Bulletin of the American Physical Society
2006 APS March Meeting
Monday–Friday, March 13–17, 2006; Baltimore, MD
Session W46: Compound Semiconductor Defects and Dopants |
Hide Abstracts |
Sponsoring Units: FIAP Chair: H. T. Johnson, University of Illinois, Champagne Room: Baltimore Convention Center 349 |
Thursday, March 16, 2006 2:30PM - 2:42PM |
W46.00001: Silicon-interstitials-based Benchmarking of DFT Exchange-correlation Potentials K. P. Driver, W. D. Parker, R. G. Hennig, J. W. Wilkins, C. J. Umrigar, R. Martin, E. Batista, B. Uberuaga, J. Heyd, G. Scuseria Diffusion Monte Carlo (DMC) benchmarks DFT functionals: LDA, GGA, and HSE [1]. Extensive DFT studies on single-, di-, and tri-interstitials [2] provide stable structures and converged energies. For single-interstitial formation energies, our DMC results confirm earlier work [3], with 1.5 and 1.0 eV underpredictions for LDA and GGA, respectively. We continue to observe this trend in most di- and tri-interstitials. Additionally, we find HSE reproduces DMC results for single-interstitals. Preliminary analysis indicates that large LDA and GGA discrepancies with DMC occur for highly distorted defect configurations. \begin{itemize} \item[{[1]}] J.~Heyd \emph{et al.}, J.Chem.Phys. \textbf{118}, 8207 (2003). \item[{[2]}] D.~A.~Richie \emph{et al.}, Phys. Rev. Lett. \textbf{92}, 45501 (2004). \item[{[3]}] W.~-K.~Leung \emph{et al.}, Phys. Rev. Lett. \textbf{83}, 2351 (1999). \end{itemize} [Preview Abstract] |
Thursday, March 16, 2006 2:42PM - 2:54PM |
W46.00002: Random doping and oxide roughness induced fluctuations in nanoscale semiconductor devices Petru Andrei Random doping and oxide roughness induced fluctuations in nanoscale semiconductor devices are analyzed by using self-consistent Poisson-Schr\"{o}dinger computations. A very fast and robust technique based on linearization of the transport equations is presented for the computation of fluctuations of various parameters (such as threshold voltages, terminal currents, and cutoff frequencies) of the semiconductor device. This technique is computationally much more efficient than the traditional Monte-Carlo approach and yields information on the sensitivity of device parameters fluctuations to the locations of doping and oxide thickness fluctuations. Hence, it can be used in the design of fluctuation resistant structures of semiconductor devices. Sample simulation results obtained by using the linearization technique are reported for MOSFET devices with channel lengths under 25 nm and compared with results obtained by using the Monte-Carlo technique. [Preview Abstract] |
Thursday, March 16, 2006 2:54PM - 3:06PM |
W46.00003: Thermodynamics of semiconductor doping and stoichiometry from first-principles methods Stephan Lany, Alex Zunger The theoretical investigation of semiconductor doping involves assessment of opposing physical effects: While \textit{extrinsic} doping can shift the Fermi level in the desired direction, such shifts disturb the balance between \textit{intrinsic} defects (vacancies, interstitials, etc.) leading to the creation of ``killer-defects'' and, in some materials, to considerable deviation from ideal stoichiometry. We have developed a self-consistent procedure which uses first- principles calculated formation enthalpies of impurities and intrinsic defects, to predict the correlation between doping, stoichiometry, and equilibrium carrier density. With this method, we explore the deviation from ideal stoichiometry and the electrical properties of the photovoltaic materials CuInSe$_{2}$ and CuGaSe$_{2}$, sampling the entire space of thermodynamical variables. We find a systematic correlation between stoichiometry and conductivity type, which in case of CuInSe$_{2}$ spans the whole range from $p$- to $n$-type. Application to the case of N-doping of ZnO, identifies the narrow window for growth conditions that lead to $p$-type doping by N$_{O}$ acceptors. [Preview Abstract] |
Thursday, March 16, 2006 3:06PM - 3:18PM |
W46.00004: Stoichiometry Driven Impurity Configurations in Compound Semiconductors G. Chen, I. Miotkowski, S. Rodriguez, A. K. Ramdas Precise stoichiometry and departures therefrom in the composition of the tetrahedrally coordinated compound semiconductors allow impurity incorporation in more than one configuration. Ultra-high resolution infrared spectroscopy of CdTe:O at low temperatures reveals a unique pair of sharp lines, a non-degenerate $\nu _{1}$ = 1096.78 cm$^{-1}$ and a doubly degenerate $\nu _{2}$ = 1108.35 cm$^{-1}$ at 5 K, associated with the local vibrational modes of O$_{Te}$ in a (O$_{Te}$ -- V$_{Cd})$ complex in crystals grown with (CdTe + CdO + excess Te) or (CdTe + TeO$_{2})$ which enhances the occurrence of Cd vacancy (V$_{Cd})$; in contrast, a single, triply degenerate sharp line at $\nu _{0}$ = 349.79 cm$^{-1}$ observed at 5 K occurs in CdTe grown with (CdTe + CdO + excess Cd) in which the appearance of V$_{Cd}$ is inhibited. In the former, oxygen, O$_{Te}$, is bonded to three nearest neighbor Cd's with a nearby V$_{Cd}$. The latter corresponds to O$_{Te}$ attached to all the four nearest neighbor Cd cations. With increasing temperature, $\nu _{1}$ and $\nu _{2}$ approach each other and behave as a single triply degenerate line at $\nu _{0}^{\ast }$ for temperature T $\ge $ T$^{\ast } \quad \sim $ 300 K; the uniaxial (C$_{3v})$ symmetry of (O$_{Te}$ -- V$_{Cd})$ transforms to T$_{d}$ symmetry at T$^{\ast }$, acquired due to an increasing rate of bond switching among the four possible O$_{Te}$ -- V$_{Cd}$ directions as T approaches T$^{\ast }$. [Preview Abstract] |
Thursday, March 16, 2006 3:18PM - 3:30PM |
W46.00005: \textit{Ab initio }studies of defects in CdTe and HgTe with symmetrized basis Yia-Chung Chang, Hyejung Kim We have performed \textit{ab initio} pseudopotential calculations of the total energies and atomic realxations of neutral and charged Cd vacancies in CdTe and Hg vacancies in HgTe. Our method takes advantage of the high point symmetry of the system, which enables us to use large supercells containing up to 64 atoms per unit cell. Supercells of 8, 16, 32, 54, and 64 atoms are employed and spin-orbit interactions are included. Considering only symmetric relaxations of neighboring atoms in CdTe system of a 64-atom supercell, we find that first-neighbor atoms around a Cd vacancy move toward the vacancy with a 9{\%} contraction in bond length and second-neighbor atoms move toward the vacancy with a 2.5{\%} lowering contraction in inter-atomic distance. The relaxation lowers the total energy by about 0.2eV. Similarly in HgTe, the contractions are 7{\%} and 1.5{\%}, respectively for first-neighbor and second-neighbor atoms. Three defect levels (with symmetry$\Gamma _{6}$,$\Gamma _{7}$, and$\Gamma _{8})$ are found for both CdTe and HgTe systems and one of the levels is located in the band gap while two of them are buried under the valence band maximum. Separate calculations are carried out using a full potential linearized augmented Slater-type orbital (LASTO) method and give consistent results. Self-energy corrections due to many-body effect are also estimated with the GW approach. [Preview Abstract] |
Thursday, March 16, 2006 3:30PM - 3:42PM |
W46.00006: Dislocation filtering by buffer layer interfaces in InSb/Al$_{x}$In$_{1-x}$Sb heterostructures grown on GaAs (001) substrates Madhavie EdirIsooriya, Tetsuya Mishima, Michael Santos Recent efforts have been devoted to the development of InSb-based transport devices, including mesoscopic magnetoresistors and field-effect transistors, on GaAs (001) substrates. The small effective mass of electrons in InSb leads to a high mobility at room temperature. Remotely-doped InSb quantum wells offer the additional advantage of a conducting layer that is close to the surface. One key to maximizing the performance of such devices is the reduction of dislocation and micro-twin densities induced by the large lattice mismatch between InSb/Al$_{x}$In$_{1-x}$Sb and GaAs. We investigated the dislocation filtering effects of the interfaces formed between an Al$_{y}$In$_{1-y}$Sb interlayer and an Al$_{x}$In$_{1-x}$Sb matrix layer with y$>$x. Transmission electron microscopy analysis shows that the interlayer interfaces filter out threading dislocations. We improve this dislocation filtering by optimizing the interlayer thickness, the number of interlayers, and the growth conditions. [Preview Abstract] |
Thursday, March 16, 2006 3:42PM - 3:54PM |
W46.00007: Thermodynamic model for the structure of the 90 degree partial dislocation in diamond cubic semiconductors. S. P. Beckman, D. C. Chrzan Recent studies of the 90 degree partial dislocation in diamond cubic semiconductors indicate that the structure of the core is not homogeneous, but rather is a combination of reconstructions and low energy structural excitations, such as kinks and anti-phase defects. As a result, direct investigation of the macroscopic properties of dislocation core by \textit{ab initio} methods is unfeasible. A model is presented that maps the complicated structure of a dislocation core onto a one-dimensional spin lattice. At each lattice site two spins are present, one to represent the reconstructed bonds, and the other kink structures. The model is sufficiently complex to allow expression of the essential nature of the structural excitations along a dislocation line. This Ising-like model can be investigated within a Monte Carlo framework. [Preview Abstract] |
Thursday, March 16, 2006 3:54PM - 4:06PM |
W46.00008: Strain relaxation and crystal quality in compositionally graded GaAsSb/GaAs metamorphic buffer layers Benny Perez Rodriguez, Joanna Mirecki Millunchick We have compared linearly graded, step graded, and constant composition layers of GaAs(1-x)Sb(x)/GaAs grown by Molecular Beam Epitaxy to determine which grading schemes result in the highest crystalline quality, while relaxing the lattice parameter most effectively. The incorporation rates used throughout the experiment for Ga and As were kept constant at 0.96 and 1.11 ML/s respectively. The Sb incorporation rate was varied from 0 to 0.63 ML/s to obtain a final composition of the topmost layers of x=0.5. The real-time stress evolution was obtained using an in situ multi-beam optical stress sensor. In our experiments, aggressive grading of the Sb flux results in decreased Sb incorporation at low x, a higher residual stress, and a bifurcation in the tilt of the sample. Less aggressive grading increases results in more uniform incorporation and lower residual stress, but the tilt remains. The tilt may be reduced by incorporating large steps in the grading, and completely eliminated when a constant composition layer of GaAs(0.5)Sb(0.5) is deposited directly on GaAs. The defect density of constant composition layers is somewhat higher than linearly graded layers with the same thickness and final composition. However, increasing the thickness of the layer reduces the defect density. [Preview Abstract] |
Thursday, March 16, 2006 4:06PM - 4:18PM |
W46.00009: Mutual Passivation of Donors and Isovalent Nitrogen in GaAs Jingbo Li, Pierre Carrier, Su-Huai Wei, Shu-Shen Li, Jian-Bai Xia Using large supercell total energy and band structure calculations, we have studied the mutual passivation mechanism of isovalent N and shallow donors in GaAs. We find that all the donor impurities, Si$_{Ga}$, Ge$_{Ga}$, S$_{As}$, and Se$_{As}$, bind to N in GaAs:N, which has a large N-induced band gap reduction relative to GaAs. For group-IV impurity such as Si, the formation of the nearest-neighbor Si$_{Ga}$-N$_{As}$ defect complex creates a deep donor level below the conduction band minimum (CBM). The coupling between this defect level with the CBM pushes the CBM upwards, thus restoring the GaAs band gap; the lowering of the defect level relative to the isolated Si$_ {Ga}$ shallow donor level is responsible for the increased electrical resistivity. Therefore, Si and N mutually passivate each other's electrical and optical activities in GaAs. For group-VI shallow donors such as S, the Coulomb binding between S$_{As}$ and N$_{As}$ does not form a direct bond and a deep level inside the gap; thus, no mutual passivation exists in the GaAs:(S+N) system. We also explained the difference between the mutal passivation of Si and N and the mutal passivation of H and N in GaAs. Our study provides a deep understanding of the mutual passivation mechanism and explained some of the recent puzzling experimental observations. [Preview Abstract] |
Thursday, March 16, 2006 4:18PM - 4:30PM |
W46.00010: Effect of indium on the localized vibrational mode of nitrogen in GaN$_x$As$_{1-x}$ A. M. Teweldeberhan, Stephen Fahy The effect of the substitution of nearest-neighbor gallium atoms by indium (In-N-Ga, In-N-In) on the frequency of the localized vibrational mode of substitutional nitrogen in the dilute nitride, GaN$_x$As$_{1-x}$, has been studied within first-principles density functional theory, using a supercell approach. The splitting of the highly localized triply-degenerate mode into singly- and doubly-degenerate modes is obtained and compared with available Raman and FTIR spectroscopy measurements. The results are in good agreement with the experimental values. [Preview Abstract] |
Thursday, March 16, 2006 4:30PM - 4:42PM |
W46.00011: Effects of edge and screw dislocations on optical properties of Wurtzite GaN Jeong Ho You, H.T. Johnson The wide bandgap and high temperature stability of GaN makes it a desirable material for applications such as blue light-emitting diodes, blue lasers, and high-power transistors. Despite these advantages, the large lattice mismatch in most epitaxial GaN leads to a high density of dislocations, on the order of 10$^{9}$cm$^{-2}$ for edge dislocations and 10$^{8}$cm$^{-2}$ for screw dislocations in WZ GaN. Edge dislocations are electron acceptors and take on a negative charge. Open-core screw dislocations are essentially voids, or nanopipes, in the material. The presence of these defects, plus the strain field associated with each dislocation type, change the density of states and reduce the PL intensity in typical epitaxial GaN device layers. In the present work, the effects of edge and screw dislocations in WZ GaN have been studied computationally as a function of dislocation density. Spectral properties are determined by solving a 6x6 multiband kp Hamiltonian in three-dimensions using a real-space finite element method. Results compare favorably to available experimental data. [Preview Abstract] |
Thursday, March 16, 2006 4:42PM - 4:54PM |
W46.00012: Vacancy induced localized states in graphene J.M.B. Lopes dos Santos, Vitor M. Pereira, F. Guinea, N.M.R. Peres, A.H. Castro Neto We show, analytically, that vacancies in an half-filled honeycomb lattice induce the formation of quasi-localized electronic states. If particle-hole symmetry is broken, these states become resonances close to the Fermi level. We also calculate numerically the electronic density of states for a finite density of vacancies, and discuss the issue of electronic localization in these systems. Our results have also relevance for the problem of disorder in d-wave superconductor. [Preview Abstract] |
Thursday, March 16, 2006 4:54PM - 5:06PM |
W46.00013: Low temperature photoluminescence in the strongly disordered dilute nitride GaAsN. Ivana Bosa, Dermot McPeake, Stephen Fahy Photoluminescence (PL) due to the radiative recombination of excitons is used to study the structural disorder of alloys. We calculate the PL spectra at finite temperatures in the dilute nitride GaAsN. The shape of PL spectra is determined by the exciton occupation under steady-state optical excitation. Exciton energy levels and wave functions are calculated numerically in a supercell geometry with a strong random alloy potential acting on the electron. The distribution of electrons in states of the random potential is found by solving a kinetic equation, including phonon-assisted transitions between states and radiative recombination. Results are compared to the recent experiments. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700