Showing papers by "Alex Zunger published in 2004"

First-principles investigation of the assumptions underlying Model-Hamiltonian approaches to ferromagnetism of 3d impurities in III-V semiconductors

Abstract: We use first-principles calculations for transition-metal impurities V, Cr, Mn, Fe, Co, and Ni in GaAs, as well as Cr and Mn in GaN, GaP, and GaSb, to identify the basic features of the electronic structures of these systems. The microscopic details of the hole state such as the symmetry and the orbital character, as well as the nature of the coupling between the hole and the transition-metal impurity, are determined. This could help in the construction of model Hamiltonians to obtain a description of various properties beyond what current first-principles methods are capable of. We find that the introduction of a transition-metal impurity in III-V semiconductor introduces a pair of levels with t 2 symmetry-one localized primarily on the transition-metal atom, referred to as crystal-field resonance (CFR), and the other localized primarily on the neighboring anions, referred to as the dangling bond hybrid (DBH). In addition, a set of nonbonding states with e symmetry, localized on the transition-metal atom, are also introduced. Each of the levels is also spin split. Considering Mn in the host crystal series GaN→GaP→GaAs→GaSb, we find that while in GaN the hole resides in the t C F R level deep in the band gap, in GaAs and GaSb it resides in the t D B H level located just above the valence-band maximum. Thus, a DBH-CFR level anticrossing exists along this host-crystal series. A similar anticrossing occurs for a fixed host crystal (e.g., GaAs) and changing the 3d impurity along the 3d series: V in GaAs represents a DBH-below-CFR limit, whereas Mn corresponds to the DBH-above-CFR case. Consequently, the identity of the hole-carrying orbital changes. The symmetry (e vs t 2 ) and the character (DBH vs CFR), as well as the occupancy of the gap level, determine the magnetic ground state favored by the transition-metal impurity. LDA+U calculations are used to model the effect of pushing the occupied Mn states deeper into the valence band by varying U. We find that this makes the DBH state more hostlike, and at the same time diminishes ferromagnetism. While the spin-splitting of the host valence band in the presence of the impurity has been used to estimate the exchange coupling between the hole and the transition-metal impurity, we show how using this would result in a gross underestimation of the coupling.

Direct carrier multiplication due to inverse Auger scattering in CdSe quantum dots

Abstract: Many optoelectronic devices could achieve much higher efficiencies if the excess energy of electrons excited well above the conduction band minimum could be used to promote other valence electrons across the gap rather than being lost to phonons. It would then be possible to obtain two electron–hole pairs from one. In bulk materials, this process is inherently inefficient due to the constraint of simultaneous energy and momentum conservation. We calculated the rate of these processes, and of selected competing ones, in CdSe colloidal dots, using our semi-empirical nonlocal pseudopotential approach. We find much higher carrier multiplication rates than in conventional bulk materials for electron excess energies just above the energy gap Eg. We also find that in a neutral dot, the only effective competing mechanism is Auger cooling, whose decay rates can be comparable to those calculated for the carrier multiplication process.

Unusual Directional Dependence of Exchange Energies in GaAs Diluted with Mn: Is the RKKY Description Relevant?

Abstract: Ferromagnetism in Mn-doped GaAs, the prototypical dilute magnetic semiconductor (DMS), has so far been attributed to hole mediated RKKY-type interactions. First-principles calculations reveal a strong direction dependence of the ferromagnetic (FM) stabilization energy for Mn pairs, a dependence that cannot be explained within RKKY. In the limit of a hostlike hole engineered here where the RKKY model is applicable, the exchange energies are strongly reduced, suggesting that this limit cannot explain the observed ferromagnetism. The dominant contribution stabilizing the FM state is found to be maximal for -oriented Mn pairs and minimal for -oriented Mn pairs, providing an alternate explanation for magnetism in such materials in terms of energy lowering due to p-d hopping interactions, and offering a new design degree of freedom to enhance FM.

Mixed-basis cluster expansion for thermodynamics of bcc alloys

Abstract: To predict the ground-state structures and finite-temperature properties of an alloy, the total energies of many different atomic configurations $\mathbf{\ensuremath{\sigma}}\ensuremath{\equiv}{{\ensuremath{\sigma}}_{i};i=1,\dots{},N}$, with $N$ sites $i$ occupied by atom A $({\ensuremath{\sigma}}_{i}=\ensuremath{-}1)$, or B $({\ensuremath{\sigma}}_{i}=+1)$, must be calculated accurately and rapidly. Direct local-density approximation (LDA) calculations provide the required accuracy, but are not practical because they are limited to small cells and only a few of the ${2}^{N}$ possible configurations. The ``mixed-basis cluster expansion'' (MBCE) method allows to parametrize LDA configurational energetics ${E}_{\mathrm{LDA}}[{\ensuremath{\sigma}}_{i};i=1,\dots{},N]$ by an analytic functional ${E}_{\mathrm{MBCE}}[{\ensuremath{\sigma}}_{i};i=1,\dots{},N]$. We extend the method to bcc alloys, describing how to select ${N}_{\ensuremath{\sigma}}$ ordered structures (for which LDA total energies are calculated explicitly) and ${N}_{F}$ pair and multibody interactions, which are fit to the ${N}_{\ensuremath{\sigma}}$ energies to obtain a deterministic MBCE mapping of LDA. We apply the method to bcc Mo-Ta. This system reveals an unexpectedly rich ground-state line, pitting Mo-rich (100) superlattices against Ta-rich complex structures. Predicted finite-$T$ properties such as order-disorder temperatures, solid-solution short-range order and the random alloy enthalpy of mixing are consistent with experiment.

Theory of Excitonic Spectra and Entanglement Engineering in Dot Molecules

Abstract: We present results of correlated pseudopotential calculations of an exciton in a pair of vertically stacked $\mathrm{I}\mathrm{n}\mathrm{G}\mathrm{a}\mathrm{A}\mathrm{s}/\mathrm{G}\mathrm{a}\mathrm{A}\mathrm{s}$ dots. Competing effects of strain, geometry, and band mixing lead to many unexpected features missing in contemporary models. The first four excitonic states are all optically active at small interdot separation, due to the broken symmetry of the single-particle states. We quantify the degree of entanglement of the exciton wave functions and show its sensitivity to interdot separation. We suggest ways to spectroscopically identify and maximize the entanglement of exciton states.

Efficient Inverse Auger Recombination at Threshold in CdSe Nanocrystals

Abstract: We apply the semiempirical nonlocal pseudopotential method to the investigation of prospects for direct carrier multiplication (DCM) in neutral and negatively charged CdSe nanocrystals In this process, known in the bulk as impact ionization, a highly excited carrier transfers, upon relaxation to the band edge, its excess energy Δ to a valence electron, promoting it across the band gap and thus creating two excitons from one For excess energies just a few meV above the energy gap Eg (the DCM threshold), we find the following: (i) DCM is much more efficient in quantum dots than in bulk materials, with rates of the order of 1010 s-1 In conventional bulk solids, comparable rates are obtained only for excess energies about 1 eV above Eg (ii) Unlike the case in the bulk, in both neutral and charged nanocrystals the DCM rate is not an increasing function of the excess energy but oscillates as Δ moves in and out of resonance with the energy of the discrete spectrum of these 0D systems, (iii) The main contribution to the DCM rates is found to come from the dot surface, as in the case of Auger multiexciton recombination rates, (iv) Direct radiative recombination of excited electron-hole pairs and phonon-assisted decay are slower than DCM, but (v) the rate of Auger cooling (where the relaxation energy of an excited electron is used to excite a hole into deeper levels) can be of the same order of magnitude as that of the DCM process Furthermore, for excess energies well above the DCM threshold, the presence of an energy gap within the hole manifold considerably slows DCM compared to Auger cooling (AC), which is not affected by it Achieving competitive DCM processes will, therefore, require the suppression of Auger cooling, for example, by removing the hole from the dot or by trapping it at the surface

Why can CuInSe2 be readily equilibrium-doped n-type but the wider-gap CuGaSe2 cannot?

Abstract: The wider-gap members of a semiconductor series such as diamond→Si→Ge or AlN→GaN→InN often cannot be doped n-type at equilibrium. We study theoretically if this is the case in the chalcopyrite family CuGaSe2→CuInSe2, finding that: (i) Bulk CuInSe2 (CIS, Eg=1.04eV) can be doped at equilibrium n-type either by Cd or Cl, but bulk CuGaSe2 (CGS, Eg=1.68eV) cannot; (ii) result (i) is primarily because the Cu-vacancy pins the Fermi level in CGS farther below the conduction band minimum than it does in CIS, as explained by the “doping limit rule;” (iii) Cd doping is better than Cl doping, in that CdCu yields in CIS a higher net donor concentration than ClSe; and (iv) in general, the system shows massive compensation of acceptors (CdIII,VCu) and donors (ClSe,CdCu,InCu).

Efficient inverse Auger recombination at threshold in CdSe nanocrystals

Abstract: We apply the semiempirical nonlocal pseudopotential method to the investigation of prospects for direct carrier multiplication (DCM) in neutral and negatively charged CdSe nanocrystals. In this process, known in the bulk as impact ionization, a highly excited carrier transfers, upon relaxation to the band edge, its excess energy Δ to a valence electron, promoting it across the band gap and thus creating two excitons from one. For excess energies just a few meV above the energy gap Eg (the DCM threshold), we find the following: (i) DCM is much more efficient in quantum dots than in bulk materials, with rates of the order of 1010 s-1. In conventional bulk solids, comparable rates are obtained only for excess energies about 1 eV above Eg. (ii) Unlike the case in the bulk, in both neutral and charged nanocrystals the DCM rate is not an increasing function of the excess energy but oscillates as Δ moves in and out of resonance with the energy of the discrete spectrum of these 0D systems, (iii) The main contribution to the DCM rates is found to come from the dot surface, as in the case of Auger multiexciton recombination rates, (iv) Direct radiative recombination of excited electron-hole pairs and phonon-assisted decay are slower than DCM, but (v) the rate of Auger cooling (where the relaxation energy of an excited electron is used to excite a hole into deeper levels) can be of the same order of magnitude as that of the DCM process. Furthermore, for excess energies well above the DCM threshold, the presence of an energy gap within the hole manifold considerably slows DCM compared to Auger cooling (AC), which is not affected by it. Achieving competitive DCM processes will, therefore, require the suppression of Auger cooling, for example, by removing the hole from the dot or by trapping it at the surface.

Structural complexity in binary bcc ground states: The case of bcc Mo-Ta

Abstract: Traditional sorting diagrams for ground states $(T=0$ stable atomic configurations) of bcc-based binary alloys predict simple crystal structures when simple parametric interactions (e.g., first few pairs) are assumed. However, the range and magnitude of interactions for real systems is not a priori known, and could lead to much greater structural complexity. We combine a density functional theory based, deterministic mixed-basis cluster expansion with an exhaustive enumeration scheme of $3\ifmmode\times\else\texttimes\fi{}{10}^{6}$ possible structures to determine the ground states of the bcc alloy Mo-Ta. The result is a rich ground-state line, changing one's outlook on bcc structural stability. We find Mo-rich (100) superlattices (including ${C11}_{\mathrm{b}}$ and $B2)$ coexisting with complex large-cell structures $({\mathrm{Mo}}_{4}{\mathrm{Ta}}_{9}$ and ${\mathrm{Mo}}_{4}{\mathrm{Ta}}_{12}).$ We demonstrate that a systematic cluster expansion construction scheme which includes both high-order pairs and many-body figures is a necessity to capture the ground states of Mo-Ta.

Metal-dimer atomic reconstruction leading to deep donor states of the anion vacancy in II-VI and chalcopyrite semiconductors.

Abstract: First-principles total-energy calculations reveal a novel local atomic reconstruction mode around anion vacancies in II-VI and chalcopyrite compounds resulting from the formation of metal dimers. As a consequence, the neutral Se vacancy has an unexpected low symmetry in ZnSe and becomes a deep donor in both ZnSe and $\mathrm{C}\mathrm{u}\mathrm{G}\mathrm{a}\mathrm{S}{\mathrm{e}}_{2}$, contrary to the common belief regarding chalcopyrites. The calculated optical transition energies explain the hitherto puzzling absorption bands observed in the classic experiments of the color center in ZnS.

Trends in ferromagnetism, hole localization, and acceptor level depth for Mn substitution in GaN, GaP, GaAs, and GaSb

Abstract: We examine the intrinsic mechanism of ferromagnetism in dilute magnetic semiconductors by analyzing the trends in the electronic structure as the host is changed from GaN to GaSb, keeping the transition metal impurity fixed. In contrast with earlier interpretations which depended on the host semiconductor, it is found that a single mechanism is sufficient to explain the ferromagnetic stabilization energy for the entire series.

Trends in ferromagnetism, hole localization, and acceptor level depth for Mn substitution in

Abstract: We examine the intrinsic mechanism of ferromagnetism in dilute magnetic semiconductors by analyzing the trends in the electronic structure as the host is changed from GaN to GaP, GaAs and GaSb, keeping the transition metal impurity fixed. In contrast with earlier interpretations which depended on the host semiconductor, we found that a single mechanism is sufficient to explain the ferromagnetic stabilization energy for the entire series.

Electronic structure and ferromagnetism of Mn-substituted CuAlS 2 , CuGaS 2 , CuInS 2 , CuGaSe 2 , and CuGaTe 2

Abstract: The electronic and magnetic properties of Mn doping at either cation sites in the class of I--III--${\mathrm{VI}}_{2}$ chalcopyrites are studied by first-principles calculation. It is found that Mn doping at the III site provides holes and stabilizes the ferromagnetic interaction between neutral Mn defects; the neutral ${\mathrm{Mn}}_{\mathrm{Cu}}^{0}$ also stabilizes the ferromagnetism, although it provides electrons to the conduction band, instead of holes. The ferromagnetic stability is generally weaker when the cation or the anion becomes heavier in these chalcopyrites, i.e., along the sequences ${\mathrm{CuAlS}}_{2}\ensuremath{\rightarrow}{\mathrm{CuGaS}}_{2}\ensuremath{\rightarrow}{\mathrm{CuInS}}_{2}$ and ${\mathrm{CuGaS}}_{2}\ensuremath{\rightarrow}\phantom{\rule{0ex}{0ex}}{\mathrm{CuGaSe}}_{2}\ensuremath{\rightarrow}{\mathrm{CuGaTe}}_{2}.$ Interestingly, ${\mathrm{CuAlO}}_{2}$ in the chalcopyrite structure is predicted to have lower FM energy than ${\mathrm{CuAlS}}_{2}$ despite its lighter anion and shorter bonds. In general, III site substitution gives stabler ferromagnetism than Cu substitution. Thus, the preferred growth conditions are Cu-rich and III-poor, which maximize ${\mathrm{Mn}}_{\mathrm{III}}$ replacement. In n-type samples, when ${\mathrm{Mn}}_{\mathrm{III}}$ is negatively charged, the antiferromagnetic coupling is preferred. In p-type samples, the ground state of positively charged ${\mathrm{Mn}}_{\mathrm{Cu}}^{+}$ is also antiferromagnetism. The main feature of the calculated electronic properties of Mn defect at either Cu or III site is explained using a simple picture of dangling bond hybride and crystal-field resonance.

Site preference for Mn substitution in spintronicCuMIIIX2VIchalcopyrite semiconductors

Abstract: The quest for combining semiconducting with ferromagnetic properties has recently led to the exploration of Mn substitutions not only in binary (GaAs, CdTe), but also in ternary semiconductors such as chalcopyrites AB I I I X V I 2. Here, however, Mn would substitute any of the two metal sites A or B. The site preference of Mn doping in CuM I I I X V I 2 chalcopyrite is crucial because it releases different type of carriers: electrons for substitution on the Cu sites, and holes for substitution on the M I I I sites. Using first-principles calculation we show that Mn prefers the M I I I site under Cu-rich and III-poor conditions, and the Cu site under III-rich condition. We establish the chemical potential domains for pure CuAlS 2 , CuGaS 2 , CuInS 2 , CuGaSe 2 , and CuGaTe 2 stability. We show that the solubility of Mn on the M I I I (Cu) site increases (decreases) as the Fermi level moves toward the conduction-band minimum (n-type conditions). It is further found that domains of chemical stability of all these chalcopyrites may be largely reduced by Mn incorporation.

Strain-induced interfacial hole localization in self-assembled quantum dots: Compressive InAs/ GaAs versus tensile InAs/ InSb

Abstract: Using an atomistic pseudopotential approach, we study how the shape of the dot (spherical vs lens shaped) affects the position-dependent strain and the electronic properties of tensile sInAs/ InSbd and compressive sInAs/ GaAsd quantum dots We compare the strain profiles, strained modified band offsets, confined levels, and atomistic wave functions of these dots We show (i) how the existence of position-dependent strain in nonflat heterostructures can control the electronic properties, leading, for example, to interfacial localization of hole states on the interface of matrix-embedded dots and (ii) how the dots shape can control the level sequence and degeneracy For example in spherical dots, one finds degenerate light-hole (LH) and heavy-hole (HH) states, whereas in lens-shaped dots one can have as the highest-occupied hole state either (a) a LH state inside the dot, becoming a HH state outside the dot sInAs/ InSb tensile case) or (b) a HH state inside the dot, becoming a LH states outside the dot (InAs/ GaAs compressive case)

Anisotropy of interband transitions in InAs quantum wires: An atomistic theory

Abstract: The electronic and optical properties of [001]-oriented free-standing InAs cylindrical quantum wires (QWRs) with diameters 10--100 \AA{} are calculated using an atomistic, empirical pseudopotential plane-wave method. We analyze the effect of different degrees of mixing between valence bands on the optical properties of these nanostructures, by switching on and off the spin-orbit interaction. The fundamental transition in these QWRs exhibit a large anisotropy, with emission polarized prevalently along the wire axis $z$. The magnitude of such an anisotropy is found to depend on both degree of valence band mixing and wire size. In higher energy interband transitions, we find anisotropies close to 100% with emission polarized perpendicular to the wire axis. Furthermore, in large wires, transitions involving highly excited valence states show in-plane polarization anisotropies between the [110] and $[1\overline{1}0]$ directions. InAs wires can therefore switch between $z$-polarized to $xy$-polarized emission/absorption for different excitation energies. This makes them ideally suited for application in orientation-sensitive devices.

Comparison of predicted ferromagnetic tendencies of Mn substituting the Ga site in III–V’s and in I–III–VI2 chalcopyrite semiconductors

Abstract: We report density-functional calculations of the ferromagnetic (FM) stabilization energy δ=EFM−EAFM for differently oriented Mn pairs in III–V’s (GaN, GaP, GaAs) and chalcopyrite (CuGaS2, CuGaSe2, CuGaTe2) semiconductors. Ferromagnetism is found to be the universal ground state (δ GaP>GaAs, whereas in chalcopyrites it is CuGaS2>CuGaSe2>CuGaTe2. Considering both groups, the order is GaN→GaP→GaAs→CuGaS2→CuGaSe2→GaSb≈CuGaTe2. The stronger FM stabilization in III–V’s is attributed to the stronger covalent coupling between the Mn 3d and the anion p orbitals. In contrast to expectations based on Ruderman–Kittel–(Kasuya)–Yosida, (i) all Mn–Mn pair separations show FM, with no FM to antiferromagnetic oscillations and, (ii) FM is orientationally dependent, with 〈110〉 Mn–Mn pairs being the most FM.

Type I to type II transition at the interface between random and ordered domains of AlxGa1−xN alloys

Abstract: We analyze the optical and transport consequences of the existence of ordered and random domains in partially ordered samples of AlxGa1−xN alloys. Using atomistic empirical pseudopotential simulations, we find that the band alignment between random and ordered domains changes from type I to type II at x≃0.4. This leads to an increase by two to three orders of magnitude in the radiative lifetime of the electron–hole recombination. This can explain the experimentally observed mobility-lifetime product behaviors with changing Al concentration. The type I to type II transition results from a competition between the ordering-induced band folding effect and hole confinement on Ga-rich monolayers within the ordered structure.

Penetration of electronic perturbations of dilute nitrogen impurities deep into the conduction band ofGaP1−xNx

Abstract: The electronic structure consequences of the perturbations caused by dilute nitrogen impurities in $\mathrm{GaP}$ are studied by means of supercell calculations using a fully atomistic empirical pseudopotential method. We find that numerous localized states are introduced by a single $\mathrm{N}$ atom and $\mathrm{N}$ clusters, not only close to the band edge but also throughout the $\mathrm{GaP}$ conduction band, up to $\ensuremath{\sim}1\phantom{\rule{0.3em}{0ex}}\mathrm{eV}$ above the conduction band edge. These localized states suggest an alternative interpretation for a previously puzzling observation of splitting of photoluminescence excitation intensity at the $\mathrm{GaP}$ ${\ensuremath{\Gamma}}_{1c}$ energy into two features, one blueshifting and the other staying pinned in energy with increasing $\mathrm{N}$ concentration.

Theory of excitons, charged excitons, exciton fine-structure and entangled excitons in self-assembled semiconductor quantum dots

Abstract: We show how the atomistic pseudopotential many-body theory of InGaAs/GaAs addresses some important effects, including (i) the fine-structure splittings (originating from interband spin exchange), (ii) the optical spectra of charged quantum dots and (iii) the degree of entanglement in a quantum dot molecule.

Can a half-metallic zincblende-type structure be stabilized via epitaxy?

Abstract: The need for spin-injectors having the same zincblende-type crystal structure as conventional semiconductor substrates has created significant interests in theoretical predictions of possible metastable ``half-metallic'' zincblende ferromagnets, which are normally more stable in other structure-types, e.g., NiAs. Such predictions were based in the past on differences $\Delta_{\rm bulk}$ in the total-energies of the respective {\em bulk} crystal forms (zincblende and NiAs). We show here that the appropriate criterion is comparing difference $\Delta_{\rm epi}(a_s)$ in {\em epitaxial} total-energies. This reveals that even if $\Delta_{\rm bulk}$ is small, still for MnAs, CrSb, CrAs, CrTe, $\Delta_{\rm epi}(a_s) > 0$ for all substrate lattice constant $a_s$, so the zincblende phase is not stabilized. For CrS we find $\Delta_{\rm epi}(a_s) 6.2$ A and half-metallic.