Issues with salt

[jevkus]

I would like to report some strange behavior of salt(). Not sure if it is a bug, because in attempts to figure this out rewrote this post many times and got confused even more.

I have a S=1/2 coupled to a proton (fL = 51.4 MHz) by A = 120 MHz and would like to get the line at higher frequencies right. I simulate 3 test spectra with identical parameters, but 'matrix', 'perturb2' and 'perturb1' methods. I have pasted an example below.

'perturb2' seems to be even worse than 'perturb1'. What could be a reason of this? Am I doing something wrong?

Code: Select all

A = 120;

Sys.S = 1/2;
Sys.lw = 0.5;
Sys.lwEndor = 0.3;
Sys.g = 2.0023;
Sys = nucspinadd(Sys,'1H',A);

Exp.mwFreq = 34;
Exp.Range = [0 120];
Exp.ExciteWidth = 20;
Exp.Field = 1213.2+mhz2mt(A)/2;
Opt.Threshold=1e-15;

Opt.Method = 'matrix';
[~,specmat] = salt(Sys,Exp,Opt);
Opt.Method = 'perturb2';
[~,specper] = salt(Sys,Exp,Opt);
Opt.Method = 'perturb1';
[freq,specper1] = salt(Sys,Exp,Opt);

figure(1)
clf
plot(freq,specmat/max(specmat),'-b',freq,specper/max(specper),...
    '-r',freq,specper1/max(specper1),'-k')
legend('matrix','perturb2','perturb1')

Thanks a lot for any kind of help!

EasySpin 5.0.2 and 5.1.7
MATLAB 8.2.0.701 (R2013b)

[Stefan Stoll]

This is indeed a bug - thanks for reporting! The second-order frequency for the mS = -1/2 ENDOR line in the case of strong hyperfine coupling is wrong due to an incorrect sign in the second-order correction.

The bug is fixed in version 5.1.8, which you can download now.

[jevkus]

That was quick! Thanks!