LDOS ATK 11.2 alpha version

[jdgayles16]

Hi,

Can I use old .nc from 10.2 to calculate the LDOS with the new alpha version? I tried and I get some pictures I dont really understand. I thought it would be similar to results i got with atk 2008.

[jdgayles16]

I tried some other systems and all my results look exactly the same, I guess have to re-converge in the new version, which should be quicker now =)

[jdgayles16]

I keep getting images like this for the ldos, whatever the system

[Anders Blom]

The isovalues might not be the same in the new version. Can you attach a picture from the old? (And if needed, email the input files so we can investigate.)

[jdgayles16]

I tried with the li-h2 tutorial and i get similar results. I will attach my script and a picture.

Code: python

###############################################################
# TwoProbe configuration
###############################################################

###############################################################
# Left electrode
###############################################################

# Set up lattice
vector_a = [10.0, 0.0, 0.0]*Angstrom
vector_b = [0.0, 10.0, 0.0]*Angstrom
vector_c = [0.0, 0.0, 6.6]*Angstrom
left_electrode_lattice = UnitCell(vector_a, vector_b, vector_c)

# Define elements
left_electrode_elements = [Lithium, Lithium, Lithium]

# Define coordinates
left_electrode_coordinates = [[ 5. ,  5. ,  1.1],
                              [ 5. ,  5. ,  3.3],
                              [ 5. ,  5. ,  5.5]]*Angstrom

# Set up configuration
left_electrode = BulkConfiguration(
    bravais_lattice=left_electrode_lattice,
    elements=left_electrode_elements,
    cartesian_coordinates=left_electrode_coordinates
    )

###############################################################
# Right electrode
###############################################################

# Set up lattice
vector_a = [10.0, 0.0, 0.0]*Angstrom
vector_b = [0.0, 10.0, 0.0]*Angstrom
vector_c = [0.0, 0.0, 6.6]*Angstrom
right_electrode_lattice = UnitCell(vector_a, vector_b, vector_c)

# Define elements
right_electrode_elements = [Lithium, Lithium, Lithium]

# Define coordinates
right_electrode_coordinates = [[ 5. ,  5. ,  1.1],
                               [ 5. ,  5. ,  3.3],
                               [ 5. ,  5. ,  5.5]]*Angstrom

# Set up configuration
right_electrode = BulkConfiguration(
    bravais_lattice=right_electrode_lattice,
    elements=right_electrode_elements,
    cartesian_coordinates=right_electrode_coordinates
    )

###############################################################
# Central region
###############################################################

# Set up lattice
vector_a = [10.0, 0.0, 0.0]*Angstrom
vector_b = [0.0, 10.0, 0.0]*Angstrom
vector_c = [0.0, 0.0, 24.9]*Angstrom
central_region_lattice = UnitCell(vector_a, vector_b, vector_c)

# Define elements
central_region_elements = [Lithium, Lithium, Lithium, Lithium, Lithium, Hydrogen, Hydrogen,
                           Lithium, Lithium, Lithium, Lithium, Lithium]

# Define coordinates
central_region_coordinates = [[  5. ,   5. ,   1.1],
                              [  5. ,   5. ,   3.3],
                              [  5. ,   5. ,   5.5],
                              [  5. ,   5. ,   7.7],
                              [  5. ,   5. ,   9.9],
                              [  5. ,   5. ,  12.1],
                              [  5. ,   5. ,  12.8],
                              [  5. ,   5. ,  15. ],
                              [  5. ,   5. ,  17.2],
                              [  5. ,   5. ,  19.4],
                              [  5. ,   5. ,  21.6],
                              [  5. ,   5. ,  23.8]]*Angstrom

# Set up configuration
central_region = BulkConfiguration(
    bravais_lattice=central_region_lattice,
    elements=central_region_elements,
    cartesian_coordinates=central_region_coordinates
    )

device_configuration = DeviceConfiguration(
    central_region,
    [left_electrode, right_electrode]
    )

###############################################################
# Calculator
###############################################################
#----------------------------------------
# Basis Set
#----------------------------------------
basis_set = GGABasis.DoubleZetaPolarized

#----------------------------------------
# Exchange-Correlation
#----------------------------------------
exchange_correlation = GGA.PBE

#----------------------------------------
# Numerical Accuracy Settings
#----------------------------------------
left_electrode_numerical_accuracy_parameters = NumericalAccuracyParameters(
    k_point_sampling=(1, 1, 100),
    )

right_electrode_numerical_accuracy_parameters = NumericalAccuracyParameters(
    k_point_sampling=(1, 1, 100),
    )

#----------------------------------------
# Iteration Control Settings
#----------------------------------------
left_electrode_iteration_control_parameters = IterationControlParameters(
    linear_dependence_threshold=4e-05,
    )

right_electrode_iteration_control_parameters = IterationControlParameters(
    linear_dependence_threshold=4e-05,
    )

#----------------------------------------
# Poisson Solver Settings
#----------------------------------------
left_electrode_poisson_solver = MultigridSolver(
    boundary_conditions=[[PeriodicBoundaryCondition,PeriodicBoundaryCondition],
                         [PeriodicBoundaryCondition,PeriodicBoundaryCondition],
                         [PeriodicBoundaryCondition,PeriodicBoundaryCondition]]
    )

right_electrode_poisson_solver = MultigridSolver(
    boundary_conditions=[[PeriodicBoundaryCondition,PeriodicBoundaryCondition],
                         [PeriodicBoundaryCondition,PeriodicBoundaryCondition],
                         [PeriodicBoundaryCondition,PeriodicBoundaryCondition]]
    )

#----------------------------------------
# Electrode Calculators
#----------------------------------------
left_electrode_calculator = LCAOCalculator(
    basis_set=basis_set,
    exchange_correlation=exchange_correlation,
    numerical_accuracy_parameters=left_electrode_numerical_accuracy_parameters,
    iteration_control_parameters=left_electrode_iteration_control_parameters,
    poisson_solver=left_electrode_poisson_solver,
    )

right_electrode_calculator = LCAOCalculator(
    basis_set=basis_set,
    exchange_correlation=exchange_correlation,
    numerical_accuracy_parameters=right_electrode_numerical_accuracy_parameters,
    iteration_control_parameters=right_electrode_iteration_control_parameters,
    poisson_solver=right_electrode_poisson_solver,
    )

#----------------------------------------
# Device Calculator
#----------------------------------------
calculator = DeviceLCAOCalculator(
    basis_set=basis_set,
    exchange_correlation=exchange_correlation,
    electrode_calculators=
        [left_electrode_calculator, right_electrode_calculator],
    )

device_configuration.setCalculator(calculator)
nlprint(device_configuration)
device_configuration.update()
nlsave('analysis.nc', device_configuration)

local_device_density_of_states = LocalDeviceDensityOfStates(
    configuration=device_configuration,
    energy=0*eV,
    kpoints=MonkhorstPackGrid(1,1,1),
    contributions=All,
    energy_zero_parameter=AverageFermiLevel,
    infinitesimal=1e-06*eV,
    self_energy_calculator=KrylovSelfEnergy(),
    spin=Spin.Up,
    )
nlsave('analysis.nc', local_device_density_of_states)

[Anders Blom]

Thanks, we'll definitely have a look at it. This is why we have alpha-version, to make things work for the final release. Thanks for helping!

[Anders Blom]

I have a feeling this is not related to the LDOS, but rather to the keyword

linear_dependence_threshold=4e-05,

which VNL in the alpha-version puts into the script by mistake. I removed it and ran your script, see attached image. (I did not run with 11.2.a1, I used our latest development edition, but that is what becomes the next release anyway.)

I'd rather not run with the keyword above (because they make the convergence really poor), so perhaps instead you can just run with those removed to see if all is fine for you too