it seems this forum is an exclusive chance to communicate with EPR-simulator-profs

My task is to simulate a cw-EPR spectrum (X-band) of a powder of silicon nanoparticles (crystalline according to electron diffraction).
The picture is below. The problem is to simulate the double-peak (in the red circle).

I've tried to simulate a complex signal (isotropic one at g=2.0055 + anisotropic from Pb-centers) but after 3 days of simulations my emotion is despair

There are too many variables (g-factors, hstrains, weights) to do it manually, but the esfit function does not, it seems, consider this double-peak!
Can anyone help me to choose an appropriate method and target (data as is, integral, etc?) for this task? Maybe some advice abous spin system or something else?

Code just in case:
Code: Select all
clear
[G,spc]= eprload('file')
Exp.mwFreq = 9.6471;
Exp.nPoints = 1024;
Exp.ModAmp = 0.1; %mT 1 mT = 10 G
Exp.CenterSweep=[344.6 10]; %mT 1 mT = 10 G
Sys1.g = [2.0055];
Sys1.HStrain = [20]; % mHz. 10G=28mHz
Vary1.g = [0.0005];
Vary1.HStrain = [15];
Sys2.g = [2.009 2.0015];
Sys2.HStrain = [20 10];
Vary2.g = [0.002 0.001];
Vary2.HStrain = [10 6];
Sys1.weight = 1;
Sys2.weight = 1;
Vary1.weight = 0.8;
Vary2.weight = 0.8;
FitOpt.Method = 'genetic fcn';
FitOpt.PopulationSize = 40; % def=20
FitOpt.Scaling = 'lsq';
esfit('pepper',spc,{Sys1,Sys2},{Vary1,Vary2},Exp,[],FitOpt);%multiple components