Recently the concept of injection current \dotJ_inj and shift current J_sh as new optical effects has been introduced [1]. These effects are related to the optical rectification \chi_2(0;ømega,-ømega). As second-order nonlinear effects they both vanish for centrosymmetric materials, but they occur in the presence of a surface. First studies have indicated [2] that the surface-state related optical contributions dominate the spectra. Moreover, both effects reveal a high reconstruction and adsorbate sensitivity which makes them promising canditates for new surface characterization techniques.
We present calculations for both effects as well as experimental results for the injection current. As an example we have chosen the technologically important clean Si(100) surface. First-principle calculations within the DFT-LDA have been done using the p(2x2) reconstruction. The experiments on vicinal samples have been carried out using a femtosecond optical source of wavelenght 1.2...1.6 \mu m (0.78...1.0 eV). Our results are discussed in terms of the electronic structure, and conclusions about the qualitity of the measurements and the validity of our theoretical approach are made. [2.ex]
[1] J.E.~Sipe and A.I.~Shkrebtii, Phys.Rev.~B 61, 5337
(2000). [.0ex] [2] B.~Adolph and J.E.~Sipe, (submitted to
Phys.~Rev.~B).
[W30.002] Analysis of the different contributions to the rectification current density
B. Adolph, F. Nastos, J.E. Sipe (University of Toronto)
Recently a new general formalism has been presented for the second-order nonlinear response of semiconductors within the independent-particle approximation [1]. As a result it was possible to identify several different physical contributions to the optical rectification \chi_2(0;ømega,-ømega), for frequencies ømega both below and above the band gap. Here, we go a step further. For arbitrary frequencies ømega_\beta and ømega_\gamma with ømega_\beta+ømega_\gamma close to zero, we demonstrate that it is possible to reduce the general second-order nonlinear susceptibility \chi_2(-ømega_\beta-ømega_\gamma;ømega_\beta,ømega_\gamma) to single contributions depending only on either ømega_\beta or ømega_\gamma. This reduction allows to calculate \chi_2(-ømega_\beta-ømega_\gamma;ømega_\beta,ømega_\gamma) for all combinations ømega_\beta,ømega_\gamma and, thus, to indentify all different physical contributions to the general second-order nonlinear susceptibility.
Based on first-principle bandstructure calculations within the DFT-LDA, we have performed calculations of all the different contributions to the second-order nonlinear response for bulk GaAs. We discuss in detail the different physical terms as well as their different spectral behaviour.
[2.ex]
[1] J.E.~Sipe and A.I.~Shkrebtii, Phys.Rev.~B 61, 5337
(2000).
[W30.003] Second harmonic generation from single domain Si(100) surfaces
Jorge Mejia, Bernardo Mendoza (Centro de Investigaciones en Optica, León-México)
We apply a microscopic formulation based on the semi-empirical tight-binding method^1 to calculate the second harmonic spectra of a clean single-domain Si(100)c(4\times 2) surface. The dimer formed between the top-most Si atoms and the sublayer atoms give a surface with a C_1v symmetry class, characterized by ten non-zero elements of the non-linear surface susceptibility tensor \chi. This symmetry gives rise to an anisotropic second harmonic signal as a function of the azimuthal angle \phi which the incident plane of the fundamental beam of frequency ømega makes with the surface plane. In particular we focus in the spectral range of the Si E_1 critical point.
We thank support from CONACYT-México under grant 26651-E.
(1) Mendoza, Gaggiotti and Del Sole, Phys. Rev. Lett.
81, 3781 (1998).
[W30.004] Density-Matrix Description of Optical Interactions based on a Tight-Binding Representation
Verne Jacobs (Naval Research Laboratory)
A density-matrix description is employed to investigate
electromagnetic transitions of quantized electronic systems
in the presence of environmental relaxation (decoherence)
phenomena. Applications are made to atomic systems and
semiconductors (bulk crystals and heterostructures).
Time-dependent (equation-of-motion) and time-independent
(resolvent-operator) formulations are developed.
Non-equilibrium (possibly coherent) electronic-state
kinetics and homogeneous spectral-line shapes are
self-consistently determined from the time- and
frequency-domain self-energy operators. A single-electron
tight-binding representation is adopted as a starting-point
for the many-electron description of linear and non-linear
optical properties.
[W30.005] Optical Anisotropy due to Strong Linear-k Valence-Band Mixing at Heterojunctions
Bradley A. Foreman (Hong Kong University of Science and Technology)
This presentation examines linear-k terms in the \Gamma_8 valence-band Hamiltonian for heterostructures of zinc-blende--type semiconductors. In bulk crystals such terms are known to be extremely small, due to their origin as relativistic perturbations from d and f orbitals. However, in heterostructures there is a nonvanishing contribution from p orbitals. This contribution is an order of magnitude larger than the corresponding bulk term, and it gives rise to an optical anisotropy comparable to (although smaller than) that seen in recent experiments on the quantum-well Pockels effect.
[W30.006] Temperature and Carrier Density Dependent Radiative Recombination Coefficients of In_xGa_1-xAs via Screened Non-Local Exchange DFT
Clint B. Geller (Bechtel Bettis Atomic Power Laboratory, West Mifflin, PA, USA), Walter Wolf (Molecular Simulations Ltd., Cambridge, UK)
Temperature-dependent and carrier density-dependent radiative recombination coefficients for In_0.50Ga_0.50As and In_0.75Ga_0.25As have been determined using the sX-LDA (D.~M.~Bylander and L.~Kleinman, Phys.~Rev.~B 41), 7868 (1990) formalism as incorporated in the Full potential Linearized Augmented Plane Wave (FLAPW) (H.~F.~Jansen and A.~J.~Freeman, Phys.~Rev.~B 30), 561 (1984) electronic structure code. Refractive indices n(\lambda) and extinction coefficients \kappa(\lambda) are determined from the sX-LDA electronic structure as a function of temperature and carrier density via time-dependent perturbation theory including Fermi-Dirac occupation functions. Radiative recombination coefficients are determined from n(\lambda) and \kappa(\lambda) via the van Roosbroeck-Shockley Relation (W.~van Roosbroeck and W.~Shockley, Phys.~Rev.~94), 1558 (1954). For undoped In_0.50Ga_0.50As, the predicted radiative recombination coefficient of 7.28 \times 10^-10 cm^3 sec^-1 agrees with that inferred from radio frequency photoconductive decay (RFPCD) experiments on In_0.53Ga_0.47As within 6.4 %. The strong dependences of the radiative recombination coefficient of doped In_xGa_1-xAs on temperature, alloy composition and carrier density are explored using the calculated sX-LDA optical properties.
[W30.007] Anharmonic oscillator model for driven and vacuum-field Rabi oscillations in semiconductor structures
S. Rudin (U.S. Army Research Laboratory, Adelphi,Maryland 20783), T. L. Reinecke (Naval Research Laboratory, Washington D.C.20375)
We use an anharmonic oscillator model for electronic
excitations to study the dynamics of electron-hole
excitations coupled to photons in semiconductor structures.
The anharmonicity arises from the exciton-exciton
interaction. In the case of a classical applied field the
anharmonic model, unlike the harmonic oscillator model, can
describe driven Rabi oscillations of the exciton system. In
the limit of large anharmonicity a two-level description of
the electronic excitations is recovered. We also consider
the interaction with the quantized electromagnatic field in
order to apply the anharmonic model to the case of a strong
exciton-photon interaction regime in semiconductor
microcavities.
[W30.008] The Coulomb Luttinger liquid
Daw-Wei Wang (Department of Physics, University of Maryland, College Park, Maryland 20742), Andrew J. Millis (Center for Materials Theory and Department of Physics and Astronomy, Rutgers University, New Brunswick, New Jersey 08554), S. Das Sarma (Department of Physics, University of Maryland, College Park, Maryland 20742)
The effect of long-ranged Coulomb interaction on the low energy properties (momentum distribution function, density of states, electron spectral function, and 4k_F correlation function) of one-dimensional electron systems is determined theoretically. Over a wide, physically relevant energy range the behavior is found to be well described by the conventional, short-ranged interacting, Luttinger liquid results, \textitwith a scale-dependent effective exponent. An accurate empirical formula for the effective exponent is presented. We also discuss the qualitative line shape of the electron spectral function.
(D. W. Wang, A. J. Millis, and S. Das Sarma, cond-mat/0010241 (2000))
[W30.009] Dipole Interactions in the Isolated Defect Problem
Kiril Tsemekhman, Hannes Jonsson (Department of Chemistry, University of Washington)
We present the solution to the problem of artificial dipole
interactions, which appears in the repeated cell
calculations in the case of isolated defect. The defect
breaks the periodic symmetry, and replication of the defect
throughout the crystal introduces interactions between the
defects. We propose the procedure to eliminate these
interactions. The electronic states due to the defect and
the atomic (structural) relaxation are typically spatially
localized within a supercell. We construct a formalism that
allows to express the undesirable dipole interaction energy
in terms of the difference of dipole moments of the
corresponding supercells in the defect state and in the
perfect system. This difference is a well-defined quantity
contrary to the dipole moments themselves. The devised
numerical procedure allows to define the supercell in the
defect case exactly matching the one in the perfect system.
The core idea is to freeze the atoms near the supercell
boundaries in their positions in perfect system during the
molecular dynamics relaxation upon introduction of the
defect. The effective electrostatic potential is corrected
on each step of MD simulation by subtracting the
dipole-dipole terms. The procedure was successfully tested
on the self-trapped exciton configuration in alpha-quartz.
[W30.010] Local Gauge Symmetry in Empirical Tight-Binding Theory
Bradley A. Foreman (Hong Kong University of Science and Technology)
A new method for incorporating electromagnetic fields into empirical tight-binding theory is derived by applying the principle of local gauge symmetry to the empirical Hamiltonian. The coupling to the field involves no additional fitting parameters, and resolves an ambiguity found in earlier work based on the Peierls substitution.
[W30.011] Scattering of excitons by acoustic phonons in semiconductors with a complex valence band structure and multiphonon processes
S. Rudin (U.S. Army Research Laboratory), T.L. Reinecke (Naval Research Laboratory)
The scattering of excitons by acoustic phonons via the deformation potential interaction in semiconductors determines the temperature dependence of the optical linewidth in the range below 100 - 150 K, in direct band gap materials. The effects of this interaction are usually treated in the lowest order of the exciton-phonon interaction, and by assuming a simple band model for the valence band, i.e. neglecting light and heavy hole coupling. In this work we study the exciton-acoustic phonon scattering in bulk materials taking into account the complex valence band structure within the isotropic approximation for the hole band and the deformation potential for carrier-phonon scattering. We also evaluate the effects of the multiphonon processes in the scattering cross-section and relate the results to experimental results for the contribution of acoustic phonons to the exciton linewidth in GaAs.