Quantum ESPRESSO Calculations - Advanced Options Dialog Box

Set options for running periodic DFT calculations. The options are saved as preferences and used in subsequent calculations.

  • Features
  • Additional Resources

Quantum ESPRESSO Calculations - Advanced Options Dialog Box Features

The options are contained in several tabs. Some tabs have the same options, but these options are applied in different parts of the job. The variables set in the input file by the options are reported in parentheses. The tabs present may vary according to the calculation that is being done.

Theory tab

Set options for the theoretical treatment used in the calculations.

Spin treatment option menu

Choose a spin treatment from this option menu:

  • Non-polarized—non-polarized calculation (equivalent to closed-shell, spin-restricted).
  • Spin-polarized—spin-polarized calculation with magnetization along the z axis (equivalent to spin-unrestricted).
  • Non-collinear—spin-polarized calculation with magnetization not confined to the z axis.
  • Spin-orbit—calculation with spin-orbit coupling included via pseudopotentials in the noncollinear formalism.
DFT+U treatment option menu

Use this menu to turn on DFT+U treatment (sets lda_plus_u, and lda_plus_u_kind = 0). Note that DFT+U currently works only for selected elements (the d and f block elements and C, H, O, N, As, Ga, In). You will need to set the Hubbard U (Hubbard_U) and the Hubbard J0 (Hubbard_J0) parameters in the Initial Parameters dialog box (click Initial Parameters in the main QE panel).

Density functional type option menu

Choose the type of density functional to use in the calculations. The types available are: LDA, GGA, VDW, Hybrid, Range-corrected. When you choose the functional type, the Density functional option menu is loaded with the available functional names. Note that hybrid and range-corrected functionals are much more resource-intensive than the other types, so you should only run calculations with these functionals on small systems. Hybrid and range-corrected functionals can only be used for single point calculations and atomic position optimizations, and cannot be used for property calculations. For atomic position optimizations with a pseudopotential and a hybrid functional, you can only use norm-conserving (NC) pseudopotentials.

Density functional option menu

Choose a density functional of the type specified on the Density functional type option menu. When choosing a pseudopotential, you should try to choose one that was generated with the functional you chose. The supported functionals are:

  • LDA (Local Density Approximation): PZ
  • GGA (Generalized Gradient Approximation): BP, PBE, RPBE, REVPBE, PBESOL, BLYP, OLYP, PW91, WC, EV93
  • VDW (van der Waals): VDW-DF, VDW-DF-CX, VDW-DF-C09, VDW-DF-OB86, VDW-DF-OBK8, VDW-DF2, VDW-DF2-C09, VDW-DF2-B86R, RVV10, BEEF [39]
  • Hybrid: B3LYP, PBE0, GauPBE
  • Range-separated: HSE (HSE06 functional)
Dispersion correction option menu

Choose a dispersion (van der Waals) correction from this option menu:

  • None—Do not add a dispersion correction.
  • DFT-D—Use Grimme's semiempirical D2 correction [4, 5].
  • DFT-D3—Use Grimme's semiempirical D3 correction [6]. This is the only dispersion method available for phonon calculations. You can optionally select the Use three-body term option along with this correction.
  • TS—Use the Tkatchenko-Scheffler dispersion corrections with first-principles derived C6 coefficients [7].
  • XDM—Use the exchange-hole dipole-moment model [8, 9].

This option menu is not available for VDW functionals.

Reciprocal space cutoff text box

Reciprocal space cutoff for correcting Coulomb potential divergencies at small q vectors (ecutvcut). This feature is only available for hybrid functionals.

Q mesh text boxes

Number of points in each dimension of the three-dimensional mesh for q sampling of the Fock operator (nqx1, nqx2, nqx3). Can be smaller than the number of k-points. This feature is only available for hybrid functionals.

Use symmetry option

Use symmetry in the calculation. The space group is assigned at the end of the calculation. If symmetry is not used:

  • if a list of k points is provided in input, it is used "as is": symmetry-equivalent k-points are not generated, and the charge density is not symmetrized
  • if a uniform (Monkhorst-Pack) k-point grid is provided in input, it is expanded to cover the entire Brillouin zone, irrespective of the crystal symmetry

Turning off this option can be advantageous:

  • in low-symmetry large cells, if you cannot afford a k-point grid with the correct symmetry
  • in calculations for isolated atoms
Use primitive cell option

If symmetry is in use and the input structure can be reduced to a primitive cell, convert the input structure and use the primitive cell. Sets use_primitive=true.

Brillouin zone partition options

Select an option for the partitioning of the Brillouin zone, and specify any related parameters.

Grid plane distance option

Specify the Monkhorst-Pack grid of k-points in terms of the distance between grid planes. This option is useful when the shape of the unit cell and the first Brillouin zone is not isotropic. You can control the grid size and location with the following options:

Distance text box

Specify the distance between grid planes, in inverse angstroms.

When Include Γ-point is not selected, the updated Monkhorst-Pack mesh is an even, shifted grid unless (1) the structure is hexagonal or (2) the space group is face centered and the Use symmetry option is selected.

Include Γ-point option

Include the Γ point in the grid. With this option, the offset of the grid is set to zero.

Update K-point mesh button

Click this button to update the values in the Monkhorst-Pack mesh and Number of k-points text boxes to the values generated from the grid plane distance settings for the structure in the Workspace.

Number of k-points text box

After clicking the Update K-point mesh button, this text box displays the number of irreducible k-points that are used in the calculation. Noneditable.

Monkhorst-Pack option and text boxes

Generate Monkhorst-Pack grids with the specified number of points along each reciprocal lattice vector.

Shift mesh option

Shift the mesh (grid) by half the grid step in each direction. If this option is not set and the mesh is 1x1x1, the Γ point is included in the grid.

Γ-point only (real wavefunctions) option

Use only the Γ point (the mesh is 1×1×1). This makes the calculations much faster, due to the minimal mesh and the real wave function.

SCF tab

Set options for the self-consistent field calculation of the electronic structure.

Self-consistent accuracy section

Set options for the accuracy of the SCF.

Custom energy cutoff for wavefunctions option and text box

Specify the cutoff on the kinetic energy of the plane waves included for wave function calculations, in rydbergs. By default, the cutoff is determined automatically by Quantum ESPRESSO, based on the calculation.

Custom energy cutoff for charge density option and text box

Specify the cutoff on the kinetic energy of the plane waves included for charge density calculations, in rydbergs. By default, the cutoff is determined automatically by Quantum ESPRESSO, based on the calculation.

Update from Pseudopotentials button

Update the custom energy cutoffs for wavefunctions and charge density from the values stored with the current pseudopotentials (the defaults). The values in the two boxes above are replaced.

Max steps in SCF text box

Set the maximum number of SCF iterations. If convergence is not reached within the specified number of steps, use the QM Monitor Panel to check whether the system is steadily converging. If so, continue the calculations specifying a higher number of maximum steps.

Energy convergence threshold text box

Set the threshold for determining when the electronic energy is converged, in rydbergs.

Diagonalization option menu

Choose the diagonalization method. (Sets the diagonalization keyword.)

  • Davidson—use Davidson iterative diagonalization with overlap matrix. This is fast, but occasionally fails.
  • Conjugate Gradient—use conjugate-gradient-like band-by-band diagonalization. This is slower but uses less memory.
  • Iterative PPCG—use Projected Preconditioned Conjugate Gradient (PPCG) iterative diagonalization [54].
  • Iterative ParO—use Parallel Orbital-updating (ParO) iterative diagonalization [55].
  • RMM Davidson—use RMM-DIIS (Residual Minimization Method - Direct Inversion in the Iterative Subspace) with alternating calls to Davidson iterative diagonalization.
  • RMM ParO—use RMM-DIIS (Residual Minimization Method - Direct Inversion in the Iterative Subspace) with alternating calls to Parallel Orbital-updating (ParO) iterative diagonalization.
Mixing mode option menu

Choose a method for charge density mixing. (Sets the mixing_mode keyword.)

  • Broyden—plain Broyden charge density mixing
  • Simple Thomas-Fermi—Broyden mixing with simple Thomas-Fermi screening. Useful for highly homogeneous systems.
  • Local Thomas-Fermi—Broyden mixing with local-density-dependent Thomas-Fermi screening. Useful for highly inhomogeneous systems.
Mixing factor text box

Weight of the new density from an iteration used to generate the input density for the next iteration in the Broyden mixing method. A mixing factor of 0.3 means 30% of the new density and 70% of the old density, and so represents substantial damping of the change in density. A larger factor can improve the speed of convergence for well-behaved systems. A small value may be necessary when the convergence is oscillatory or the changes in density are large, as they often are at the beginning of the SCF. (Sets the mixing_beta keyword.)

Mixing iterations option menu

Set the number of iterations used in the charge density mixing scheme. The amount of memory used increases with the number of iterations. (Sets the mixing_ndim keyword.)

Empty bands fraction text box

Number of empty bands, as a fraction of the number of occupied bands. This defines the unoccupied space, and thereby defines the size of the basis set used for the SCF.

Occupations section

Set parameters for the occupation of the electronic states. (Sets the occupations keyword.)

Smearing option, menu, and text box

Choose a method for smearing the occupations at the Fermi level. This improves convergence of the SCF, and is intended for metals where there is no band gap. (Sets the smearing keyword.)

  • Gaussian—ordinary Gaussian spreading.
  • Methfessel-Paxton—Methfessel-Paxton first-order spreading [12].
  • Marzari-Vanderbilt—Marzari-Vanderbilt cold smearing [13].
  • Fermi-Dirac—smearing with the Fermi-Dirac function.

The text box gives the size of the spreading or smearing for Brillouin-zone integration.

Fixed option

This option is intended for insulators with a gap, where no smearing is needed.

Tetrahedra option

Use the optimized tetrahedra method for the Brillouin zone integration [10, 11]. This option is especially suited for calculation of the density of states and phonons. It requires a uniform grid of k-points, automatically generated. It is not suitable for geometry optimizations because it is not variational. (Keyword value tetrahedra_opt.)

Optimization tab

Set options for optimizing the geometry.

General optimization options section

Set general optimization options for structural relaxation. These options apply when optimizing the atomic positions with or without lattice optimization.

Optimization algorithm option menu

Choose one of the following optimization algorithms (ion_dynamics):

  • BFGS quasi-newton algorithm—use the BFGS quasi-newton algorithm, based on the trust radius procedure, for structural relaxation (bfgs).
  • Damped relaxation—use damped (quick-minimization Verlet) dynamics for structural relaxation (damp).
Number of steps text box

Set the maximum number of geometry optimization steps. If convergence is not reached within the specified number of steps, use the QM Monitor Panel to check whether the system is steadily converging. If so, continue the calculations specifying a higher number of maximum steps.

Total energy threshold text box

Set the energy threshold used to determine if the geometry is converged (etot_conv_thr).

Force threshold text box

Set the force threshold used to determine if the geometry is converged (forc_conv_thr).

Lattice optimization options section

Set options for the optimization of the lattice (calculation=vc-relax), with optional constraints on the cell parameters.

Optimization algorithm option menu

Choose one of the following algorithms for the lattice optimization. (cell_dynamics).

  • BFGS quasi-newton algorithm—the default. If this algorithm is used, the same algorithm must be selected in the General optimization options section (bfgs).
  • Damped dynamics of P-R lagrangian—use damped (Beeman) dynamics of the Parrinello-Rahman extended lagrangian (damp-pr).
  • Damped dynamics of Wentzcovitch lagrangian—use damped (Beeman) dynamics of the new Wentzcovitch extended lagrangian (damp-w).
Cell optimization option menu

Specify which of the cell parameters are to be optimized. It is assumed that two of the axes are in the xy plane.

  • All axes and angles are free—optimize all cell parameters without restriction (cell_dofree='all').

  • All axes and angles are free, volume fixed—allow variation of all cell parameters but constrain the volume to be constant (cell_dofree='shape').

  • Only x and y components are free—allow variation of the cell parameters in the xy plane, and fix the rest. No constraint is placed on the volume (cell_dofree='2Dxy').

  • Only x and y components are free, x-y plane fixed—allow variation of the cell parameters in the xy plane, while keeping the area of the face in the xy plane fixed. The remaining parameters are fixed (cell_dofree='2Dshape').

  • The volume changes, keeping all angles fixed—allow variation of the cell volume but constrain angles to be constant (cell_dofree='volume').

Target pressure text box

Set the target pressure for optimization of the lattice (press).

Pressure threshold text box

Set the pressure threshold used to determine if the lattice is converged (press_conv_thr).

Optimize FFT grid option

Choose to optimize the Fast Fourier Transform (FFT) grid at every relaxation step when Optimize atomic positions and cell is selected. This option is most useful when calculating elastic constants as it allows for a more accurate calculation of stress.

(P)DOS tab

The controls in the sections in this tab are the same as in the Theory and SCF tabs. You might want to choose different options for calculating the density of states than what you use for the SCF. Note that the grid spacing for the energy is fixed at 0.01 eV.

Brillouin zone partition options

Select an option for the partitioning of the Brillouin zone, and specify any related parameters.

Grid plane distance option

Specify the Monkhorst-Pack grid of k-points in terms of the distance between grid planes. This option is useful when the shape of unit cell and the first Brillouin zone is not isotropic. You can control the grid size and location with the following options:

Distance text box

Specify the distance between grid planes, in inverse angstroms.

Include Γ-point option

Include the Γ point in the grid. With this option, the offset of the grid is set to zero.

Update K-point mesh button

Click this button to update the values in the Monkhorst-Pack mesh text boxes to the values generated from the grid plane distance settings for the structure in the Workspace.

Number of k-points text box

After clicking the Update K-point mesh button, this text box displays the number of irreducible k-points that are used in the calculation. Noneditable.

Monkhorst-Pack option and text boxes

Generate Monkhorst-Pack grids with the specified number of points along each reciprocal lattice vector.

Shift mesh option

Shift the mesh (grid) by half the grid step in each direction. If this option is not set and the mesh is 1x1x1, the Γ point is included in the grid.

Occupations section

Set parameters for the occupation of the electronic states.

Smearing option, menu, and text box

Choose a method for smearing the occupations at the Fermi level. See above for the list of methods.

Fixed option

This option is intended for insulators with a gap, where no smearing is needed.

Tetrahedra option

Use a Brillouin zone integration scheme that involves partitioning space in to tetrahedra. This option is especially suited for calculation of the density of states [12]. It requires a uniform grid of k-points, automatically generated. It is not suitable for optimizations because it is not variational.

Band structure tab

In this tab you can make settings for the calculation of the band structure.

Line density between edge points text box

Specify the number of k-points to use along the line segment between each pair of edge points in the path.

Custom k-point path option

Select this option to specify a custom path between k-points for the band structure. The path is specified by a list of points which are traversed in the order given in the table below.

K-points table

This table lists the k-points in the path, defined by their fractional coordinates and a label. All cells are editable. Greek letters can be specified by name (not symbol), and are case-insensitive.

Add button

Add a row to the table. The new row is placed at the end of the table.

Delete button

Select the rows you want to delete, then click this button to delete them.

Convergence tests tab

Set options for wave function (basis set) and k-point convergence tests. If both tests are selected, the calculations are run at all pairs of points, and can be run simultaneously. The results can be examined in the Convergence Tests Viewer Panel.

Run self-consistent tests option and section

Run tests for the convergence of the wave function with respect to the energy cutoff, by incrementing the energy cutoff over a specified range and redoing the SCF at each value. See Custom energy cutoff for wavefunctions option and text box (above) for a definition of the cutoff. If this test is selected, the values specified here are used instead of the cutoff in the SCF tab.

Minimum energy cutoff text box

Specify the minimum value of the energy cutoff for the tests (the initial value).

Maximum energy cutoff text box

Specify the maximum value of the energy cutoff for the tests. The largest cutoff used is less than or equal to this value. For example if the minimum is 20.0, the step is 10.0 and the maximum is 45.0, the largest cutoff used is 40.0.

Step size text box

Specify the size of the increment to the energy cutoff. The cutoff is incremented by this value while the cutoff is less than or equal to the maximum.

Charge density multiplier text box

Specify the multiplier to be applied to the energy cutoff to determine the charge density cutoff. The default value is 5.0, which is good for ultrasoft pseudopotentials. You can decrease the value to 4.0 for norm-conserving pseudopotentials, as this is more appropriate for these pseudopotentials.

Run Brillouin zone partition tests option and section

Run tests for the convergence of the energy with the partitioning of the Brillouin zone. If this test is selected, the values specified here are used instead of the partitioning in the Theory tab.

Grid plane distance option

Specify the Monkhorst-Pack grid of k-points in terms of the distance between grid planes. This permits a structure-agnostic test. The following settings can be used to define the sequence of grids that are used for the partitioning tests.

Max text box

Specify the maximum grid plane distance to use in the tests. This value is used in the first step, and is decremented for each subsequent step.

Decrement text box

Specify the amount by which the grid plane distance is decremented at each step (other than the first) in the tests.

Steps text box

Specify the total number of steps to use in the partitioning tests.

Update K-point mesh button

Click this button to update the values in the Monkhorst-Pack mesh text boxes to the values generated using the maximum grid plane distance. This distance defines the initial partitioning; the numbers in the text boxes (which are noneditable for this method choice) might not change, but the values are recorded for the tests.

Monkhorst-Pack option and text boxes

Run tests of the convergence with the Monkhorst-Pack partitioning. The partitioning depends on the structure. The initial partitioning is specified in the text boxes.

Increment text boxes

Specify the increment in the number of points in each direction. The increment for a given direction can be set to zero, to bypass testing the convergence in that direction (e.g. for a slab).

Steps text box

Specify the number of steps in the partitioning (including the first), i.e. the total number of partitionings.

Include gamma point text box

Include the G point in the grid (sets the offset to zero).

Custom NEB tab

Make settings for nudged elastic band calculations. This tab is only present when this dialog box is opened from the Nudged Elastic Band Calculations Panel.

Method option menu

Choose a method for the calculations to be performed, from Nudged elastic band[51], Classical NEB[50], Elastic band[52], Spline method[53], and String method[53]. The default is Nudged elastic band.

Optimization algorithm option menu

Choose an optimization algorithm for locating the point on the minimum energy path (gradients zero perpendicular to the path). The choices are:

  • ODE—An adaptive scheme for solving ordinary differential equations (ODE)[53].
  • Static—A linear optimization algorithm[53].
Force norm threshold text box

Enter the convergence threshold, in eV/Å. The convergence threshold is defined as the norm of the force orthogonal to the path. The simulation stops when the norm is smaller than the threshold value entered here.

Number of steps text box

Number of optimization steps. At each step, the geometry of each image is optimized so that the force perpendicular to the path is zero (to within a threshold).

Climbing image option

Choose whether to use the climbing image algorithm for finding the actual transition state. If this option is selected, the highest-energy image is allowed to climb along the path to the transition state, without the effect of the springs. If this option is not selected, the normal NEB algorithm does not attempt to locate the transition state, it merely produces points that are fairly evenly distributed along the path.

Dynamics tab

Set options for a Born-Oppenheimer molecular dynamics calculation. This tab is only present if you select Molecular dynamics in the Geometry section of the Quantum ESPRESSO Calculations Panel.

Dynamics algorithm option menu

Choose the algorithm for performing the dynamics, from Verlet algorithm or Over-damped Langevin. (ion_dynamics)

Ionic temperature option menu

Choose the method for controlling the temperature of the ions (atoms) in the simulation (ion_temperature). The available methods are:

  • Velocity rescaling—control temperature by velocity rescaling (rescale-v)
  • Temperature rescaling—control temperature by velocity rescaling (rescale-T)
  • Berendsen thermostat—control temperature using “soft” velocity rescaling (berendsen)
  • Andersen thermostat—control temperature using an Andersen thermostat (andersen)
  • Stochastic-velocity rescaling—control temperature by stochastic-velocity rescaling (svr)
  • Initialized but uncontrolled—initialize velocities to the specified temperature but leave them uncontrolled. (initial)
  • Uncontrolled—temperature is not controlled. (not_controlled)

The ensemble used is NVT for the rescaling methods and the thermostats, and NVE for the uncontrolled temperatures.

Simulation time text box

Specify the total simulation time in picoseconds.

Time step text box

Specify the time step to be used, in femtoseconds.

Temperature text box

Specify the temperature of the simulation, in kelvin. This is the starting temperature, and the target temperature for most thermsostats. (tempw)

Delta T text box

Multiply the instantaneous temperatures at each step by this factor. The instantaneous temperature is calculated at the end of every ionic move and before rescaling. (delta_t)

Raise steps text box

This number specifies the number of steps (nraise) used in various thermostats for the temperature control, as follows:

  • Velocity rescaling—rescale the average temperature computed over the specified number of steps to the input temperature.
  • Berendsen thermostat—the “rise time” parameter tau is given by this number of steps times the time step size.
  • Andersen thermostat—the collision frequency parameter nu is the inverse of tau, defined above.

Phonons tab

Specify the parameters of the Monkhorst-Pack mesh and define the points to be used for the phonon band dispersion calculation. The phonon bands and their intensities (both IR and Raman) can be viewed in the Vibrations Panel, which can also be used to display an animation of the vibrations in the Workspace.

Monkhorst-Pack text boxes

Set up the Monkhorst-Pack partitioning for the phonon calculation in the text boxes.

Self-consistency threshold text box

Specify the threshold for self consistency.

Compute Raman activities option

Compute Raman activities for the phonon bands. The Raman spectrum can then be displayed in the Spectrum Plot Panel.

Line density between edge points text box

Specify the number of k-points to use along the line segment between each pair of edge points in the path.

Custom k-point path option

Select this option to specify a custom path between k-points for the band structure. The path is specified by a list of points which are traversed in the order given in the table below.

K-points table

This table lists the k-points in the path, defined by their fractional coordinates and a label. All cells are editable. Greek letters can be specified by name (not symbol), and are case-insensitive.

Add button

Add a row to the table. The new row is placed at the end of the table.

Delete button

Select the rows you want to delete, then click this button to delete them.

Dielectric function tab

Set options for the calculation of the dielectric function. Note that the calculation requires fixed occupations.

Inter-band broadening (insulators) text box

Specify the inter-band broadening for insulators.

Equation of state tab

Set up strain values for equation of state calculations.

Min linear strain text box

Set the minimum strain value applied in each direction (a, b, c) to produce a volume strain.

Max linear strain text box

Set the maximum strain value applied in each direction (a, b, c) to produce a volume strain.

Number of steps text box

Set the number of strain steps from the minimum to the maximum strain value (inclusive). The generated strain values are interpolated linearly.

NMR tab

Set options for the calculation of NMR chemical shifts. This tab is only present if you select NMR in the property section of the Quantum ESPRESSO Calculations Panel.

Displacement wave vector text box
Set the displacement wave vector for the linear response calculation of the NMR chemical shifts, in units of bohr-1.

Reset button

Reset the options to their default values.

Related Topics