Advanced Nanoscale Simulation: Atoms, Materials and Devices
Sunday June 1, 2008, 8:00 am  6:00 pm, Boston, Massachusetts
Core Challenges on the Nanoscale
Nanostructured materials are defined as materials with a constituent dimension less than 100nm (0,0001mm) in at least one dimension.
Characterization
For traditional materials, the microstructure that can be seen under a conventional light microscope and the composition, i.e. the weight fractions of the different elements in the mixture, determine their properties.
For nanostructured materials, the exact composition, i.e. what atomic species are present, and configuration, i.e. where the atoms and sometimes even electrons are exactly located on the atomic scale, determine the materials properties. Examples include the exact contact structure between leads and molecules in molecular electronic devices.
Conventional tools, such as light microscopy, Xray diffraction, and mass spectrometry, cannot deliver such information. Combinations of atomiclevel modeling and atomicresolution characterization by analytical electron microscopy have shown the most promise in this area so far.
Fabrication  Process Modeling
For traditional materials, average properties, such as concentrations in continuum, or phasefield modeling, and equilibrium thermodynamics, using phase diagrams and diffusivity properites, are sufficient to model the evolution of a material during processing.
For nanostructured materials, single atoms need to be traced during the processing to predict the final structure and its properties.
However, traditional atomiclevel techniques, such as molecular dynamics, are only capable of following a few atoms reliably for times on the order of nanoseconds, i.e. not long enough for most processing steps.
The recent development of new techniques such as transitionstatetheory based methods, accelerated dynamics, such as “HyperMD”, and physicsbased Monte Carlo methods, get around this problem and enable nanoscale process modeling.
StructureProperty Relationships
For traditional materials, the properties are determined by their microstructure and composition. The structureproperty relationships are in many cases still poorly understood from a fundamental point of view, but a vast body of experience and theory exists how to change the structure to improve the properties.
For nanostructured materials, the structureproperty relationships can only be found on the atomic level and with no background experience to indicate what to expect. An example is the scaling relationship between physical dimensions and electron transport in nanoelectronics devices which strongly differs from those of conventional devices.
The theory that is currently being developed is less heuristic and more fundamental than for traditional materials. This opens the chance of designing and understanding materials more reliably with the help of computer simulations.
Materials Design
For traditional materials, revolutionary new materials are in most cases discovered by accident. Most of the “design” work is spent on
incremental improvement of known types of materials by, e.g., by alloying it with various amounts of impurities.
For nanostructured materials, computational design of hypothetical new materials is possible, including the study of stability and prediction of properties. In combination with a mature materials synthesis facility, completely new materials can be virtually designed and fabricated.
Electron Transport on the NanoScale
Traditional transport models for semiconductor devices treat electrons and holes as classical particles or a classical carrier gas. A vast software resource exists for the modeling of semiconductor devices using such classical approaches.
As the device lengths decrease down to the nanometer scale, the boundaries between device and material become blurry. Quantum mechanical effects of the underlying material structure dominate device characteristics. Quantum mechanical treatments for carrier transport and materials need to be married. This is the domain where materials science meets device engineering.
Technology Focus
Besides exciting novel capabilities and opportunities, nanostructured materials pose a number of challenges that were not present in the field of traditional materials science. Whereas trialanderror engineering was a successful approach for most of the traditional materials development, this is difficult in the field of nanoscale materials, where fundamental understanding on the atomic level is required for successful materials design. In this context, computer modeling poses incredible new chances to support new materials development for targeted design and fast timetomarket products.
Course Content
This is an intermediate level course suitable for anyone interested in computational materials design of nanoelectronics devices.
We will introduce the area of simulation techniques and tools on the nanoscale level and discuss applications and approaches with focus on advanced materials design and nanoscale engineering for nanoelectronics devices. We will address core areas of nanoscale computation, including computerassisted characterization, nanoscale process modeling, and structureproperty relationships for properties such as conductance, which governs electron transport. We will also discuss methods for ‘computational alchemy’, which means computational design of new materials with sets of desired properties.
Course Objectives
On completion of the course, you will be
 Familiar with the challenges in the design of nanostructured materials and of nanoelectronics devices
 Able to assess the role that materials computation can play to meet these challenges
 Familiar with strategies, methods, and programs for such computations and the availability of appropriate software
 Familiar with successful applications of nanoscale simulation tools and how they can serve as templates for your applications
Course Outline
Nanoscale Modeling
 What Makes a Material ‘nano’?
 The Role of Surfaces and Interfaces
 Importance of Modeling for Nanomaterials
 Nanoscale Challenges for Modeling and Simulation
Review of Singlescale Nanoscale Simulation and Modeling Methods
 Classical Molecular Dynamics Simulations
 Electronicstructure Methods with Focus on AbInitio Calculations
 Monte Carlo Simulations
 Mesoscopic Methods
 Continuum Methods
Case Studies: Simulating Simple Nanosystems
Characterization
 Concept of Simulationassisted Nanoscale Characterization
 Candidatestructure Selection
 Simulation Methods to extract atomicscale Information  from not necessarily atomicresolution  experimental techniques such as TEM images, EELS and XPS spectra, or Raman and infrared spectra
Case Studies: Nanoelectronics Devices
Fabrication  Process modeling
 Concept of Nanoscale Process Modeling
 Multiscale and Atomicscale Process Modeling
 Abinitio to Continuum
 Abinitio to Monte Carlo
 Accelerated Dynamics Methods
Case Studies: Molecular Electronics and Nanoelectronics Devices
StructureProperty Relationships in Nanoelectronic Devices
 Concept of Nanoscale Property Simulation:
 Calculation of spatially resolved band structures
 Identification of traps and charged defects
 Electron transport from electronic structure calculations
Case Studies: Nanoelectronics Devices — CNT transistors and others
Materials Design
 Concept of ‘Computational Alchemy’
 Free Energies as Measure of Stability
 Property Calculations: Mechanical, electronic, and optical
 Optimization Techniques to find materials with the optimum properties
Case Studies: Computerassisted Materials Design
Course Instructors
Wolfgang Windl, Ph.D., Associate Professor at the Ohio State University, Columbus, USA. Professor Windl works in the area of Nanoscale Computational Materials Science. His central field of expertise is in the area of atomistic simulations of materials and nanoelectronic devices, especially within densityfunctional theory. Previously, he spent four years with Motorola, first as Senior Staff Scientist in Motorola's Computational Materials Group at Los Alamos National Laboratory, and later as a Principal Staff Scientist in the Digital DNA Laboratories in Austin, Texas, where he was working in the area of multiscale modeling of semiconductor processing. Before that, he held postdoctoral positions at Los Alamos National Laboratory and Arizona State University. He received his diploma and doctoral degree in physics from the University of Regensburg, Germany. Wolfgang Windl is on the editorial board of the Journal of Computational Electronics and the Journal of Theoretical and Computational Nanoscience. Among others, he has been Chairman of the International Conference on Computational Nanoscience (www.nsti.org). Among others, he is recipient of a 2006 FraunhoferBessel Award, jointly from the Fraunhofer Society and Humboldt foundation in Germany, a 2004 Nanotechnology Industrial Impact Award, and 1998 and 1999 Patent and Licensing Awards from Los Alamos National Laboratory for his contributions to the molecular modeling code CLSMAN.
