Session D32: Focused Session: Computational Nanoscience II

The effect of polytype on energy gap in SiC nano-clusters

The size dependence of energy gap is perhaps the most remarkable aspect of quantum confinements in low dimensional systems. Numerous models have been proposed to describe the quantum confined electronic states in Si, CdSe, etc, providing a precise description of the bandgap as a function of nano-crystal dimensions. Recently, ab-initio studies of energy gap in cubic SiC nanoparticles as a function of both size and surface composition have been reported. SiC is a remarkable semiconductor with over 200 different crystal structures (polytypes) due to different stacking orders. The always-indirect band gap of bulk SiC varies substantially among the different polytypes (2.4 eV for 3C-SiC to 3.3 eV for 2H-SiC). We have investigated the effect of polytypism on the size-dependency of energy gap in SiC nano-clusters using ab-initio calculations. For clusters smaller than 1 nm, all SiC polytypes show identical energy gap-size dependencies. For SiC nano-particles larger than 1 nm, the effect of crystal structure becomes apparent approaching the bulk values.   

Structural and electronic properties of InP nanowires: role of surface dangling bonds on nanowire facets

InP nanowires are one of intriguing targets in the sources and detectors in fiber optic communications and high-speed electronic applications. In spite of this technological importance, however, understanding of atomic structures and electronic properties of InP nanowires still remain unclear. Here, we present first principles pseudopotential calculations that clarify structural stability and electronic properties of InP nanowires vertically grown on InP(111) substrates. Our calculations with diameter less than 23 \AA\ demonstrate that the nanowires with zinc blende structure are less favorable than those with wurtzite structure, in which the surface dangling bonds on nanowire facets are found to be crucial to determine the stability. An analysis of the nanowire cohesive energy based on the number of the dangling bonds predicts that the nanowires are bistable forming both wurtzite and zinc blende structures at large diameter around 120 \AA, which leads to the exhibition of polytypes being consistent with experiments. In addition, the calculated Kohn-Sham energy bands for stable wurtzite nanowires show that the surface dangling bond states determine the band character and the gap energy for the nanowires with diameter less than 9 \AA\ while the gap energy of nanowires with diameter larger than 14 \AA\ is dependent on only the nanowire size.   

Atomistic Simulations of Long-Range Strain and Close-Range Electronic Structure in Self-Assembled Quantum Dot Systems.

The electronic structure in self-assembled quantum dots depends on the detailed quantum dot configuration inside the embedding matrix and the nearest neighbor quantum dots through electronic and strain interactions. However, realistic determination of strain requires a large computational domain. To tackle this problem for an embedded InAs quantum dot NEMO-3D uses the atomistic VFF Keating model containing up to 64 million atoms. Interatomic distance changes obtained are used to influence the sp3d5s* tight-binding electronic Hamiltonian with 21 million atoms. Targeted eigenstates with correct symmetry are found reliably even in such large systems. Our investigations show a dramatic dependence of the dot states on the size of the strain domain and the boundary conditions. NEMO-3D is also used to study the electronic states in coupled quantum dots in stacks of 2 and 7 dots. There is an interesting interplay between strain induced and size induced state distributions.   

Numerical modeling InAs/GaAs quantum ring capacitance spectroscopy using non-parabolic approximation.

Direct observation for discrete energy spectra of InAs/GaAs quantum dots (QD) and rings (QR) is possible by capacitance-voltage (CV) end far infrared (FIR) spectroscopy [1]. Existing theoretical explanations of experimental data are limited by the parabolic potential model [1,2] and infinite confinement, which they are using. In present work a single subband model for InAs/GaAs QD(QR) is used. The finite confinement band-gap potential is estimated by the band gap difference of InAs quantum object and GaAs substrate [3]. The non-parabolic approximation is defined by electron effective mass dependence on the confinement energy according to the Kane formula. The 3D confined energy problem is solved numerically by the finite element method. Obtained results for single electron levels are in good agreement with the CV spectroscopy. The calculations also reproduce experimental value for the energy-gate-voltage conversion coefficient equal to 7. Our estimation for the magnitude of the electron effective mass agrees with experimental data. The results are compared with the parabolic potential model calculations. \newline [1] A. Lorke, et al., PRL \textbf{84} 2223 (2000). \newline [2] A. Emperador, et al., PR B.\textbf{ 62} 4573 (2000). \newline [3] I. Filikhin et al., Modelling Simul. Mater. Sci. Eng. \textbf{12 }1121 (2004).   

A charge patching method calculation of a quantum dot/quantum well nanosystem.

First principles density functional calculations typically involve finding self-consistent solution to the Kohn-Sham equations, scaling with the cube of system size. To study large systems, such as semiconductor nanocrystals, an approximate ab initio potential may be constructed by patching together local charge motifs determined from self-consistent calculations on small prototype systems, and the eigenvalues determined using the folded spectrum method for a few band-edge states. In this talk, I will discuss the recent applications of this method to CdS/CdSe/CdS colloidal quantum dot quantum wells. Results on the effect of core, well, and shell thicknesses on the wavefunction and optical properties will be discussed. We find the conduction band wavefunction to be significantly less confined to the CdSe quantum well layer than predicted by k.p theory, and discuss the implications of this result on the theoretical interpretation of recent time-resolved Faraday rotation experiments. We will also briefly discuss the extensions of this approach to the explicit treatment of surface ligand effects and transition-metal doped nanocrystals.   

Discrete size series of CdSe quantum dots: A combined computational and experimental investigation

Ab initio computational studies were performed for CdSe nanocrystals over a wide range of sizes in conjunction with recent experimental work. Substantial relaxations and coordination of surface atoms were found to play a crucial role in determining the nanocrystal stability and optical properties. While optimally $($three-fold$)$ coordinated surface atoms resulted in stable closed-shell structures with large optical gaps, sub-optimal coordination gave rise to lower stability and negligible optical gaps. These computations are in qualitative agreement with recent chemical etching experiments suggesting that closed shell nanocrystals contribute strongly to photoluminescence quantum yield while clusters with non-optimal surface coordination do not.   

Excitonic absorption of million-atom (In,Ga)As/GaAs self-assembled quantum dots

We calculate the optical absorption spectra of (In,Ga)As/GaAs self-assembled quantum dots by adopting an atomistic pseudopotential approach to calculate the single-particle electron and hole confined states of the dot followed by a calculation of the neutral exciton $X^0$ states $|\Psi^{(\nu)}(X^0)\rangle$ based on the configuration-interaction approach. We predict three types of {\em allowed} transitions that would be naively expected to be forbidden. (i) Transitions involving low-lying electron and hole states that are forbidden in simple effective mass models (e.g. $1S$-$2S$, $1S$-$1P$) become allowed by virtue of single-particle band-mixing. (ii) Transitions involving a deep hole state, with a mixture of heavy-hole and light-hole character, and an electron in the lowest state are found to have oscillator strengths that are comparable in magnitude to those of the expected allowed transitions. Thus, simple models based on single-band envelope functions cannot predict these transitions. (iii) Transitions whose intensity is large due to many-body borrowing of oscillator strength from allowed transitions. The transitions in (i) and (ii) appear, respectively, as low-energy and high-energy satellites of the allowed {\em P}-{\em P} transitions, as observed in PLE.   

Optimized Configuration Interaction Method for Electronic Excitations in Nanostructures

The Configuration Interaction (CI) method has been widely used to calculate electronic excitations in semiconductor nanostructures. The main drawbacks of this method are its slow convergence with the number of configurations and the difficulty to select a-priori the most relevant configurations. Here we report a new approach for the selection of the CI basis set: For a given number N$_{C}$ of configurations, we use heuristic search methods to find the set of N$_{C}$ configurations that minimizes the excited-state energy of the nanostructure. We demonstrate this method for single excitons and biexcitons in CdSe quantum dots. We show that the best configurations not only are different from what one would expect based on the single-particle energy ladder, but also, they often do not correspond to the configurations that have the largest contribution to the full CI wave function. We also show that a few (less than 100) optimized configurations provide excitation energies with accuracy comparable to much larger (10,000 or more) non-optimized configuration basis sets.   

Molecular models for nanoporous amorphous carbons via a novel monte carlo algorithm

Nanoporous carbons (NPCs) are interesting amorphous phases of carbon that can exhibit very favorable gas permeation selectivity, which is an area of research. However, computational studies of such amorphous, nonequilibrium phases are hindered by the lack of a robust method to generate candidate molecular structures that validate against the known properties of such materials. A new monte carlo algorithm has been developed to create structural models for amorphous carbons. The simulation method mimics the experimental preparation of nanoporous carbons (NPC) by pyrolysis from polyfurfuryl alcohol as a guideline. The resulting molecular structures exhibits properties that compare favorably to those observed experimentally for real NPCs. These atomistic NPC models are approaching a realistic representation of NPCs used for gas separations and as such, are being used to study the diffusion of small gas molecules in these materials. Comparisons are made to other methods in the literature and possible improvements are discussed.   

Solvophobic Solvation Forces between Nanoparticles: Size, Shape, and Solvent Effects

The hydrophobic effect is an important phenomenon that can play a central role in biological and colloidal systems. Although hydrophobic hydration has been extensively studied, this is just a specialized case of the more general phenomenon of ``solvophobic solvation'', which can occur in a wide variety of applications involving solutes in non-aqueous solutions, including colloid and polymer suspensions and assemblies. We report results from molecular-dynamics simulations of solvophobic nanoparticles immersed in n-decane liquid. Analogous to aqueous systems, we observe dewetting in the inter-particle region and attractive solvation forces when the particle separation becomes smaller than a critical value. We show that the critical separation can be affected by particle size and shape, and, in contrast to predictions by theories of hydrophobic hydration; it can be larger for small particles than for large ones. The complex size and shape dependence of solvophobic inter-particle forces allows new prospects for creating selective nonaqueous colloidal assemblies.   

The Formation of Haeckelite Structures Induced by Vacancy Defects in Graphene Layers of Carbon Nanotube

The formation of haeckelite structures induced by vacancy defects in graphene layers of carbon nanotube are investigated by tight-binding molecular dynamics (TBMD) simulations and by first principles total energy calculations. It is observed in the TBMD simulations that two single vacancies coalesce into a 5-8-5 double vacancy at the temperature of 3,000 K, and it is further reconstructed into a new defect structure, the 555-777 defect, by the Stone-Wales type transformation at higher temperatures. First principles calculations confirm that the 555-777 defect is energetically much more stable than two separated single vacancies, and the energy of the 555-777 defect is also slightly lower than that of the 5-8-5 double vacancy. In TBMD simulation, it is also found that the four single vacancies reconstruct into two collective 555-777 defects which is the unit for the hexagonal haeckelite structure proposed by Terrones et al. [Phys. Rev. Lett. {\bf 84}, 1716 (2000)]   

Viscoelastic Properties of Model Polymer Nanocomposites

The study of the dynamics of polymers in the presence of nano-scale particles is not only of fundamental interest but is also key to developing the tools necessary to engineer polymer nanocomposite materials. In this work, we investigate the effect of nanoscopic filler particles on the viscoelastic behavior of a model polymer melt. Molecular Dynamics simulations of bead-spring polymer systems containing roughly spherical nano-filler particles were performed. Equilibrium and non-equilibrium methods were used to determine the stress relaxation and shear-dependent viscosity of both the neat melt and filled systems. We will discuss the results of these simulations, as well as an interesting correlation between certain structural quantities and the observed mechanical reinforcement.