Hi Exciting users
I have followed the tutorial, but I don’t know if the Exciting code can calculate the excitons at different K points (Such as the hole locates at the M point, and the electron locates at the Gamma point). Can q-dependent BSE cacluate this property?
The exciting code is able to do q-dependent BSE but it is not possible to locate the electron and hole at a specific point in the reciprocal space directly.
Though you could try the following workaround:
The element qpointset can be used for q-dependent BSE. Here you specify all q-points to calculate. Technically, it displaces the occupied states about +q/2 and the unoccupied states about -q/2.
The attribute vkloff sets up an offset to the k-points.
Thus the first k-point for the occupied states is vkloff / N_k + q/2 and for the unoccupied states vkloff / N_k - q/2.
If you want to calculate excitons where the hole locates at the M point and the electron at the Gamma point you could do the following:
Set the Number of k-points (ngridk and ngridq in the xs element) to 1 1 1
Set a qpoint element in the qpointset element to q = Gamma - M
Set vkloff to ** - (Gamma + M) / 2**
The occupied stats (where the hole sits) are now calculated at the k-point vkloff / N_k + q/2 = -(Gamma + M) / 2 + (Gamma - M) / 2 = M*,
And the unoccupied states (where the electron sits) are calculated at the k-point vkloff / N_k - q/2 = -(Gamma + M) / 2 - (Gamma - M) / 2 = -Gamma = Gamma*.
Is this what you need? Also, I’ve never tried this, so I can not guarantee good results!!!
Dear maurerben,
I did the tests according to the methods you suggested. When set the Number of k-points (ngridk and ngridq in the xs element) to 1 1 1, no documents related to EPSILON and LOSS were produced. The EXCCOEFF_BSE-singlet-TDA-BAR_SCR-full_QMT001.OUT also was not produced. The input.xml has been pasted.