Page 1 of 1

High-order zfs, wignerd

Posted: Tue Mar 21, 2017 3:22 pm
by Mantas_Sim
Hi!

I am trying to simulate a Mn2+ spectrum with higher order zfs parameters B(k,q). In the manual it is written that currently it is not possible to include tilt angles for the principal frames of these higher order tensors. But there is an option provided to use Wigner rotation matrices:

"Alternatively, you can use wignerd to compute a Wigner rotation matrix that can be used to tilt the tensors explicitly.
angles = rand(1,3)*pi; % Euler tilt angles, in radians
B2 = [3 4 5 0 2]; % B(2,q) tensor components
R = wignerd(2,angles); % rotation matrix for rank-2 tensor
B2 = R*B2; % rotated tensor
Sys.B2 = B2.';"

Here I see two problems. First of all, shouldn't B2 = R*B2 be B2 = R*B2' ? Secondly, wignerd provides complex rotation matrix which then is multiplied by B2 giving complex components of Sys.B2. But those should be real. Is this a just a mistake in the manual? How to tilt the higher-order tensors then?

Thanks!
Mantas

Re: High-order zfs, wignerd

Posted: Thu Mar 23, 2017 10:01 am
by Stefan Stoll
This is indeed not correct. There needs to be a transpose, and the rotation should not give a complex-valued result, since we are not using spherical tensor operators, but the tesseral linear combinations of them.

We will fix this. Thanks for reporting!