Problem with Spin polarization graphs

[NMR-trainee]

Greetings, Prof. Kuprov.

I'm currently making spin polarization graphs, and some of my results are unphysical. My experiment has one hyperpolarized OH group, and the spin polarization graphs show how this polarization changes between all the atoms in the system, while the total should remain constant.

Most of my results are physical, but then again most of my 0 T field calculations give unphysical results. I have attached a picture of my code and the unphysical graph that it produces for this system. The black line is the total polarization and red is the hyperpolarized group. I have tested different timesteps and step lenghts.

I was wondering, if spinach has something specific going on with zero field sys.magnet? And if there is something that I should take into account here.

Other thing that came to mind, was if I should multiply the evolutions of H1 and H3 by 3, since the system has 3 of those atoms. In another system, I received correct results this way, but do I then need to have all 3 symmetrical atoms in the detection state, or only one of them?

As always, many thanks for the help!

[kuprov]

Simple - the 'nmr' assumptions (in the Hamitonian function call) are essentially high-field secular approximation. That does of course break down at zero field. Use 'labframe' assumptions instead, further details here:

https://spindynamics.org/wiki/index.php?title=Assume.m

[NMR-trainee]

Thanks! This is just what I needed. Now, if I use 9.4T or 6.5mT sys.magnet, should I then use 'nmr'?

[kuprov]

That depends on what your dynamics is. The safe choice is always 'labframe', although the calculation might be very slow and require very short time step because you are explicitly digitising very high frequencies.

You can use 'nmr' when the rotating frame is a good approximation. That is roughly when the amplitude of your interactions is much smaller than the amplitude of the Zeeman interaction.

[NMR-trainee]

This was very helpful, thank you!