Octopus输入文件实例

周铁戈

Octopus是一款开源的第一性原理计算软件，采用密度泛函理论（DFT）计算体系基态。可以利用的交换关联泛函有LDA、GGA、Meta-GGA 和OEP等。能够采用含时密度泛函理论（TDDFT）计算体系的响应，给出体系实空间的电荷密度的实时演化。周期体系的TDDFT计算是Octopus软件相对于其他第一性原理计算软件的突出优点。

笔者曾利用Octopus软件对铁基超导体、铜氧化物超导体和Nb单质超导进行了一系列研究，主要计算体系电荷密度随时间的演化，借此探索超导电子的配对机制。本文给出作者研究的三个体系的Octopus输入文件，仅供参考。

一、铁基超导体KFeSe的TDDFT计算。首先要计算出体系的基态，即设定CalculationMode = gs，然后再修改为CalculationMode = td，进行TDDFT计算。计算中笔者使用外加时变电场激发体系电荷密度的变化，激发之后撤销外加电场，最后给出体系电荷密度的自由振动。下面是输入文件：

CalculationMode = td

PeriodicDimensions = 3

BoxShape = parallelepiped

ExperimentalFeatures = yes

EigenSolver = rmmdiis

ConvRelDens = 1.0e-7

ExcessCharge = 2.0  # 减少2个价电子

MaximumIter = 600

Output = density

OutputFormat = cube

UnitsOutput = eV_Angstrom

a = 1.0*angstrom

%Species

"Fe" | species_pseudo | hubbard_l | 2 | hubbard_u | 4.5/27.21138

%

PseudopotentialSet = hscv_pbe

DFTULevel = dft_u_empirical

%LatticeParameters

  a | a | a

%

%LatticeVectors

     3.9479032341976 |   0.0000000000000 |   0.0000000000000

     0.0000000000000 |   3.9479032335976 |   0.0000000000000

     0.0000000000000 |   0.0000000000000 |  12.2633618360410

%

%ReducedCoordinates

"Fe" |  0.000000000000 | 0.500000000000 | 0.250000000000

"Fe" |  0.000000000000 | 0.500000000000 | 0.750000000000

"Fe" |  0.500000000000 | 0.000000000000 | 0.250000000000

"Fe" |  0.500000000000 | 0.000000000000 | 0.750000000000

"Se" |  0.000000000000 | 0.000000000000 | 0.368205065662

"Se" |  0.000000000000 | 0.000000000000 | 0.631794934337

"Se" |  0.500000000000 | 0.500000000000 | 0.868206065663

"Se" |  0.500000000000 | 0.500000000000 | 0.131793934337

"K"  |  0.000000000000 | 0.000000000000 | 0.000000000000

"K"  |  0.500000000000 | 0.500000000000 | 0.500000000000

%

Spacing = 0.35

FromScratch = no

ParDomains = no

ParKPoints = auto

SpinComponents = polarized

GuessMagnetDensity = user_defined

%AtomsMagnetDirection

 5.0

 5.0

-5.0

-5.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

%

LCAOStart= lcao_states

%KPointsGrid

6 | 6 | 4

%

KPointsUseSymmetries = no

KPointsUseTimeReversal = no

ExtraStates = 15

FilterPotentials = filter_none

AOThreshold = 0.05

OutputLDA_U = occ_matrices + effectiveU

Tf  = 20/eV

dt = 0.002/eV

TDPropagator = aetrs

TDMaxSteps = Tf/dt

TDTimeStep = dt

amplitude = 1*eV/angstrom

omega = 1.8*eV

tau0 = 5.0/eV

t0 = tau0

%TDExternalFields

  electric_field | 1 | 1 | 1 | omega | "envelope_cos"

%

%TDFunctions

  "envelope_cos" | tdf_cosinoidal | amplitude | tau0 | t0

%

TDOutput = laser + multipoles

二、铜氧化物超导体TlBa2CaCu2O7的TDDFT研究，采用轻微改变一个Tl原子的位置产生电荷扰动。具体计算中先把一个Tl原子偏离平衡位置，计算基态，然后把Tl原子位置复原，再进行TDDFT计算。下面是输入文件：

CalculationMode = td

PeriodicDimensions = 3

BoxShape = parallelepiped

ExperimentalFeatures = yes

EigenSolver = rmmdiis

ConvRelDens = 1.0e-6

MaximumIter = 650

Output = density

OutputFormat = cube

UnitsOutput = eV_Angstrom

a = 1.0*angstrom

%Species

"Cu" | species_pseudo | hubbard_l | 2

%

PseudopotentialSet = hscv_pbe

DFTULevel = dft_u_acbn0

#DFTULevel = dft_u_empirical

%LatticeParameters

  a | a | a

%

%LatticeVectors

5.3999928121  |  0.0000000000  |   0.0000000000

0.0000000000  |  5.3999921647  |   0.0000000000

0.0000000000  |  0.0000000000  |  12.8208424994

%

%ReducedCoordinates

"Tl" |  0.500000000000 | 0.500000000000 | 0.500000000000

"Tl" |  0.000000000000 | 0.000000000000 | 0.500000000000

"Ba" |  0.500000000000 | 0.000000000000 | 0.289365868572

"Ba" |  0.000000000000 | 0.500000000000 | 0.289365858572

"Ba" |  0.500000000000 | 0.000000000000 | 0.710634131496

"Ba" |  0.000000000000 | 0.500000000000 | 0.710634131496

"Ca" |  0.500000000000 | 0.000000000000 | 0.000000000000

"Ca" |  0.000000000000 | 0.500000000000 | 0.000000000000

"Cu" |  0.500000000000 | 0.500000000000 | 0.132782612859

"Cu" |  0.000000000000 | 0.000000000000 | 0.132782612859

"Cu" |  0.500000000000 | 0.500000000000 | 0.867217418172

"Cu" |  0.000000000000 | 0.000000000000 | 0.867217418172

"O"  |  0.750000000000 | 0.250000000000 | 0.120153183183

"O"  |  0.250000000000 | 0.749999950000 | 0.120153123183

"O"  |  0.750000000000 | 0.250000000000 | 0.879846846795

"O"  |  0.250000000000 | 0.749999957000 | 0.879846846795

"O"  |  0.750000000000 | 0.749999957000 | 0.120153183183

"O"  |  0.250000000000 | 0.250000000000 | 0.120153183183

"O"  |  0.750000000000 | 0.749999957000 | 0.879846816795

"O"  |  0.250000000000 | 0.250000000000 | 0.879846843795

"O"  |  0.500000000000 | 0.500000000000 | 0.328394882794

"O"  |  0.000000000000 | 0.000000000000 | 0.328393482794

"O"  |  0.500000000000 | 0.500000000000 | 0.671606517238

"O"  |  0.000000000000 | 0.000000000000 | 0.671605517238

"O"  |  0.500000000000 | 0.000000000000 | 0.500000000000

"O"  |  0.000000000000 | 0.500000000000 | 0.500000000000

%

Spacing = 0.39

FromScratch = yes

ParDomains = no

ParKPoints = auto

SpinComponents = polarized

GuessMagnetDensity = user_defined

%AtomsMagnetDirection

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 3.0

-3.0

 3.0

-3.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

 0.0

%

LCAOStart= lcao_states

%KPointsGrid

 2 | 2 | 1

%

KPointsUseSymmetries = no

KPointsUseTimeReversal = no

ExtraStates = 25

FilterPotentials = filter_none

AOThreshold = 0.04

OutputLDA_U = occ_matrices + effectiveU

TDFreezeU = yes

Tf  = 40/eV

dt = 0.002/eV

TDPropagator = aetrs

TDMaxSteps = Tf/dt

TDTimeStep = dt

三、Nb单值超导的TDDFT研究，Nb不符合同位素效应。采用随时间变化电场产生扰动，外加电场在一定时间后消失，研究电荷自由振荡的规律。下面是输入文件：

CalculationMode = td

PeriodicDimensions = 3

BoxShape = parallelepiped

ExperimentalFeatures = yes

EigenSolver = rmmdiis

ConvRelDens = 1.0e-6

MaximumIter = 450

Output = density

OutputFormat = cube

UnitsOutput = eV_Angstrom

SmearingFunction = fermi_dirac

a = 1.0*angstrom

%Species

"Nb" | species_pseudo

%

PseudopotentialSet = hscv_pbe

#DFTULevel = dft_u_acbn0

#DFTULevel = dft_u_empirical

%LatticeParameters

  a | a | a

%

%LatticeVectors

6.6012001000047  |  0.0000000000000  |   0.0000000000000

0.0000000000000  |  6.6012001000047  |   0.0000000000000

0.0000000000000  |  0.0000000000000  |   6.6012008000047

%

%ReducedCoordinates

"Nb"  |        0.000000     |    0.000000    |     0.000000

"Nb"  |        0.250000     |    0.250000    |     0.250000

"Nb"  |        0.500000     |    0.000000    |     0.000000

"Nb"  |        0.750000     |    0.250000    |     0.250000

"Nb"  |       -0.000000     |    0.500000    |     0.000000

"Nb"  |        0.250000     |    0.750000    |     0.250000

"Nb"  |        0.500000     |    0.500000    |     0.000000

"Nb"  |        0.750000     |    0.750000    |     0.250000

"Nb"  |       -0.000000     |   -0.000000    |     0.500000

"Nb"  |        0.250000     |    0.250000    |     0.750000

"Nb"  |        0.500000     |   -0.000000    |     0.500000

"Nb"  |        0.750000     |    0.250000    |     0.750000

"Nb"  |       -0.000000     |    0.500000    |     0.500000

"Nb"  |        0.250000     |    0.750000    |     0.750000

"Nb"  |        0.500000     |    0.500000    |     0.500000

"Nb"  |        0.750000     |    0.750000    |     0.750000

%

Spacing = 0.35

FromScratch = yes

ParDomains = no

ParKPoints = auto

SpinComponents = polarized

GuessMagnetDensity = paramagnetic

LCAOStart= lcao_states

%KPointsGrid

 4 | 4 | 4

%

KPointsUseSymmetries = no

KPointsUseTimeReversal = no

ExtraStates = 25

FilterPotentials = filter_none

AOThreshold = 0.04

OutputLDA_U = occ_matrices + effectiveU

OutputInterval = 50

TDFreezeU = yes

Tf  = 40/eV

dt = 0.002/eV

TDPropagator = aetrs

TDMaxSteps = Tf/dt

TDTimeStep = dt

amplitude = 1*eV/angstrom

omega = 1.8*eV

tau0 = 10.0/eV

t0 = tau0

%TDExternalFields

  electric_field | 1 | 1 | 1 | omega | "envelope_cos"

%

%TDFunctions

  "envelope_cos" | tdf_cosinoidal | amplitude | tau0 | t0

%

TDOutput = laser + multipoles