Contents

• DISCLAIMER
• 1. Introduction
  • 1.1 Citation
  • 1.2 User Feedback
  
  
• 2. Getting Started
  • 2.1 Input File Structure
  • 2.2 Simple Input File -- SCF geometry optimization
  • 2.3 Water Molecule Sample Input File
  • 2.4 Input Format and Syntax for Directives
    • 2.4.1 Input Format
    • 2.4.2 Format and syntax of directives
  
  
• 3. NWChem Architecture
  • 3.1 Database Structure
  • 3.2 Persistence of data and restart
  
  
• 4. Functionality
  • 4.1 Molecular electronic structure
  • 4.2 Relativistic effects
  • 4.3 Pseudopotential plane-wave electronic structure
  • 4.4 Molecular dynamics
  • 4.5 Python
  • 4.6 Parallel tools and libraries (ParSoft)
  
  
• 5. Top-level directives
  • 5.1 START and RESTART -- Start-up mode
  • 5.2 SCRATCH_DIR and PERMANENT_DIR -- File directories
  • 5.3 MEMORY -- Control of memory limits
  • 5.4 ECHO -- Print input file
  • 5.5 TITLE -- Specify job title
  • 5.6 PRINT and NOPRINT -- Print control
  • 5.7 SET -- Enter data in the RTDB
  • 5.8 UNSET -- Delete data in the RTDB
  • 5.9 STOP -- Terminate processing
  • 5.10 TASK -- Perform a task
    • 5.10.1 TASK Directive for Electronic Structure Calculations
    • 5.10.2 TASK Directive for Special Operations
    • 5.10.3 TASK Directive for the Bourne Shell
    • 5.10.4 TASK Directive for QM/MM simulations
  • 5.11 CHARGE -- Total system charge
  • 5.12 ECCE_PRINT -- Print information for Ecce
  
  
• 6. Geometries
  • 6.1 Keywords on the GEOMETRY directive
  • 6.2 SYMMETRY -- Symmetry Group Input
  • 6.3 Cartesian coordinate input
  • 6.4 ZMATRIX -- Z-matrix input
  • 6.5 ZCOORD -- Forcing internal coordinates
  • 6.6 Applying constraints in geometry optimizations
  • 6.7 SYSTEM -- Lattice parameters for periodic systems
  
  
• 7. Basis sets
  • 7.1 Basis set library
  • 7.2 Explicit basis set definition
  • 7.3 Combinations of library and explicit basis set input
  
  
• 8. Effective Core Potentials
  • 8.1 Scalar ECPs
  • 8.2 Spin-orbit ECPs
  
  
• 9. Relativistic All-electron Approximations
  • 9.1 Douglas-Kroll approximation
  • 9.2 Dyall's Modified Dirac Hamitonian approximation
  
  
• 10. Hartree-Fock or Self-consistent Field
  • 10.1 Wavefunction type
  • 10.2 SYM -- use of symmetry
  • 10.3 ADAPT - symmetry adaptation of MOs
  • 10.4 TOL2E -- integral screening threshold
  • 10.5 VECTORS -- input/output of MO vectors
    • 10.5.1 Superposition of fragment molecular orbitals
    • 10.5.2 Atomic guess orbitals with charged atoms
  • 10.6 Accuracy of initial guess
  • 10.7 THRESH -- convergence threshold
  • 10.8 MAXITER -- iteration limit
  • 10.9 PROFILE -- performance profile
  • 10.10 DIIS -- DIIS convergence
  • 10.11 DIRECT and SEMIDIRECT -- recomputation of integrals
    • 10.11.1 Integral File Size and Format for the SCF Module
  • 10.12 SCF Convergence Control Options
  • 10.13 NR -- controlling the Newton-Raphson
  • 10.14 LEVEL -- level-shifting the orbital Hessian
  • 10.15 Orbital Localization
  • 10.16 Printing Information from the SCF Module
  • 10.17 Hartree-Fock or SCF, MCSCF and MP2 Gradients
  
  
• 11. DFT for Molecules (DFT)
  • 11.1 Specification of Basis Sets for the DFT Module
  • 11.2 VECTORS and MAX_OVL -- KS-MO Vectors
  • 11.3 XC and DECOMP -- Exchange-Correlation Potentials
    • 11.3.1 Exchange-Correlation Functionals
    • 11.3.2 Combined Exchange and Correlation Functionals
    • 11.3.3 Meta-GGA Functionals
  • 11.4 Sample input file
  • 11.5 ITERATIONS -- Number of SCF iterations
  • 11.6 CONVERGENCE -- SCF Convergence Control
  • 11.7 SMEAR -- Fractional Occupation of the Molecular Orbitals
  • 11.8 GRID -- Numerical Integration of the XC Potential
    • 11.8.1 Angular grids
    • 11.8.2 Partitioning functions
    • 11.8.3 Radial grids
    • 11.8.4 Grid Scheme
  • 11.9 TOLERANCES -- Screening tolerances
  • 11.10 DIRECT and NOIO -- Hardware Resource Control
  • 11.11 ODFT and MULT -- Open shell systems
  • 11.12 SIC -- Self-Interaction Correction
  • 11.13 MULLIKEN -- Mulliken analysis
  • 11.14 BSSE -- Basis Set Superposition Error
  • 11.15 Print Control
  
  
• 12. Spin-Orbit DFT (SODFT)
• 13. COSMO
• 14. GAPSS-no longer distributed with NWChem
  • 14.1 Specification of Basis Sets for the GAPSS Module
  • 14.2 CORRELATION and EXCHANGE -- Exchange-Correlation Potentials
    • 14.2.1 Optional Functionals
  • 14.3 Numerical Integration
  • 14.4 Summations and Integrations for a Periodic System
  • 14.5 SCF Iterative Procedure
  
  
• 15. CIS, TDHF, and TDDFT
  • 15.1 Overview
  • 15.2 Performance of CIS, TDHF, and TDDFT methods
  • 15.3 Input syntax
  • 15.4 Keywords of TDDFT input block
    • 15.4.1 CIS and RPA -- the Tamm-Dancoff approximation
    • 15.4.2 NROOTS -- the number of excited states
    • 15.4.3 MAXVECS -- the subspace size
    • 15.4.4 SINGLET and NOSINGLET -- singlet excited states
    • 15.4.5 TRIPLET and NOTRIPLET -- triplet excited states
    • 15.4.6 THRESH -- the convergence threshold of Davidson iteration
    • 15.4.7 MAXITER -- the maximum number of Davidson iteration
    • 15.4.8 TARGET and TARGETSYM-- the target root and its symmetry
    • 15.4.9 SYMMETRY -- restricting the excited state symmetry
    • 15.4.10 ALGORITHM -- algorithms for tensor contractions
    • 15.4.11 FREEZE -- the frozen core/virtual approximation
    • 15.4.12 PRINT -- the verbosity
  • 15.5 Sample input
  
  
• 16. Tensor Contraction Engine Module: CI, MBPT, and CC
  • 16.1 Overview
  • 16.2 Performance of CI, MBPT, and CC methods
  • 16.3 Algorithms of CI, MBPT, and CC methods
    • 16.3.1 Spin, spatial, and index permutation symmetry
    • 16.3.2 Runtime orbital range tiling
    • 16.3.3 Dynamic load balancing parallelism
    • 16.3.4 Parallel I/O schemes
  • 16.4 Input syntax
  • 16.5 Keywords of TCE input block
    • 16.5.1 HF, SCF, or DFT -- the reference wave function
    • 16.5.2 CCSD,CCSDT,CCSDTQ,CISD,CISDT,CISDTQ, MBPT2,MBPT3,MBPT4, etc. -- the correlation model
    • 16.5.3 THRESH -- the convergence threshold of iterative solutions of amplitude equations
    • 16.5.4 MAXITER -- the maximum number of iterations
    • 16.5.5 IO -- parallel I/O scheme
    • 16.5.6 DIIS -- the convergence acceleration
    • 16.5.7 FREEZE -- the frozen core/virtual approximation
    • 16.5.8 PRINT -- the verbosity
  • 16.6 Sample input
  
  
• 17. MP2
  • 17.1 FREEZE -- Freezing orbitals
  • 17.2 TIGHT -- Increased precision
  • 17.3 PRINT and NOPRINT
  • 17.4 VECTORS -- MO vectors
  • 17.5 RI-MP2 fitting basis
  • 17.6 FILE3C -- RI-MP2 3-center integral filename
  • 17.7 RIAPPROX -- RI-MP2 Approximation
  • 17.8 Advanced options for RI-MP2
    • 17.8.1 Control of linear dependence
    • 17.8.2 Reference Spin Mapping for RI-MP2 Calculations
    • 17.8.3 Batch Sizes for the RI-MP2 Calculation
    • 17.8.4 Energy Memory Allocation Mode: RI-MP2 Calculation
    • 17.8.5 Local Memory Usage in Three-Center Transformation
  • 17.9 One-electron properties and natural orbitals
  
  
• 18. Multiconfiguration SCF
  • 18.1 ACTIVE -- Number of active orbitals
  • 18.2 ACTELEC -- Number of active electrons
  • 18.3 MULTIPLICITY
  • 18.4 SYMMETRY -- Spatial symmetry of the wavefunction
  • 18.5 STATE -- Symmetry and multiplicity
  • 18.6 VECTORS -- Input/output of MO vectors
  • 18.7 HESSIAN -- Select preconditioner
  • 18.8 LEVEL -- Level shift for convergence
  • 18.9 PRINT and NOPRINT
  
  
• 19. Selected CI
  • 19.1 Background
  • 19.2 Files
  • 19.3 Configuration Generation
    • 19.3.1 Specifying the reference occupation
    • 19.3.2 Applying creation-annihilation operators
    • 19.3.3 Uniform excitation level
  • 19.4 Number of roots
  • 19.5 Accuracy of diagonalization
  • 19.6 Selection thresholds
  • 19.7 Mode
  • 19.8 Memory requirements
  • 19.9 Forcing regeneration of the MO integrals
  • 19.10 Disabling update of the configuration list
  • 19.11 Orbital locking in CI geometry optimization
  
  
• 20. Coupled Cluster Calculations
  • 20.1 MAXITER -- Maximum number of iterations
  • 20.2 THRESH -- Convergence threshold
  • 20.3 TOL2E -- integral screening threshold
  • 20.4 DIISBAS -- DIIS subspace dimension
  • 20.5 FREEZE -- Freezing orbitals
  • 20.6 IPRT -- Debug printing
  • 20.7 PRINT and NOPRINT
  • 20.8 Methods (Tasks) Recognized
  • 20.9 Debugging and Development Aids
    • 20.9.1 Switching On and Off Terms
  
  
• 21. Geometry Optimization with DRIVER
  • 21.1 Convergence criteria
  • 21.2 Available precision
  • 21.3 Controlling the step length
  • 21.4 Maximum number of steps
  • 21.5 Discard restart information
  • 21.6 Regenerate internal coordinates
  • 21.7 Initial Hessian
  • 21.8 Mode or variable to follow to saddle point
  • 21.9 Optimization history as XYZ files
  • 21.10 Print options
  
  
• 22. Geometry Optimization with STEPPER
  • 22.1 MIN and TS -- Minimum or transition state search
  • 22.2 TRACK -- Mode selection
  • 22.3 MAXITER -- Maximum number of steps
  • 22.4 TRUST -- Trust radius
  • 22.5 CONVGGM, CONVGG and CONVGE -- Convergence criteria
  • 22.6 Backstepping in STEPPER
  • 22.7 Initial Nuclear Hessian Options
  
  
• 23. Hybrid Calculations with ONIOM
  • 23.1 Real, model and intermediate geometries
    • 23.1.1 Link atoms
    • 23.1.2 Numbering of the link atoms
  • 23.2 High, medium and low theories
    • 23.2.1 Basis specification
    • 23.2.2 Effective core potentials
    • 23.2.3 General input strings
  • 23.3 Use of symmetry
  • 23.4 Molecular orbital files
  • 23.5 Restarting
  • 23.6 Examples
    • 23.6.1 Hydrocarbon bond energy
    • 23.6.2 Optimization and frequencies
    • 23.6.3 A three-layer example
    • 23.6.4 DFT with and without charge fitting
  
  
• 24. Hessians
  • 24.1 Hessian Module Input
    • 24.1.1 Defining the wavefunction threshold
    • 24.1.2 Profile
    • 24.1.3 Print Control
  
  
• 25. Vibrational frequencies
  • 25.1 Vibrational Module Input
    • 25.1.1 Hessian File Reuse
    • 25.1.2 Redefining Masses of Elements
    • 25.1.3 Animation
    • 25.1.4 An Example Input Deck
  
  
• 26. DPLOT
  • 26.1 GAUSSIAN -- Gaussian Cube format
  • 26.2 TITLE -- Title directive
  • 26.3 LIMITXYZ -- Plot limits
  • 26.4 SPIN -- Density to be plotted
  • 26.5 OUTPUT -- Filename
  • 26.6 VECTORS -- MO vector file name
  • 26.7 WHERE -- Density evaluation
  • 26.8 ORBITAL -- Orbital sub-space
  • 26.9 Examples
    • Charge Density
    • Molecular Orbital
  
  
• 27. Electron Transfer Calculations with ET
  • 27.1 VECTORS -- input of MO vectors for ET reactant and product states
  • 27.2 FOCK/NOFOCK -- method for calculating the two-electron contribution to the ET Hamiltonian
  • 27.3 SWITCH2E -- overlap threshold which determines the method for calculating the two-electron contribution to the ET Hamiltonian
  • 27.4 TOL2E -- integral screening threshold
  • 27.5 Example
  
  
• 28. Properties
  • 28.1 Subdirectives
    • 28.1.1 Nbofile
  
  
• 29. Electrostatic potentials
  • 29.1 Grid specification
  • 29.2 Constraints
  • 29.3 Restraints
  
  
• 30. Prepare
  • 30.1 Default database directories
  • 30.2 System name and coordinate source
  • 30.3 Sequence file generation
  • 30.4 Topology file generation
  • 30.5 Appending to an existing topology file
  • 30.6 Generating a restart file
  
  
• 31. Molecular dynamics
  • 31.1 Introduction
    • 31.1.1 Spacial decomposition
    • 31.1.2 Topology
    • 31.1.3 Files
    • 31.1.4 Databases
    • 31.1.5 Force fields
  • 31.2 Format of fragment files
  • 31.3 Creating segment files
  • 31.4 Creating sequence files
  • 31.5 Creating topology files
  • 31.6 Creating restart files
  • 31.7 Molecular simulations
  • 31.8 System specification
  • 31.9 Restarting and continuing simulations
  • 31.10 Parameter set
  • 31.11 Energy minimization algorithms
  • 31.12 Multi-configuration thermodynamic integration
  • 31.13 Time and integration algorithm directives
  • 31.14 Ensemble selection
  • 31.15 Velocity reassignments
  • 31.16 Cutoff radii
  • 31.17 Polarization
  • 31.18 External electrostatic field
  • 31.19 Constraints
  • 31.20 Long range interaction corrections
  • 31.21 Fixing coordinates
  • 31.22 Special options
  • 31.23 Autocorrelation function
  • 31.24 Print options
  • 31.25 Periodic updates
  • 31.26 Recording
  • 31.27 Program control options
  
  
• 32. Analysis
  • 32.1 System specification
  • 32.2 Reference coordinates
  • 32.3 File specification
  • 32.4 Selection
  • 32.5 Coordinate analysis
  • 32.6 Essential dynamics analysis
  • 32.7 Trajectory format conversion
  • 32.8 Electrostatic potentials
  
  
• 33. Combined quantum and molecular mechanics
• 34. File formats
• 35. Pseudopotential plane-wave density functional theory (NWPW)
  • 35.1 PSPW Tasks
    • 35.1.1 Simulation Cell
    • 35.1.2 (
    • 35.1.3 DPLOT
    • 35.1.4 Wannier
    • 35.1.5 Self-Interaction Corrections
    • 35.1.6 Point Charge Analysis
    • 35.1.7 Car-Parrinello
    • 35.1.8 PSP_GENERATOR
    • 35.1.9 WAVEFUNCTION_INTITIALIZER
    • 35.1.10 V_WAVEFUNCTION_INITIALIZER
    • 35.1.11 WAVEFUNCTION_EXPANDER
    • 35.1.12 STEEPEST_DESCENT
  • 35.2 Band Tasks
    • 35.2.1 Brillouin Zone
  • 35.3 PAW Tasks
  • 35.4 Pseudopotential and PAW basis Libraries
  • 35.5 NWPW RTDB Entries and DataFiles
    • 35.5.1 Ion Positions
    • 35.5.2 Ion Velocities
    • 35.5.3 ANALYSIS: Mulliken RTDB data
    • 35.5.4 Wavefunction Datafile
    • 35.5.5 Velocity Wavefunction Datafile
    • 35.5.6 Formatted Pseudopotential Datafile
    • 35.5.7 One-Dimensional Pseudopotential Datafile
    • 35.5.8 PSPW Car-Parrinello Output Datafiles
  • 35.6 Car-Parrinello Scheme for Ab Initio Molecular Dynamics
    • 35.6.1 Verlet Algorithm for Integration
    • 35.6.2 Constant Temperature Simulations: Nose-Hoover Thermostats
  • 35.7 PSPW Tutorial 1: Minimizing the geometry for a C molecule
  • 35.8 PSPW Tutorial 2: Running a Car-Parrinello Simulation
  • 35.9 PSPW Tutorial 3: Running a Simulated Annealing Car-Parrinello Simulation
  • 35.10 PSPW Tutorial 4: optimizing a unit cell and geometry for Silicon-Carbide
  • 35.11 Band Tutorial 1: Minimizing the energy of a silicon-carbide crystal by running a PSPW and Band simulation in tandem
  • 35.12 PAW Tutorial
  • 35.13 NWPW Capabilities and Limitations
  • 35.14 Questions and Difficulties
  
  
• 36. Controlling NWChem with Python
  • 36.1 How to input and run a Python program inside NWChem
  • 36.2 NWChem extensions
  • 36.3 Examples
    • 36.3.1 Hello world
    • 36.3.2 Scanning a basis exponent
    • 36.3.3 Scanning a basis exponent revisited.
    • 36.3.4 Scanning a geometric variable
    • 36.3.5 Scan using the BSSE counterpoise corrected energy
    • 36.3.6 Scan the geometry and compute the energy and gradient
    • 36.3.7 Reaction energies varying the basis set
    • 36.3.8 Using the database
    • 36.3.9 Handling exceptions from NWChem
    • 36.3.10 Accessing geometry information -- a temporary hack
    • 36.3.11 Scaning a basis exponent yet again -- plotting and handling child processes
  • 36.4 Troubleshooting
  
  
• 37. Interfaces to Other Programs
  • 37.1 NBO -- Natural Bond Orbital Analysis
  • 37.2 DIRDYVTST -- DIRect Dynamics for Variational Transition State Theory
    • 37.2.1 Introduction
    • 37.2.2 Files
    • 37.2.3 Detailed description of the input
  
  
• 38. Acknowledgments
• A. Standard Basis Sets
• B. Sample input files
  • B.1 Water SCF calculation and geometry optimization in a 6-31g basis
    • B.1.1 Job 1. Single point SCF energy
    • B.1.2 Job 2. Restarting and perform a geometry optimization
  • B.2 Compute the polarizability of Ne using finite field
    • B.2.1 Job 1. Compute the atomic energy
    • B.2.2 Job 2. Compute the energy with applied field
  • B.3 SCF energy of HCO using ECPs for C and O
  • B.4 MP2 optimization and CCSD(T) on nitrogen
  
  
• C. Examples of geometries using symmetry
  • C.1 methanol
  • C.2 water
  • C.3 acetylene
  • C.4 ethylene
  • C.5 methane
  • C.6 buckminsterfullerene
  • C.7 porphyrin
  • C.8 iron penta-carbonyl
  • C.9 sodium crown ether
  • C.10 ammonia
  • C.11 benzene
  • C.12
  • C.13 ferrocene
  • C.14
  • C.15 trans-dichloroethylene
  • C.16
  • C.17 cyclopentadiene anion
  • C.18 gold tetrachloride
  
  
• D. Running NWChem
  • D.1 Sequential execution
  • D.2 Parallel execution on UNIX-based parallel machines including workstation clusters using TCGMSG
  • D.3 Parallel execution on UNIX-based parallel machines including workstation clusters using MPI
  • D.4 Parallel execution on MPPs
  • D.5 IBM SP
  • D.6 Cray T3E
  • D.7 Linux
  • D.8 Alpha systems with Quadrics switch
  • D.9 Windows 98 and NT
  • D.10 Tested Platforms and O/S versions
  
  
• About this document ...