My research activities focus mainly on the development of electronic structure and quantum transport models and methodologies for nanostructures including quantum dots, carbon nanotubes, silicon nanowires and biosensors, semiconductor devices and analog circuit design, simulations and characterization, numerical algorithms, large-scale parallel cluster computing and building user interfaces for community nanotechnology application packages. Listed below is a description of my research activities and interests.
Post-Doctoral Research Associate (Spring, 2005−)
School of Electrical and Computer Engineering, Purdue University, West Lafayette, Indiana, USA.
1. Development of a comprehensive full three-dimensional quantum simulator for nanowires, nanotubes, molecular devices, and other novel nanoscale semiconductor devices. The simulator will have the capability of simulating different material systems with different geometric structures/configurations. Different schemes will be implemented for electrostatics and transport kernels nature of which range from purely classical to full quantum mechanical. Important features of the simulator: (a) Electrostatics being solved on the very atomistic level on the fundamental zincblende/diamond/graphine lattice using both finite element and finite difference schemes. (b) New algorithms for quantum transport based on non-equilibrium Green’s function such as basis-set reduction technique, contact block reduction, fast inverse using nested dissection, wavefunction decomposition, divide and conquer methodology, Sancho-Rubio algorithm and recursive algorithm. (c) Fully parallel and portable applications using Message passing Interface (MPI) and OpenMP and open source packages. (d) State-of-art graphical user interface using in-house rapid application infrastructure (RAPPTURE). (d) Open source code and application development using SVN repository system, and (e) Availability of the application through web-based infrastructure on www.nanohub.org.
2. Study on Charge and Spin -based Quantum Qubit Systems using the NEMO 3D simulator. This is a package that currently allows calculating single-particle electronic states of semiconductor nanostructures. With an appropriate choice of boundary conditions, it can be used for calculating electronic structures of bulk, quantum dots, quantum wires, nanocrystals etc. Summary of Current NEMO 3D capabilities: NEMO 3D includes spin in its fundamental atomistic tight binding representation. Spin is therefore not added in as an afterthought into the theory, but spin-spin interactions are naturally included in the Hamiltonian. Effects of interaction with external electromagnetic fields are also included. NEMO 3D enables the computation of strain and electronic structure in an atomistic basis for over 64 and 23 million atoms, corresponding to volumes of (110nm)3 and (77nm)3, respectively. These volumes can be spread out arbitrarily over thin layer geometry. For example if a thin layer of 15 nm height is considered, the corresponding widths in the x-y plane correspond to 298nm for strain calculations and 178nm for electronic structure calculations. No other atomistic tool can currently handle such volumes needed for realistic device simulations. NEMO 3D runs on serial and parallel platforms, local cluster computers as well as the NSF Teragrid. About 400,000 atoms are treated efficiently on a single 32bit CPU. NEMO 3D uses an atomistic valence force field method (strain) and the empirical tight binding method (electronic structure). A 64 million atom strain simulation requires less than 9 hours to complete on 48 Itanium CPUs of the NSF Teragrid. A 23 million atom electronic structure calculation requires about 12 hours on 64 CPUs. NEMO 3D has been publicly released by JPL and is freely available without any export restrictions. Calculations are currently limited to an electronic structure analysis and carrier transport is not included in the model right now. This application package is being used to advance the experimental construction of P-impurity-based qubit gates through atomistic modeling within collaboration between the Australian Research Council Centre of Excellence for Quantum Computer Technology (CQCT) and Purdue University. NEMO 3D can model systems of 64 million atoms, which is large enough to cover the interaction between multiple P impurities. NEMO 3-D will answer fundamental questions that cannot be answered without an atomistic approach. These questions address issues of P impurity in Si descriptions in realistic gated potentials, actual sensitivity to donor placement, strain engineering to mitigate donor placement sensitivity, gate control, spin read-out, shark-shift control, and conductance spectroscopy. The fundamental questions related to a Si:P system that need to be addressed are grouped into classes of charge qubits, spin qubits, and conductance spectroscopy: (1) Charge qubit: (a) basic P and P-P+ description and energy scales (b) verify extent of actual sensitivity to relative donor placement (c) verify gate control and E-V characteristics. (2) Spin qubit: (a) Spin readout: adiabatic transition to P− state (b) Stark shift control of hyperfine coupling (c) Basic P-P system and exchange coupling J in two electron description. (3) Spectroscopy through transport: I-V characteristics of tunnel devices. (4) Resolution of architecture issues such as nearest neighbor interaction based computation.
3. Study of Symmetry Breaking and Fine Structure Splitting in Self-Assembled Zincblende Quantum Dots through Atomistic Simulations of Long-Range Strain and Piezoelectric Field. The symmetry in quantum dots realized from III-V materials is lowered due to two fundamental symmetry breaking mechanisms: (a) the underlying crystal, which lacks inversion symmetry, and (2) the presence of strain which induces piezoelectric charges. Strain in self-assembled quantum dots is a long-range phenomenon and the dot states depend strongly on both the size of the strain domain and the boundary conditions. Both sources of symmetry breaking influence the fine structure splitting (splitting of the bright exciton) in self-assembled quantum dots. Results show a significant dependence of the dot states and optical polarization on the size of the strain domain and the boundary conditions.
4. Non-equilibrium Green’s function (NEGF) simulation and study of impacts of defects on the transport properties of single-walled metallic Carbon nanotubes. The simulator for the realistic 3D CNT field effect transistors (CNTFETs) is based on Finite Element Method (FEM) for the electrostatic treatment and nearest neighbor Tight-Binding (TB) method for electron transport, employing Non-Equilibrium Green’s Function (NEGF) techniques. The major findings have been that the presence of a single vacancy locally modulates the LDOS significantly in the device. More importantly, regardless of the chirality of the nanotube, the transmission is reduced throughout the entire energy spectrum (by one quantum unit in some regions). We are currently employing this approach to model the aberrant nature of 1/f noise in metallic nanotubes in collaboration with an experimental group at Purdue University.
5. Study of bandstructure and crystal orientation effects in novel nanoscale devices.
6. Development and implementation of novel algorithms and methodologies for solving non-equilibrium Green’s functions from large sparse matrix systems. Collaborating with Stanford University, NASA Ames Research Center, University of Alabama and Purdue University in order to achieving the goal. Different parallel computing methodologies using MPI and OpenMP are being studied and implemented in these applications for achieving optimum performance gain.
7. Modeling and characterization of low frequency noise in nanoscale MOSFETs. The effects of Halo tilt angle and dose and energy of implants on the 1/f noise characteristics are the focus of this study. Modeling is being done with a full three dimensional quantum corrected Monte Carlo simulator while the experimental data are received from partners in Intel and Texas Instruments.
8. Cooperating with the NCN (Network for Computational Nanotechnology) team towards providing and developing generalized numerical computational tools for use in the community software in the nanohub. These tools, software and scientific applications will be available and expected to be of great use and importance to the scientific research communities all over the world.
9. Collaborated with Rensselaer Polytechnic Institute, NASA Marshall Space Flight Center, Georgia State University, NASA Jet Propulsion Laboratory, California Institute of Technology in preparing a proposal for NASA Research Opportunities in Space and Earth Sciences-2005 (ROSES) on Development of a Far Infrared Simulator and Device for Exploration Applications.
10. Extending the thermalized effective potential approach to sub-band modeling to simulate thin SOI and FinFET devices including quantum transport effects in the lateral direction. A new proposal has been in line that focus on the development of a meshless simulator based on the classical Fast Multipole Method (FMM) to correctly predict transport of ions through arbitrary biological protein channels.
Graduate (Ph.D.) Research Assistant (2001−2004)
Department of Electrical Engineering, Arizona State University, Tempe, Arizona, USA.
1. Developed a parameter-free quantum field approach for use in conjunction with particle-based simulations. The method is based on a perturbation theory around thermodynamic equilibrium and leads to a quantum field formalism in which the size of an electron depends upon its energy. The approach when used in the simulations of a conventional nanoscale 25 nm n-channel MOSFET device is found to produce correct experimentally verified threshold voltage shifts of about 220 mV and drain current degradation of about 30%. To further test the applicability, the quantum field formalism is used to calculate the threshold voltage and output characteristics of recently proposed single-gated (SG) and dual-gated (DG) fully-depleted silicon-on-insulator (FDSOI) devices. It is observed that the method quite correctly retrieves the trend in the threshold voltage shift with the variation of silicon film thickness. The simulation results are verified with the available experimental and/or theoretical data.
2. Developed a full 3D Monte-Carlo particle based simulator coupled with an effective potential scheme to investigate the quantum effects occurring in recently proposed narrow-width SOI devices. From the results obtained, it is found that the two-dimensional (2D) carrier confinement in the device structure results not only in a significant increase in the threshold voltage but also in its pronounced channel width dependency.
3. Investigations of the impact of unintentional/discrete doping on the performance of SOI devices. To treat the Coulomb (electron-ion and electron-electron) interactions properly, three different but consistent real-space molecular dynamics (MD) schemes have been implemented: the particle-particle-particle-mesh (P3M) method, the corrected Coulomb approach and the Fast Multipole Method (FMM). It is believed that the FMM algorithm has been used for the first time in the simulations of semiconductor devices. Demonstrated is the fact that unintentional doping in the channel region leads to considerable fluctuations in the device intrinsic parameters. Strong correlation between the location of the impurity/dopant atom and the magnitude of the drain current was observed and the underlying reasons were analyzed within a quantum-mechanical framework.
4. Developing a self-consistent event-biasing scheme for statistical enhancement in the Monte-Carlo device simulations. Enhancement algorithms are especially useful when the device behavior is governed by rare events in the carrier transport process. The method is used for the simulations of subthreshold conduction in a 15 nm n-channel MOSFET. It is found that event-biasing experiments recover precisely the physical averages and the self-consistent field and reduces the time necessary for computation of the desired device characteristics.
5. Modeling and simulation of Focused Ion Beam MOSFET (FIBMOS) devices using quasi-3D Monte-Carlo particle based simulator coupled with an effective potential scheme. It is found that the short-channel effects are suppressed in the FIBMOS structures compared to their conventional counterparts.
6. Modeling and simulation of Schottky Junction Transistors (SJTs) using quasi-3D Monte-Carlo particle based simulator. Gate tunneling was modeled by an Airy function approach. A nonperturbative mode of surface roughness scattering was included in the simulator. The device was found to offer higher mobility and transconductance than its conventional counterpart.
7. Study and simulation of ballistic transport in mesoscopic devices using basically the transfer matrix formalism. Classical solution of the electrostatic confinement through 3D Poisson solver was combined with the Landauer’s formalism to calculate the on-state current quantum mechanically (treating the device as a quantum wire with a constriction).
8. Worked on developing Hydrodynamic and Extended Drift-Diffusion models to simulate Thin Film Electro-Luminescent (TFEL) devices.
9. HgFET Pseudo-MOSFET (ψ-MOSFET) characterization/measurement/modeling of silicon-on-insulator (SOI) material by Four Dimensions CV Map 92-B System allowing threshold voltage, electron and hole mobility, doping density, oxide charge, interface trap density, etc. to be determined. A SIMOX wafer (p-type, 8.5-14 ohm-cm) was used in this study.
Short- and Long- Term Research Plan
1. Continue working on the projects in collaboration with my current collaborator.
2. Collaborating with the computation electronics group of Arizona State University on a proposal to Study the self-heating effects in contemporary SOI devices. The work would integrate THERMAL3D simulations (on a circuit level) with Monte Carlo BTE solvers (on the device level) that utilize microscopic phonon-theory to calculate the phonon scattering rates and the appropriate relaxation constants.
3. Implementation of gate tunneling model with the non-equilibrium Green’s functions framework for the simulations of Carbon nanotube based Flash memory cells.
4. Study and design bio-sensors based on NEMO 3D package.
5. Study coupled quantum dot systems for cascaded Laser applications.
6. Investigation of the origin of finite hole g-factor in self-assembled InAs quantum dots through atomistic simulations.
Noteworthy Presentations
Green's Function
SOI Devices Characterization