dominant diffusing species, while in n -type material, it is the doubly negative phosphorus-vacancy complex. For a Fermi level position of Ev +0.5 eV, a barrier for phosphorus diffusion via the doubly negative phosphorus-vacancy defect of ∼2.5 eV was calculated, which is roughly ∼1 eV below the equivalent process in Si.
Large vacancy clusters, or voids, formed during crystal growth have been reported in Ge. The divacancy is a precursor to such clusters, and is believed to be stable up to 150 or 180 °C. It is also believed to form in Ge irradiated at room temperature where single vacancies are mobile. Density functional theory (DFT) cluster calculations have been performed to calculate the energy barriers for migration and dissociation of the divacancy. We find that the binding energy in the neutral charge state is ~1.5 eV and increases for negatively charged states. The migration energies were found to vary from 1.0 to 1.3 eV from the singly positive to the doubly negative charge states. These results line up well with an estimate of a migration barrier of 1.0 eV for the divacancy from experimental data. Therefore, we conclude that the divacancy in germanium will anneal by migration to trapping centers.
The structure and electrical properties of of SnV2, Sn2V, and Sn2V2 complexes in Si are investigated using first-principles cluster and supercell methods. The formation of SnV2 and Sn2V2 is found to be energetically favorable, in agreement with the experimental results. All the tin-vacancy defects are found to possess deep donor and acceptor levels, although the number of the gap states decreases with increasing size of the defect. The diffusion of tin in silicon is considered and the mechanism found to be distinct from the diffusion of group V shallow donors. In contrast with these, the Sn-V interaction is found to extend only to the third nearest neighbor distance. This implies that the activation energy for Sn diffusion via vacancies should be nearly the same as self-diffusion by this mechanism. We find an activation energy of 3.5 eV which is close to some experimental findings but considerably less than given by others.
The nitrogen-vacancy (NV) center is a paramagnetic defect in diamond with applications as a qubit. Here, we investigate its electronic structure by using ab initio density functional theory for five different NV center models of two different cluster sizes. We describe the symmetry and energetics of the low-lying states and compare the optical frequencies obtained to experimental results. We compute the major transition of the negatively charged NV centers to within 25-100 meV accuracy and find that it is energetically favorable for substitutional nitrogens to donate an electron to NV0. The excited state of the major transition and the NV0 state with a neutral donor nitrogen are found to be close in energy
Density functional theory (DFT) and low-temperature scanning tunneling microscopy (STM) have been combined to examine the bonding of individual C60 molecules on Cu(111). Energy-resolved differential-conductance maps have been measured for individual C60 molecules adsorbed on a Cu(111) surface by means of low-temperature STM, which are compared to and complemented by theoretically computed spectral images. It has been found that C60 chemisorbs with a six-membered ring parallel to the surface at two different Cu(111) binding sites that constitute two exclusive hexagonal sublattices. On each sublattice, C60 is bonded in one particular rotational conformer, i.e., C60 molecules bind to the Cu(111) surface in two different azimuthal orientations differing by 60°depending on which sublattice the binding site belongs to. The binding conformation of C60 and its orientation with regard to the copper surface can be deduced by this joint experimental-theoretical approach. Six possible pairs of C60 configurations on three different Cu surface binding sites have been identified that fulfil the requirements of the two sublattices and are consistent with all experimental and theoretical data. Theory proposes that two of these configuration pairs are most likely. We have found that DFT does not get the binding energy between rotational conformers in the correct order. We also report two different C60 monolayers on Cu(111): one with alternating orientations of neighboring molecules at low temperature and the other with (4×4) structure after annealing above room temperature.
Density-functional theory is used to assess the validity of modeling metal clusters as single atoms or rings of atoms when determining adhesion strengths between clusters and single-walled carbon nanotubes (SWNTs). Representing a cluster by a single atom or ring gives the correct trends in SWNT-cluster adhesion strengths (Fe≈Co>Ni), but the single-atom model yields incorrect minimum-energy structures for all three metals. We have found that this is because of directional bonding between the SWNT end and the metal cluster, which is captured in the ring model but not by the single atom. Hence, pairwise potential models that do not describe directional bonding correctly, and which are commonly used to study these systems, are expected to give incorrect minimum-energy structures
The structures and energies of model defects consisting of copper and hydrogen in silicon are calculated using the AIMPRO local-spin-density functional method. For isolated copper atoms, the lowest energy location is at the interstitial site with Td symmetry. Substitutional copper atoms are found to adopt a configuration with D2d symmetry. We conclude that the symmetry is lowered from Td due to the Jahn-Teller effect. Interstitial hydrogen atoms are found to bind strongly to substitutional copper atoms with an energy that is more than the difference in formation energy over the interstitial site for Cu. The resulting complex has C2v symmetry in the -2 charge state where the H atom is situated about 1.54 Å away from the Cu atom in a [100] direction. In other charge states the symmetry of the defect is lowered to Cs or C1. A second hydrogen atom can bind to this complex with nearly the same energy as the first. Two structures are found for copper dihydride complexes that have nearly equal energies; one with C2 symmetry, and the other with Cs symmetry. The binding energy for a third hydrogen atom is slightly more than for the first. Calculated electronic levels for the model defects relative to one another are found to be in fair to good agreement with experimental data, except for the copper-dihydride complex. The copper trihydride complex has no deep levels in the bandgap, according to our calculations.
This article reports the results of investigations based on local-density-functional theory into the relative formation energies for single substitutional carbon atoms in nine III-V compound semiconductors. The calculations are performed using a supercell formalism derived from the AIMPRO real-space cluster method. Only a very slight trend is discernible down the periodic table. When a metal atom is replaced with carbon, it is energetically least favorable in the phosphides, very marginally lower energy in the arsenides, and ≈0.5-0.7 eV lower in the antimonides. The situation is approximately reversed when a P, As, or Sb atom is substituted by a C atom: for the In compounds the energy is ≈0.4-0.8 eV higher than for the Al and Ga compounds.
The properties of several point defects in hexagonal gallium nitride that can compensate beryllium shallow acceptors (BeGa) are calculated using the AIMPRO method based on local density functional theory. BeGa itself is predicted to have local vibrational modes (LVM's) very similar to magnesium acceptors. The highest frequency is about 663 cm-1. Consistent with other recent studies, we find that interstitial beryllium double donors and single-donor beryllium split interstitial pairs at gallium sites are very likely causes of compensation. The calculations predict that the split interstitial pairs possess three main LVM's at about 1041, 789, and 738 cm-1. Of these, the highest is very close to the experimental observation in Be-doped GaN. Although an oxygen donor at the nearest-neighboring site to a beryllium acceptor (BeGa-ON) is also a prime suspect among defects that are possibly responsible for compensation, its highest frequency is calculated to be about 699 cm-1 and hence is not related in any way to the observed center. Another mode for this defect is estimated to be about 523 cm-1 and is localized on the ON atom. These two vibrations of BeGa-ON are thus equivalent to those for the isolated substitutional centers perturbed by the presence of their impurity partners.
The properties of defect complexes consisting of a nitrogen vacancy with a substitutional beryllium or magnesium atom on neighboring lattice sites in hexagonal GaN are calculated using the AIMPRO local-density-functional theory method. Both types of defects VN-BeGa and VN-MgGa are bound with respect to their isolated constituents. They do not appear to have any electronic levels in the bandgap, and are expected to be neutral defects. Important structural differences are found. In its minimum energy configuration, the Be atom in the VN-BeGa complex lies nearly in the same plane as the three equivalent N atoms nearest to it. Thus, it has shorter Be-N bonds than the Ga-N distance in the bulk crystal, while the Mg atom in the VN-MgGa complex occupies a position closer the lattice site of the Ga atom it replaces. Hence, the VN-BeGa complex has a larger open volume than the VN-MgGa complex. This is consistent with positron annihilation experiments [Saarinen et al., J. Cryst. Growth 246, 281 (2002); Hautakangas et al., Phys. Rev. Lett. 90, 137402 (2003)]. The frequency of the highest local vibrational mode of the VN-BeGa center is calculated to be within 3-4 % of an infrared absorption line detected in Be-doped GaN
Infrared absorption measurements on n-type silicon doped with carbon and irradiated with electrons at room temperature have revealed new absorption lines at 527.4 and 748.7 cm-1, which originate from the same defect. The 748.7-cm-1 line is observed only when the sample is cooled in the dark and the spectrum is measured through a low-pass filter with cutoff frequency below 6000 cm-1. Light with frequency above 6000 cm-1 removes this line and generates the 527.4-cm-1 line. Comparison with spectra recorded on irradiated silicon doped with 13C shows that the two lines represent local vibrational modes of carbon. The annealing behavior of the 748.7-cm-1 line is identical to that of the EPR signal originating from the negative charge state of two adjacent substitutional carbon atoms (Cs-Cs)-. The 527.4- and 748.7-cm-1 lines are ascribed to the E modes of Cs-Cs in the neutral and negative charge states, respectively. The structure and local vibrational modes of (Cs-Cs)0 and (Cs-Cs)- have been calculated by ab initio local density functional theory. The calculated structures agree qualitatively with those obtained previously by Hartree-Fock methods, but the calculated Si-C and C-C bond lengths differ somewhat. The calculated local mode frequencies are in good agreement with those observed. The formation of Cs-Cs has also been investigated. It is suggested that the center is formed when a vacancy is trapped by the metastable substitutional carbon-interstitial carbon center, Cs-Ci.
Crystalline silicon samples doped with carbon were irradiated with electrons and subsequently implanted with protons. Infrared-absorption measurements revealed local modes of hydrogen and carbon at 2967.4, 911.7, and 654.7 cm-1, which originate from the same defect. Measurements on samples codoped with different carbon and hydrogen isotopes showed that the defect contains two equivalent carbon and two equivalent hydrogen atoms. From uniaxial stress measurements, the defect is found to display trigonal symmetry. Ab initio local-density-functional theory was applied to calculate the structure and local vibrational modes of defects with pairs of equivalent carbon and hydrogen atoms. Based on these results, the observed local modes are ascribed to a defect with two adjacent substitutional carbon atoms, each of which binds a hydrogen atom located between the carbon atoms.