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, Pakistanhttp://www.qau.edu.pk/profile.php?id=818020 On Tuesday, 4 August 2015, 14:21, 毛飞 <[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]> 写道: -----原始邮件----- 发件人: Mayuri Vaghela <[email protected]> 发送时间: 2015年8月4日 星期二 收件人: [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]> 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 3gnuplot> 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]> 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 MaoUniversity of South China, China ############################ Input file for the LiF ############################SystemName bulk.LiFSystemLabel bulk_LiF NumberOfAtoms 8NumberOfSpecies 2 %block ChemicalSpeciesLabel 1 3 Li 2 9 F%endblock ChemicalSpeciesLabel # basis PAO.BasisSize DZPPAO.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 BandLines1 0.0 0.0 0.0 # Gamma-point20 0.5 0.0 0.5 # X-point30 0.5 0.25 0.75 # W-point30 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 optionsMaxSCFIterations 100 # Maximum number of SCF iterDM.NumberPulay 3 # One Pulay every 3 iterDM.MixingWeight 0.35 # New DM amount for next SCF cycleDM.Tolerance 5.d-4 # Tolerance in maximum difference SolutionMethod diagon # OrderN or DiagonOccupationFunction FDElectronicTemperature 25 meV # Temp. for Fermi smearing # exchange-correlation functionalXC.functional LDAXC.authors CA # structural infomationAtomCoorFormatOut Ang # output informationWriteCoorXmol TWriteMDXmol TWriteForces T WriteKpoints .true.WriteEigenvalues .true.WriteKbands .true.WriteBands .true.WriteMullikenPop 1 # options for MDMD.MaxForceTol 0.01 eV/AngMD.TypeOfRun CGMD.NumCGSteps 400MD.Broyden.History.Steps 6MD.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
