Mingsu Si

Dear Kay, I am sorry for response late. The component 211 should be equal to 112, and the component 233 should be equal to 322. This requirement is not directly related to the symmetry of the system. Best regards, Mingsu

Dear Michael, I also have another question. In the calculation of shg, the elk code uses the formula chi2w(:)=t0 \x chi2w(:) , where t0=wkptnr/omega. Here omega is the volume of bulk. If the symmetry is two-dimensional, the z direction would have a large vaccum layer. In this case, how do we choose the vulume of the two-dimensional system?

Dear Michael, In fact, these two components xyy and yyx have finite amplitudes and are not zero. Based on elk, this system has four symmetry operations: diag{1,1,1}, diag{-1,1,1}, diag{-1,1,-1}, and diag{1,1,-1}, which are the same as those generated by vasp. In this way, the symmetry constraint can not cancel these two components as the x component appears two times. But, we cannot understand why this happens.

Dear Michael, Thank you for your quick response. The following is my elk.in. The structure is optimized by vasp. SHG is calculated by elk. ! Calculation of the non-linear reponse function χ(-2ω;ω,ω) for GaAs. Note ! that many k-points and empty states are required for good convergence. ! Various contributions to the susceptibility are also written to file. ! See Phys. Rev. B 48, 11705 (1993) and Phys. Rev. B 53, 10751 (1996). tasks 0 120 125 ! large number of empty states required nempty 5 ! scissor...

Dear everyone, I have a question about SHG for those low-dimensional systems. In bulk, such as GaAs, we can get that the tensors such as Chi_xyz and Chi_yzx are equal. However, for two-dimensional systems, the tensors such as Chi_xyy and Chi_yyx are not equal, which should be the same in the SHG formulas.