Hyperfine Tensor in Chili

[wlinde]

Hey,

I'm trying to determine whether the Chili simulation and nonlinear fitting function in EasySpin is a viable alternative to the NLSL distribution from ACERT. It seems that it should be able to operate identically. However, I am having trouble reproducing the sample-data that are released in the NLSL distribution. http://www.acert.cornell.edu/index_file ... _links.php. I think it's because I don't understand how to convert units for some of the variables.

I'm trying to simulate a spectrum using the parameters (copied from the NLSL run file):

{g} = 2.0089, 2.0021, 2.0058
{A} = 5.6, 33.8, 5.3 (gauss)
betad = 15 degrees
Rpll = 1e7
Rperp = 1e8
B0 = 3400 G
GIB = 2.0 G (p-p width of Gaussian inhomog. linewidth)

Here, g and A are identical, and my understanding is that betad is simply one of the Euler angles for the diffusional reference frame. Rpll and Rperp are the rotational diffusion coefficients and B0 is the static field. The software is specific to nitroxide radicals.

In EasySpin, I am running:

Sys = struct('g',[2.0089,2.0021,2.0058],'Nucs','14N','A',[.56,3.38,.53],'DiffFrame',[0,15*pi/180,0]); %g is unitless; A -- don't understand units;
Sys.Diff = [1e7,1e8]; %s^(-1)
Sys.lw=[.2]; %Line broadening of 2G, or 0.2 mT
Exp = struct('mwFreq',9.55,'CenterSweep',[340,10]); %I am totally guessing at the frequency, since I cannot find it in the literature.
chili(Sys,Exp);

The spectra look quite different, and I believe the reason is that I don't know how to convert the hyperfine tensor from G to MHz. In addition, I don't understand why the data-set from Cornell appears to use a static field, B0, yet the x-axis of the data clearly uses units of Gauss rather than Hz. I'm sorry I couldn't attach the data set, I attempted but apparently it can't handle .dat or .txt files.

Is there anyone who can help me with this? Hopefully I'm missing something basic -- I'm very new to EPR.

[Matt Krzyaniak]

There are two immensely useful functions in easyspin mt2mhz and mhz2mt, with those it's fairly easy to do the conversion. So: A = mt2mhz([.56,3.38,.53]);

As for the specific differences and such I can't exactly answer that, I'll leave that to Stefan.

[wlinde]

That works, thanks!

[Stefan Stoll]

Both NLSL and EasySpin calculate slow-motion spectrum for a static field and then convert the spectrum from frequency-domain to field-domain. This is a first-order-perturbation-type conversion that only works as long as the spectrum is relatively narrow, e.g. for nitroxides. EasySpin uses the center field of the given field sweep range for the frequency-domain calculation.