Hello SIESTA users, I am facing a difficulty during while trying to fully relax a 32 atom graphene like 2D system with and without a single vacancy, in the middle of the srtucture. A part of my input .fdf file is given below:
*DM.Tolerance 1.d-3DM.Require.Energy.Convergence TDM.Energy.Tolerance 1.d-5 eVPAO.BasisSize DZPPAO.EnergyShift 0.3 eVMD.TypeOfRun CGMaxSCFIterations 3000MD.NumCGsteps 3000MD.VariableCell TMD.MaxForceTol 0.02 eV/AngMeshCutoff 300 RyDM.MixingWeight 0.02DM.NumberPulay 3WriteCoorXmol TWriteMullikenPop 0XC.functional GGAXC.authors PBESolutionMethod diagonElectronicTemperature 300 K* The lattice constant of the system, the bond length and bond angle used by me has already been reported in a reputed journal, as a stable one. I have not used any constraints in my structure. I have created a 4x4x1 supercell using the reported dimensions and trying to optimize the structure with the siesta code version 3.2-pl-5. The problem I am facing is that the optimized structure is no longer of the symmetry I started with. The structure is no longer a hexagonal honeycomb one anymore. Also while using a vacancy defect in the structure the atoms are getting asymmetrically arranged breaking any symmetry that was present there. Now my question is, are there any more parameters that I should have specified in the input .fdf file? Or, will a change in the .psf file can bring about any improvement in the optimization? Also, if I apply some constraints like fixing the bond angles or that my structure should relax only in a single plane, then how can I say that all the atoms in my structure are FULLY relaxed? Please kindly help me in this regard. I feel like I am missing some important parameter in my .fdf file. Thank you. Regards, Sunetra Das.