[Wien] NMR calculation in Spin-polarized cases

[Nakamoto, Mitsunori]

Dear WIEN2k users and developers,

I'm Mitsunori, an engineer in Sony. I'm interested in calculating NMR contact shifts by using the NMR module recently implemented in WIEN2k (v13). I have a question about NMR calculations in spin-polarized cases.

First, the calculations on a paper (Phys Rev B 85, 245117) were tested, and the results were confirmed to be reproduced. Then I moved on shift calculations in more complicated materials, for example, Li6CoO4 as shown in a paper (Phys Chem C 116 27393, it seems they don't use the NMR module in WIEN2k). I expected to obtain similar values also by the NMR module, but it failed. The obtained ppm value for Li in Li6CoO4 is almost same as that in LiCl (reference material), which means the shift is nearly 0. 

My SCF calculation is the spin-polarized one and the result shows that the electronic state is as expected (S=6 per unit cell by -up setup). DOS is same as the one in the paper above. I wonder if my calculation procedure for NMR shifts is correct or not. My procedure is as follows:

 x_nmr_lapw -mode in1 -nodes ** -up -p -scratch /scr/**up/
 x_nmr_lapw -up -p

This procedure is repeated for -dn in another directory by using the same SCF result. I mean,

 x_nmr_lapw -mode in1 -nodes ** -dn -p -scratch /scr/**dn/
 x_nmr_lapw -dn -p

Then I summed up ppm values from these calculations. Is this a right way or not? Is there any important parameters as for this type of calculation?
Any advice would be appreciated. Thank you in advance.

Mitsunori Nakamoto
Mitsunori.Nakamoto at jp.sony.com
Sony Corporation