The g DFTB method applied to transport in Si nanowires and carbon nanotubes 1 Dip. di Ingegneria Elettronica, Universita` di Roma Tor Vergata 2 Computational.

Slides:



Advertisements
Similar presentations
by Alexander Glavtchev
Advertisements

Electronic transport properties of nano-scale Si films: an ab initio study Jesse Maassen, Youqi Ke, Ferdows Zahid and Hong Guo Department of Physics, McGill.
Atomistic Simulation of Carbon Nanotube FETs Using Non-Equilibrium Green’s Function Formalism Jing Guo 1, Supriyo Datta 2, M P Anantram 3, and Mark Lundstrom.
T.J. Kazmierski Circuit-level modelling of carbon nanotube field-effect transistors 1 MOS-AKMunich 14 September 2007 School of Electronics and Computer.
Research Projects Dr Martin Paul Vaughan available from available from
Adaptive Control of a Multi-Bias S-Parameter Measurement System Dr Cornell van Niekerk Microwave Components Group University of Stellebosch South Africa.
Nanostructures Research Group Center for Solid State Electronics Research Quantum corrected full-band Cellular Monte Carlo simulation of AlGaN/GaN HEMTs.
Carbon nanotube field effect transistors (CNT-FETs) have displayed exceptional electrical properties superior to the traditional MOSFET. Most of these.
Sohrab Ismail-Beigi Applied Physics, Physics, Materials Science Yale University Describing exited electrons: what, why, how, and what it has to do with.
A. Pecchia, A. Di Carlo Dip. Ingegneria Elettronica, Università Roma “Tor Vergata”, Italy A. Gagliardi, Th. Niehaus, Th. Frauenheim Dep. Of Theoretical.
Huckel I-V 3.0: A Self-consistent Model for Molecular Transport with Improved Electrostatics Ferdows Zahid School of Electrical and Computer Engineering.
Finite Size Effects in Conducting Nanoparticles: Classical and Quantum Theories George Y. Panasyuk Wright-Patterson Air Force Base, OH, February 8, 2011.
2D and time dependent DMRG
International Workshop of Computational Electronics Purdue University, 26 th of October 2004 Treatment of Point Defects in Nanowire MOSFETs Using the Nonequilibrium.
Screening of Water Dipoles inside Finite-Length Carbon Nanotubes Yan Li, Deyu Lu,Slava Rotkin Klaus Schulten and Umberto Ravaioli Beckman Institute, UIUC.
Stanford 11/10/11 Modeling the electronic structure of semiconductor devices M. Stopa Harvard University Thanks to Blanka Magyari-Kope, Zhiyong Zhang and.
Xlab.me.berkeley.edu Xlab Confidential – Internal Only EE235 Carbon Nanotube FET Volker Sorger.
Lattice regularized diffusion Monte Carlo
Analytical 2D Modeling of Sub-100 nm MOSFETs Using Conformal Mapping Techniques Benjamin Iñiguez Universitat Rovira i Virgili (URV), Tarragona, E-43001,
Outline Introduction – “Is there a limit?”
Full-band Simulations of Band-to-Band Tunneling Diodes Woo-Suhl Cho, Mathieu Luisier and Gerhard Klimeck Purdue University Investigate the performance.
Large-Scale Density Functional Calculations James E. Raynolds, College of Nanoscale Science and Engineering Lenore R. Mullin, College of Computing and.
Modern VLSI Design 3e: Chapter 2 Copyright  1998, 2002 Prentice Hall PTR Topics n Derivation of transistor characteristics.
INAC The NASA Institute for Nanoelectronics and Computing Purdue University Circuit Modeling of Carbon Nanotubes and Their Performance Estimation in VLSI.
Efficient solution algorithm of non-equilibrium Green’s functions in atomistic tight binding representation Yu He, Lang Zeng, Tillmann Kubis, Michael Povolotskyi,
Deformation of Nanotubes Yang Xu and Kenny Higa MatSE 385
Мэдээллийн Технологийн Сургууль Монгол Улсын Их Сургууль Some features of creating GRID structure for simulation of nanotransistors Bolormaa Dalanbayar,
ATK 方法的扩展及其应用 王雪峰 苏州大学物理系 ZJNU-JinHua.
NEGF Method: Capabilities and Challenges
Algorithms and Software for Large-Scale Simulation of Reactive Systems _______________________________ Ananth Grama Coordinated Systems Lab Purdue University.
VFET – A Transistor Structure for Amorphous semiconductors Michael Greenman, Ariel Ben-Sasson, Nir Tessler Sara and Moshe Zisapel Nano-Electronic Center,
ENEE 704 Summary Final Exam Topics. Drift-Diffusion 5 Equations, Five Unknowns. – n, p, Jn, Jp,  Solve Self-Consistently Analytical Method: – Equilibrium:
ICECS, Athens, December /18 From nanoscale technology scenarios to compact device models for ambipolar devices Sébastien Frégonèse, Cristell Maneux,
Laboratoire Matériaux et Microélectronique de Provence UMR CNRS Marseille/Toulon (France) - M. Bescond, J-L. Autran, M. Lannoo 4 th.
The Geometry of Biomolecular Solvation 2. Electrostatics Patrice Koehl Computer Science and Genome Center
Stefano Sanvito Computational Spintronics Group Physics Department, Trinity College Dublin CCTN’05, Göteborg 2005.
Feb 23, 2010, Tsukuba-Edinburgh Computational Science Workshop, Edinburgh Large-Scale Density-Functional calculations for nano-meter size Si materials.
Simulation of transport in silicon devices at atomistic level Introduction Properties of homogeneous silicon Properties of pn junction Properties of MOSFET.
Scaling of the performance of carbon nanotube transistors 1 Institute of Applied Physics, University of Hamburg, Germany 2 Novel Device Group, Intel Corporation,
1 Recent studies on a single-walled carbon nanotube transistor Reference : (1) Mixing at 50GHz using a single-walled carbon nanotube transistor, S.Rosenblatt,
Conduction and Transmittance in Molecular Devices A. Prociuk, Y. Chen, M. Shlomi, and B. D. Dunietz GF based Landauer Formalism 2,3 Computing lead GF 4,5.
Junction Capacitances The n + regions forms a number of planar pn-junctions with the surrounding p-type substrate numbered 1-5 on the diagram. Planar junctions.
UPoN Lyon 2008 G. Albareda 1 G.Albareda, D.Jimenez and X.Oriols Universitat Autònoma de Barcelona - Spain E.mail: Can analog and.
Influence of carrier mobility and interface trap states on the transfer characteristics of organic thin film transistors. INFM A. Bolognesi, A. Di Carlo.
Vanderbilt MURI meeting, June 14 th &15 th 2007 Band-To-Band Tunneling (BBT) Induced Leakage Current Enhancement in Irradiated Fully Depleted SOI Devices.
Computational Aspects of Multi-scale Modeling Ahmed Sameh, Ananth Grama Computing Research Institute Purdue University.
Quantum pumping and rectification effects in interacting quantum dots Francesco Romeo In collaboration with : Dr Roberta Citro Prof. Maria Marinaro University.
U Tor Vergata Charge transport in molecular devices Aldo Di Carlo, A. Pecchia, L. Latessa, M.Ghorghe* Dept. Electronic Eng. University of Rome “Tor Vergata”,
Development of an analytical mobility model for the simulation of ultra thin SOI MOSFETs. M.Alessandrini, *D.Esseni, C.Fiegna Department of Engineering.
IEE5328 Nanodevice Transport Theory
Fundamentals of DFT R. Wentzcovitch U of Minnesota VLab Tutorial Hohemberg-Kohn and Kohn-Sham theorems Self-consistency cycle Extensions of DFT.
Scattering Rates for Confined Carriers Dragica Vasileska Professor Arizona State University.
F. Sacconi, M. Povolotskyi, A. Di Carlo, P. Lugli University of Rome “Tor Vergata”, Rome, Italy M. Städele Infineon Technologies AG, Munich, Germany Full-band.
Network for Computational Nanotechnology (NCN) UC Berkeley, Univ.of Illinois, Norfolk State, Northwestern, Purdue, UTEP First Time User Guide to MOSCAP*
Quantum Capacitance Effects In Carbon Nanotube Field-Effect Devices
Nano and Giga Challenges in Microelectronics Symposium and Summer School Research and Development Opportunities Cracow Sep , 2004 Afternoon 4: Carbonanotubes.
Determination of surface structure of Si-rich SiC(001)(3x2) Results for the two-adlayer asymmetric dimer model (TAADM) are the only ones that agree with.
Fatemeh (Samira) Soltani University of Victoria June 11 th
Sandipan Dutta and James Dufty Department of Physics, University of Florida Classical Representation of Quantum Systems Work supported under US DOE Grants.
Trap Engineering for device design and reliability modeling in memory/logic application 1/ 년 02 월 xx 일 School of EE, Seoul National University 대표.
Analysis of Strain Effect in Ballistic Carbon Nanotube FETs
Contact Resistance Modeling and Analysis of HEMT Devices S. H. Park, H
DNA Functionalization of Carbon Nanotubes: Application in Device Design Patrick Bennett.
Contact Resistance Modeling in HEMT Devices
Lecture 7 DFT Applications
The Materials Computation Center. Duane D. Johnson and Richard M
Carbon Nanotubes Adam Charnas.
Nonequilibrium Green’s Function with Electron-Phonon Interactions
Algorithms and Software for Large-Scale Simulation of Reactive Systems
Algorithms and Software for Large-Scale Simulation of Reactive Systems
Presentation transcript:

The g DFTB method applied to transport in Si nanowires and carbon nanotubes 1 Dip. di Ingegneria Elettronica, Universita` di Roma Tor Vergata 2 Computational Material Science, Universitat Bremen Alessandro Pecchia 1 L. Latessa 1, Th. Frauenheim 2, A. Di Carlo 1

ACS - San Francisco 2006 Retarded (r) and advanced (a) Green functions are defined as follow Let us write H and G in a block form NEGF + DFTB = g DFTB Self-energies Device region Lead HDHD HLHL HRHR  LD  RD -- ++

ACS - San Francisco 2006 Non equilibrium density In order to compute V(r) we need the local  (r) Density Matrix We can build the n.e. density matrix LL RR

ACS - San Francisco 2006 Self-consistent loop Density matrix ρ Multigrid Poisson solver Self-consistent solutions Evaluation of Green’s function External potential Hartree term Exchange- correlation (LDA) + +

ACS - San Francisco 2006 Implementation details Is performed by countour integration and has been parallelized (MPI) All matrices stored in dense format Green’s functions computed by direct inversions Sparse storage Implementation of a block-iterative construction Old gDFTB ( ) New gDFTB (2006-)

ACS - San Francisco 2006 Block-iterative algorithm The device G.F. are computed with an iterative algorithm 1) Computation of partial Green’s 2) Computation of equilibrium Green’s 3) Additional blocks needed for non-equilibrium H PL

ACS - San Francisco 2006 Profiling SWCNT(20,0): 2880 atoms, 36 Principal Layers 2,5 nm 14,7 nm 60,2 5, Tempo (s) Old gDFTB New gDFTB Peak Memory (MB) Old gDFTB New gDFTB Single node (P GHz), Single energy point

ACS - San Francisco 2006 Self-consistent potential eV

ACS - San Francisco 2006 Poisson Equation The Poisson’s equation is solved with a 3D Multi-grid algorithm. Discretize in real space This allows to solve complex boundary conditions (bias, gate) 2-terminals gated coaxially-gated 4-terminals

ACS - San Francisco 2006 projection back in AO Now we need to project the solution into the local basis set Can be viewed as an approximation of the rigorous matrix elements of V(r). This is consistent with standard DFTB

ACS - San Francisco 2006 Summary g DFTB implementation  Construction of H and S directly in sparse format  Solution of Green’s functions via block-iterative methods  Current Bottle-neck: Poisson equation - Very efficient for 1D type systems - Memory scales linearly (depends on PL size) - Can be used for O(N) calculations even in equilibrium - Considerable speed-up and memory save - Dense matrices never allocated - Multigrid with uniform grid, dense storage! - Need to implement more efficient methods (finite elements with adaptive grids)

ACS - San Francisco 2006 Applications of g DFTB Molecular Electronics. Incoherent Transport and Inelastic Tunneling Spectroscopy A. Pecchia et al., Nano Lett. 4, 2109 (2004) G. Solomon et al., J. Chem Phys 124, (2006) A. Pecchia, A. Di Carlo, Introducing molecular electronics, Springer Series, (2005) A. Pecchia, A. Di Carlo, Molecular Electronics: Analysis design and simulations, Elsevier (2006)

ACS - San Francisco 2006 Applications to CNT and SiNW

ACS - San Francisco 2006 Coaxially gated CNT VDVD VS=0VS=0 VGVG Semiconduct ing CNT Insulator (ε r =3.9) 10 nm 1.5 nm x y z CNT contact (INFINEON - Düsberg)

ACS - San Francisco 2006 Atomic Forces V GATE = 5 V Ang. GATE Forces [Ang] Application of V G changes CNT diameter

ACS - San Francisco 2006 Screening problem Quantum correction to the induced charge V ext Distance VGVG CNT electron gas CNT is not able to accumulate the electronic charge to completely screen the gate bias (λ > electron gas extension) λ CNT completely screens the external field. Classical electrostatics: charge induced on the CNT is λ

ACS - San Francisco 2006 HOWEVER: In a CNT the DOS is not the only contribution. Many body correction should be considered (XC) Screening in CNT: DOS limit Why charge induction is limited? DOS [Latessa et al., Phys. Rev. B 72, (2005)] Pauli exclusion principle limits the induced electrons to the number allowed by filling the DOS

ACS - San Francisco 2006 Many-body corrections Compressibility of an interacting electron gas Compressibility of an non-interact electron gas  Compressibility Capacity

ACS - San Francisco 2006 Evidence of Negative K Eisenstein, Pfeiffer and West, PRL 68, 674 (1992) Eisenstein, Pfeiffer and West, PRB 50, 1760 (1994) N (10 11 cm -2 ) NcNc Compressibility of 2D electron gas Thomas-Fermi screening In 1D systems things can be more complicated because of D(E) Including exchange

ACS - San Francisco 2006 Negative C Q in CNTs Overscreening, C Q <0 [Latessa et al., Phys. Rev. B 72, (2005)]

ACS - San Francisco 2006 Negative C Q in CNTs Critical charge density n TS TOTAL SCREENING High density limit : PARTIAL SCREENING C Q approaches e 2 DOS(E F ) gDFTB calculation C Q proportional to DOS Low density limit: OVER-SCREENING Fit to analytic model Chalmers, PRB 52, (‘95) Fogler, PRL 94, (2005)

ACS - San Francisco 2006 XC in DFTB Hubbard and e-e repulsion integrals T.A. Niehaus, PRA 71, (2004)

ACS - San Francisco 2006 Diameter Dependence CNT (13,0) CNT (10,0)CNT (7,0)

ACS - San Francisco 2006 Output characteristics h V DS < 0 p p i E F,S E F,D Drain Source “Electrostatic saturation” L. Latessaet al.: IEEE Trans. Nanotechnol., in press

ACS - San Francisco 2006 Small sub-threshold swing (theoretical limit for silicon MOSFET: 60 mV/dec) I on /I off ~ 10 8 Unipolar behavior Trans-characteristics

ACS - San Francisco 2006 Band-to-well tunneling Generation of confined states in a quantum well

ACS - San Francisco 2006 SiNW FET SiO 2 shells has been removed and silicon is terminated with H D. D. D. Ma. et al., Science, vol. 299, pp , 2003 L. J. Lauhon, et al., Nature, vol. 420, pp , Coaxially gated Si nanowire FET

ACS - San Francisco 2006 Geometry relaxations d<10 nm 10<d<20 nm d>30 nm 1.22 nm 0.87 nm

ACS - San Francisco 2006 Device geometry P doped region Intrinsic region oxide 1.2 nm (2.4 nm) 7.7 nm 3.6 nm 6 nm Drain Source

ACS - San Francisco 2006 CNT vs SiNW CNT-FETSiNW-FET 6 nm

ACS - San Francisco 2006 Differences in S Coax. gated (7,0) CNTFET SiNW – FET |V GS | (V) Current, I DS (A) S = 180 mV/dec S = 75 mV/dec

ACS - San Francisco 2006 Conclusions of part II Atomistic Density Functional approach can be extended to account for current transport in molecular devices by using self-consistent non-equilibrium Green function. We use g DFTB is a good compromise between simplicity and reliability but there is room for improvement. The use of a Multigrid Poisson solver allows for study very complicated device geometries CNT and Silicon Nanowire FET has been studied with g DFTB Quantum capacitance in CNT is governed by XC Gate control in SiNW FETs is more delicate

ACS - San Francisco 2006 Surface Green’s function The surface G.F. is computed by iteration (decimation technique)   H PL g > Converged surface Green’s function Lopez Sancho et al., J. Phys. F: Met. Phys (1984); ibid., (1985)