Saffron 15N 3P Eseem, high hfcc?

[Timothee_Chauvire]

Hi everybody,
I'm trying to simulate some 15N 3P-Eseem experiment on flavoproteins using saffron.
I detect an issue that come apparently when the hfcc are high.
Here is the script that I prepare with one 15N nucleus coupling with 1 electron.
I use for the demonstration three different hfcc axial tensor (MHz): [1 15], [1 30] and [1 60].
(I know that from cw-ESR simulation, the expected Azz coupling is around 40-80 MHz)
You can see some wiggles appearing at longer evolution time, and the higher the hfcc is, at a shorter time they appear.
Any ideas or suggestions to solve this problem? (I am using the last version of Easyspin 5.2.25, and even try the development version 6.0.0)
Thanks for your help.

Timothee

[nwili]

Not a bug.

Increase the number of sampled orientations. (nKnots)
Artificial "echoes" in the time trace indicate an insufficient sampling in frequency domain, i.e. not enough orientations. You can also see it in the the spectra.
The problem is not really that the hfcc are high, but because the anisotropies of the hfccs are high.

[TimotheeChauvire]

Hello,
Thank you for your previous help Nino Wili it solved my problem.

I tried to simulate some 3P-Eseem spectra and I’m still trying to understand what the time equivalence between a Brüker experiment and a 3p-Eseem easyspin fitting. And it would be really important fro my simulation as I have a series of eight 3p-Eseem spectra with different tau values (that I am summing to avoid tau cancellation effect).
If I understand well, Brüker spectrometers define the time scale in function of T (the evolution time) for a 3p-Eseem experiment.
On the other hand, the output time given by easyspin seems T+tau. Am I right for this assumption?
Also, you described for the self-defined pulse sequence that Brüker and easyspin have different time definitions. Is it true for the already defined 3P-Eseem pulse sequence in easyspin? Should I compensate the length of the pi/2 pulse in the time definition of T and tau for the 3p-Eseem easyspin simulation?
Thank you for your help.
Best

[nwili]

Honestly I cannot really help. It also depends on how the Pulse sequence and axis was exactly defined in your Bruker PulseSpel script.
In the end, does it really matter? it's just a time shift, and usually you would compare the Magnitude FTs anyway, where the time shift should not really matter, right?

[Timothee_Chauvire]

it's just a time shift, and usually you would compare the Magnitude FTs anyway, where the time shift should not really matter, right?

I agree that the frequency domain isn't impacted. And that one 3pEseem time domain with a fixed tau value will be just time-shifted.
There is two issues however:
1) I was hoping to find a better fit for the spectra, and Saffron/esfit will do it in the time domain... I guess a shift in time could induce some problem to find the right solution.
2) And as I'm simulating different tau values spectrum, the question is: Is it correct to sum these different simulated easyspin spectra with different time axis?
I'm still exploring anyway...

[nwili]

1) you can fit them in whatever way you want? you can simply write your own function that uses saffron and then does the processing you want and then you can use esfit. Or I think you can use the fitOption that the FTs are fitted, not completely sure about that.
2) I think usually the FTs of different tau spectra are added up