Cannot see the effect of J coupling between two electrons

[asifequbal]

% comparison of field sweep and frequency sweep cw EPR
%==========================================================================
clc
clear vars
%close all
%define two spins
Sys.S = [1/2 1/2];
% full g tensors for two BDPA electron spins
Sys.g = [2.0034 2.0031 2.0027; 2.0034 2.0031 2.0027];
J12 = 100; %MHz %%THIS J COUPLING DOESN'T AFFECT SPECTRUM. WHY?
Sys.J = J12;
%D12 = [1 1 -2]*10; % units: MHz
%Sys.eeD = D12;
Sys.lw = 0.5;

% % Frequency-sweep experiment
% %------------------------------------------------------------
% % For frequency sweeps, the convolutional width should be given in MHz.
Exp1.Field = 6906; % static magnetic field, mT
Exp1.mwRange = [193.2 194.4]; % frequency sweep range, GHz
Exp1.nPoints= 8192*5;

% Simulate frequency-sweep spectrum
[nu,specnu] = pepper(Sys,Exp1);
figure (2); hold on; plot(nu,specnu*2,'-');
title('frequency sweep');
xlabel('frequency (GHz)');

[Stefan Stoll]

This is not a software bug. It's a textbook example of two coupled spins-1/2 giving singlet and triplet states.

Since your radicals have exactly the same g values (and there is no hyperfine), the system eigenstates are αα, ββ, (αβ+βα)/sqrt(2), and (αβ-βα)/sqrt(2), independent of the value of the J coupling. This means that your system is a clean sum of a triplet (S=1) and a singlet (S=0), without any singlet-triplet mixing. Since EPR transitions have the selection rule Delta-S = 0, you won't see any transitions between the singlet state and any of the three triplet states. And the two triplet transitions have the same energy, yielding a single line in the EPR spectrum, independent of J.

[asifequbal]

Thank you for the explaination!

Shouldn't the g-anisotropy and Dipolar anisotropy induce a frequency difference leading to non singlet-triplet basis?

Should I assign a different euler angle set for PAS of electron-2 ?

[Stefan Stoll]

If the two g tensors have identical principal value and orientation, then you won't get S-T mixing. Either the principal value must be different, or the orientations, or both.