J couplings from fluid-solution EPR spectra

[thanasis]

Hello,

I would like to use fluid solution EPR spectra of an asymmetric biradical with hf-couplings to determine the isotropic coupling, J.

I know there is theory for that (e.g. here, "CUNO" program, then here, here and here), but not much is said in terms of implementation.
Moreover, from what I understand garlic cannot treat more than one spins.

Is there a possibility of introducing this to garlic? If so, I could post a git request.

In the mean time, does anyone know if there a way to tackle this problem without solving it from scratch?

Thanks!

[katarkon]

The pepper() function allows to simulate the spectra of biradicals.

[thanasis]

Thanks for the rapid reply.

Indeed, pepper can handle biradicals, but powder or frozen solution spectra.

From what I understand from reading the theory, the fact that these are fluid solution spectra needs to be specifically accounted for, no?

[Matt Krzyaniak]

If you keep the interactions isotropic, pepper works well for simulating radicals in the fast motion regime. If you have some dynamics that push you slightly toward the slow motion regime, where you need Tcorr or the functionality of chili, unfortunately EasySpin can't deal with two spins at the moment.

[thanasis]

Thanks for the suggestion Matt.

I did try that, and one thing comes out immediately: in my case, and other fluid solution spectra in the literature, each spin system may exhibit very different linewidths. However, using pepper imposes a unique linewidth to the entire spectrum and simulations are really very off the mark.

In Eaton & Eaton's original CUNO program, each line was assigned a linewidth of A + BmI + CmI², where mI is the nuclear spin. Subsequently, arbitrary line widths were introduced (Inorg. Chem., 1982, 21, 1094).

See also the figure below the very narrow nitroxyl and much broader vanadyl lines:

Selection_144.png (21.35 KiB) Viewed 5169 times

[katarkon]

You may try to use chili() function. It works with arbitrary spin systems including biradicals, but calculations seem to be extremely slow.
Here the exmple script.

Code: Select all

% Simple slow-motional cw EPR spectrum simulation of a nitroxide diradical
%==========================================================================
clear, clf

% Parameters
%--------------------------------------------------------------------------
% Static parameters
Nitroxide.S=[0.5 0.5];
Nitroxide.g = [2.008,2.006,2.004;2.008,2.006,2.004];
Nitroxide.Nucs = '14N,14N';
Nitroxide.A = [20 20 20 20 85 85;20 20 20 20 85 85];
Nitroxide.J=2.0e1;

% Dynamic parameters
Nitroxide.lw = [0.01 0.1];
Nitroxide.tcorr = 3e-9;

% Experimental parameters
Experiment.mwFreq = 9.65;
Experiment.Range=[334 354];
Experiment.nPoints=8192;
%Experiment.Temperature = 290;
Opt.Verbosity=0;

% Simulation and graphical rendering
%--------------------------------------------------------------------------
tic;
chili(Nitroxide,Experiment,Opt);
toc;

[thanasis]

Thanks for this.

I am trying it out as we speak with the logtcorr near its limits (ca. -14). It is slow as you say, but that's not the main issue.

It is rather the same as above, that it gives a global line width for the entire spectrum.

[katarkon]

Alternative way is custom simulation function. You may to obtain the resonance fields and the intensities via resfields() and further define the linewidths and lineshapes for separate lines individually. The problem that resfields() sorts its output according to the energy of the levels, there are no clear way to rid it out.

[Stefan Stoll]

Supporting J-coupled spin pairs seems to be a useful feature to add to EasySpin garlic. I opened an issue on GitHub. Please post there if you have specific suggestions on how to implement this.