Exercises core loss spectra

WIEN2013Penn State

General• MATERIALS• * TiN, TiC : see User’s Guide and previous exercises• * fcc Cu : a=6.822 a.u. ; space group = 225 ; Cu at 0.0 0.0 0.0• * wurtzite GaN : H lattice, a=6.016, c=9.785 ; Ga at 1/3, 2/3, 0 and

2/3, 1/3, 0.5 ; N at 1/3, 2/3, u and 2/3, 1/3, u+1/2 (use u=0.371) • CALCULATION PARAMETERS• Use reasonable values. No need to use huge k-meshes or test

RKMAX convergence at this point. (E.g. rkmax = 6 ; nkpt = 1000 ; or whatever you think is right). Use regular run_lapw for everything (no sp, so, …)

Cu K edge XAS

• Set up a Cu calculation• Run SCF• Calculate the XAS for the K edge. The edge is at 8995

eV; core hole width is about 1.16 eV.• Compare to experiment below – why is it different?

h-GaN N K EELS• Set up a GaN calculation. Run the SCF.• Calculate the N K edge EELS, averaging over beam orientations. Use

collection angle = 0.3 mrad, convergence angle = 0.2 mrad, beam energy = 300 keV.

• Calculate orientation-sensitive spectra for a beam parallel to the c-axis ( 0 0 1) and for the beam in the a,b-plane (1 0 0).

• Compare to blue curves below ( right: 1 0 0 ; left: 0 0 1)

Cu L3 corehole EELS• Make a small supercell from your previous Cu.struct (or remove symmetry from the regular unit

cell)• Label 1 atom differently – this will be your core hole atom• Run initialization completely• Edit the case.inc and remove 1 electron from the 2p3/2 state• Either add 1 background charge in case.inm;

or• add 1 valence electron in case.in2

(don’t do both!)• run SCF• calculate the L3 spectrum in TELNES• How does it compare to an L3 edge from

your earlier ground-state calculation?• Repeat this exercise, but now remove

only 0.5 electron from the core state.

Mg K XAS (from earlier slides)

GaN N K EELS

• Create supercells of different size, e.g. 2x2x2 and 3x3x3 (and you can start with 1x1x1 if you like)

• Compare the resulting N K spectra