Hi Fei,
your unit cell is wrong.
Here is the correct conventional unit cell:
Li4 F4
1.0
4.0834274292 0.0000000000 0.0000000000
0.0000000000 4.0834274292 0.0000000000
0.0000000000 0.0000000000 4.0834274292
Li F
4 4
Cartesian
0.000000000 0.000000000 0.000000000
0.000000000 2.041713715 2.041713715
2.041713715 0.000000000 2.041713715
2.041713715 2.041713715 0.000000000
0.000000000 2.041713715 0.000000000
0.000000000 0.000000000 2.041713715
2.041713715 2.041713715 2.041713715
2.041713715 0.000000000 0.000000000
Good luck
--Zhen
On Tue, Aug 4, 2015 at 9:59 PM, 毛飞 <[email protected]> wrote:
> >Dear Khan and Rahman
>
> I calculate the band gap of LiF with the corrected input file, but the
> energy gap is still very large, is about 18 eV (see attachment). The input
> file is attached below.
>
>
>
> Any comment or suggestions?
>
> Regards,
>
>
>
> Fei
>
>
>
>
>
> ################### Input file #####################
>
> SystemName bulk.LiF
>
> SystemLabel bulk_LiF
>
>
>
> NumberOfAtoms 2
>
> NumberOfSpecies 2
>
>
>
> %block ChemicalSpeciesLabel
>
> 1 3 Li
>
> 2 9 F
>
> %endblock ChemicalSpeciesLabel
>
>
>
> # basis
>
> PAO.BasisSize DZP
>
> PAO.EnergyShift 200 meV
>
>
>
> LatticeConstant 4.02 Ang # Lattice constant alat
>
> %block LatticeParameters # conventional cell
>
> 1.00 1.00 1.00 90.0 90.0 90.0
>
> %endblock LatticeParameters
>
>
>
> %block kgrid_Monkhorst_Pack
>
> 10 0 0 0.0
>
> 0 10 0 0.0
>
> 0 0 10 0.0
>
> %endblock kgrid_Monkhorst_Pack
>
>
>
>
>
> BandLinesScale ReciprocalLatticeVectors
>
> %block BandLines
>
> 1 0.0 0.0 0.0 # Gamma-point
>
> 20 0.5 0.0 0.5 # X-point
>
> 30 0.5 0.25 0.75 # W-point
>
> 30 0.0 0.0 0.0 # Gamma-point
>
> %endblock BandLines
>
>
>
>
>
> %block ProjectedDensityOfStates
>
> -20 10 0.1 300 eV
>
> %endblock ProjectedDensityOfStates
>
>
>
> MeshCutoff 300.0 Ry
>
>
>
> # SCF options
>
> MaxSCFIterations 100 # Maximum number of SCF iter
>
> DM.NumberPulay 3 # One Pulay every 3 iter
>
> DM.MixingWeight 0.35 # New DM amount for next SCF cycle
>
> DM.Tolerance 5.d-4 # Tolerance in maximum difference
>
>
>
> SolutionMethod diagon # OrderN or Diagon
>
> OccupationFunction FD
>
> ElectronicTemperature 25 meV # Temp. for Fermi smearing
>
>
>
> # exchange-correlation functional
>
> XC.functional LDA
>
> XC.authors CA
>
>
>
>
>
> # structural infomation
>
> AtomCoorFormatOut Ang
>
>
>
> # output information
>
> WriteCoorXmol T
>
> WriteMDXmol T
>
> WriteForces T
>
>
>
> WriteKpoints .true.
>
> WriteEigenvalues .true.
>
> WriteKbands .true.
>
> WriteBands .true.
>
> WriteMullikenPop 1
>
>
>
> # options for MD
>
> MD.MaxForceTol 0.01 eV/Ang
>
> MD.TypeOfRun CG
>
> MD.NumCGSteps 400
>
> MD.Broyden.History.Steps 6
>
> MD.Broyden.Initial.Inverse.Jacobian 1.0
>
>
>
> # options for continue
>
> # MD.UseSaveCG T # neccessary!
>
> # MD.UseSaveXV T # neccessary!
>
> # DM.UseSaveDM true # to use continuae..ion files
>
>
>
>
>
> AtomicCoordinatesFormat Fractional
>
> %block AtomicCoordinatesAndAtomicSpecies # Number of irreducible
> atoms in the conventional cell is 2
>
> 0.0000000 0.0000000 0.0000000 1 Li
>
> 0.5000000 0.5000000 0.5000000 2 F
>
> %endblock AtomicCoordinatesAndAtomicSpecies
>
>
>
> 在2015-08-05，Younas Khan <[email protected]> 写道:
>
> -----原始邮件-----
> *发件人:* Younas Khan <[email protected]>
> *发送时间:* 2015年8月5日 星期三
> *收件人:* "[email protected]" <[email protected]>
> *主题:* Re: [SIESTA-L] The energy band gap of LiF
>
> I think you have incorrectly written the lattice vectors. Kindly check it,
> they may not be in proper format.
>
> Younas Khan
>
> On Tuesday, August 4, 2015, 毛飞 <[email protected]> wrote:
>
>> Dear Rahman
>>
>>
>>
>> Thank you for your reply, I will try the calculation with your
>> suggestions.
>>
>>
>>
>> Fei
>>
>> 在2015-08-04，Gul Rahman <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>>
>> 写道:
>>
>> -----原始邮件-----
>> *发件人:* Gul Rahman <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>
>> >
>> *发送时间:* 2015年8月4日 星期二
>> *收件人:* "[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>"
>> <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>
>> >
>> *主题:* Re: [SIESTA-L] The energy band gap of LiF
>>
>> Dear Fei Mao,
>> Yes, your band structure is strange, it has flat band (like molecules)
>> and the band gap is very large, which is wrong.
>> Plz, look carefully into your input file.
>> First, try to reproduce the band structure and lattice parameters of
>> bulk (primitive) LiF, then you can generate the SuperCell either using your
>> pencil in hands (:) or using the SIESTA or any other computer code. And
>> then put the Atomic coordinates of the Supercell explicitly in your fdf
>> file. You can see you have defined SuperCell (2x2x2) in your fdf file.
>> Also, try to use fractional coordinates if you are dealing with a bulk
>> system. Whenever you generate .XV file, you must visualise it to see what's
>> going on.
>> Also, your k-points sampling is not good for the bulk LiF system. You
>> must use dense k-points. I will not use 1x1x1 for bulk. I have to converge
>> it.
>> I hope it may help.
>> Regards,
>> Gul
>>
>>
>> Dr. Gul Rahman
>> Assistant Professor,
>> Department of Physics,
>> Quaid-i-Azam University,
>> Islamabad, Pakistan
>> http://www.qau.edu.pk/profile.php?id=818020
>>
>>
>>
>> On Tuesday, 4 August 2015, 14:21, 毛飞 <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>>
>> wrote:
>>
>>
>> Dear Rahman and Vaghela
>>
>> Thank you for kind replies.
>> The Fermi energy level is at -5.9 eV in the calculation, I modify the fig
>> and mark the Fermi energy by a black line (which is attached). Now, I am
>> puzzled by two points in this fig.
>>
>> 1. As shown in the fig, the conduction and valence band are both
>> composed by many straight lines, it seems to me very strange. Generally,
>> these energy levels evolves smoothly with changes in k, forming a smooth
>> band of states, and there are conduction band bottom and valence band top.
>> However, these typical characters are absence in the fig.
>>
>> 2. It can be seen from the fig that, the energy gap of LiF is about
>> 16 eV (the experimental one is 14 eV), but the LDA calculations usually
>> underestimate the gap.
>>
>> Any comment is appreciated, and thank you in advance.
>> Sincerely
>>
>> Fei Mao
>> University of South China, China
>>
>>
>>
>>
>>
>> 在2015-08-04，Mayuri Vaghela <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>>
>> 写道:
>>
>> -----原始邮件-----
>> *发件人:* Mayuri Vaghela <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>
>> >
>> *发送时间:* 2015年8月4日 星期二
>> *收件人:* [email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>
>> *主题:* Re: [SIESTA-L] The energy band gap of LiF
>>
>> first check systemlabel.EIG file,
>> you will get fermi energy from that file.
>> now draw clear line using set axis command of gnuplot at appropriate
>> energy value
>> from that you can distinguish band gap between CB &VB...
>>
>>
>> On Tue, Aug 4, 2015 at 10:47 AM, Altaf Ur Rahman <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>
>> > wrote:
>>
>> Dear when you calculate the band structure by using like gnuplot you must
>> see the fermi energy from your band data file/EIG file and then when you
>> plot the you can write in the gnuplot this sentence
>> *set zero axis 3*
>> gnuplot> plot './Band.data' u ($1)+3.23:2 w l
>>
>> here in above case your fermi energy is 3.23
>> and you can see the fermi line will appear that clearly distiguish the
>> conduction and valance band
>>
>>
>>
>> On Mon, Aug 3, 2015 at 6:04 PM, 毛飞 <[email protected]
>> <http://stu.mail.bnu.edu.cn/jy3/editor/java_script:_e(%7B%7D,'cvml','[email protected]');>
>> > wrote:
>>
>> Dear siesta developers and users
>>
>> I am a newbie in band structure calculation. I calculate the band
>> structure of LiF, and a 2*2*2 supercell of 64 atoms is selected. In this
>> calculation, the local density approximation (LDA) is adopted for the
>> exchange-correlation functional term, the band structure is obtained along
>> the high symmetric line Γ->X->W->Γ, only gamma point is used for
>> Brillouin zone sampling, more details can be found in the input file which
>> is attached below.
>>
>> I can not understand the results of the E(k) (two figs are also
>> attached), it seems there are two band gaps formed in the calculation (the
>> Fermi energy is shifted to 0 eV), I want to know which one (gap 1 or gap
>> 2) is the band gap in my calculation, and how does another gap come from?
>>
>> Is there anything wrong in the input file?
>>
>> Please help me, any suggestion or comment is appreciated.
>> Sincerely
>>
>> Fei Mao
>> University of South China, China
>>
>> ############################ Input file for the LiF
>> ############################
>> SystemName bulk.LiF
>> SystemLabel bulk_LiF
>>
>> NumberOfAtoms 8
>> NumberOfSpecies 2
>>
>> %block ChemicalSpeciesLabel
>> 1 3 Li
>> 2 9 F
>> %endblock ChemicalSpeciesLabel
>>
>> # basis
>> PAO.BasisSize DZP
>> PAO.EnergyShift 200 meV
>>
>> LatticeConstant 4.04 Ang # Lattice constant alat
>> %block LatticeVectors # Lattice vectors, in units of
>> latticeconstant
>> 2.0 0.0 0.0
>> 0.0 2.0 0.0
>> 0.0 0.0 2.0
>> %endblock LatticeVectors
>>
>>
>> %block SuperCell
>> 2 0 0 0.0
>> 0 2 0 0.0
>> 0 0 2 0.0
>> %endblock SuperCell
>>
>>
>> %block kgrid_Monkhorst_Pack
>> 1 0 0 0.0
>> 0 1 0 0.0
>> 0 0 1 0.0
>> %endblock kgrid_Monkhorst_Pack
>>
>>
>> BandLinesScale ReciprocalLatticeVectors
>> %block BandLines
>> 1 0.0 0.0 0.0 # Gamma-point
>> 20 0.5 0.0 0.5 # X-point
>> 30 0.5 0.25 0.75 # W-point
>> 30 0.0 0.0 0.0 # Gamma-point
>> %endblock BandLines
>>
>>
>> %block ProjectedDensityOfStates
>> -20 10 0.1 300 eV
>> %endblock ProjectedDensityOfStates
>>
>> MeshCutoff 300.0 Ry
>>
>> # SCF options
>> MaxSCFIterations 100 # Maximum number of SCF iter
>> DM.NumberPulay 3 # One Pulay every 3 iter
>> DM.MixingWeight 0.35 # New DM amount for next SCF cycle
>> DM.Tolerance 5.d-4 # Tolerance in maximum difference
>>
>> SolutionMethod diagon # OrderN or Diagon
>> OccupationFunction FD
>> ElectronicTemperature 25 meV # Temp. for Fermi smearing
>>
>> # exchange-correlation functional
>> XC.functional LDA
>> XC.authors CA
>>
>> # structural infomation
>> AtomCoorFormatOut Ang
>>
>> # output information
>> WriteCoorXmol T
>> WriteMDXmol T
>> WriteForces T
>>
>> WriteKpoints .true.
>> WriteEigenvalues .true.
>> WriteKbands .true.
>> WriteBands .true.
>> WriteMullikenPop 1
>>
>> # options for MD
>> MD.MaxForceTol 0.01 eV/Ang
>> MD.TypeOfRun CG
>> MD.NumCGSteps 400
>> MD.Broyden.History.Steps 6
>> MD.Broyden.Initial.Inverse.Jacobian 1.0
>>
>> # options for continue
>> # MD.UseSaveCG T # neccessary!
>> # MD.UseSaveXV T # neccessary!
>> # DM.UseSaveDM true # to use continuae..ion files
>>
>>
>> AtomicCoordinatesFormat Ang
>> %block AtomicCoordinatesAndAtomicSpecies
>> -5.963747 -2.022693 -2.026392 1
>> -6.051639 -5.949852 -2.013945 2
>> -2.147364 -2.056847 -2.009531 2
>> -2.039835 -5.953723 -2.024337 1
>> -6.085698 -1.925561 2.008577 2
>> -5.941415 -5.910826 2.052174 1
>> -2.040599 -2.032355 2.028040 1
>> -2.129653 -6.096888 2.008118 2
>> %endblock AtomicCoordinatesAndAtomicSpecies
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
--
Zhen Zhu,
Postdoctoral Researcher
Materials Department
University of California
Santa Barbara, CA 93106-5050 U.S.A.
Tel: +1-517-488-2298
Email: [email protected]
[email protected]
