Author Topic: questions about new spinach version  (Read 3199 times)

faq_user

  • Member
  • *
  • Posts: 23
    • View Profile
questions about new spinach version
« on: November 15, 2013, 06:01:08 PM »
1. Is there a function you have written for spin-locking? I had an impression there is, but I cannot find it.

2. Is there a way now to conveniently extract the density matrix information from the tensor operator basis to the zeeman base matrix? I built each state like the singlet state before, but it is tedious for larger spin system.

3. I noticed in the JACS paper, you use M06/cc-pVTZ to simulate the system, have you compared the performace with b3lyp or other functional/basis set combination?

kuprov

  • Administrator
  • Member
  • *****
  • Posts: 58
    • View Profile
Re: questions about new spinach version
« Reply #1 on: November 15, 2013, 06:01:35 PM »
1.       Spin locking implementation depends on what you need to do – in many cases it simply amounts to putting all Zeeman interactions to zero for a period of time. Which experiment are you simulating?
 
2.       Yes – the current development version of Spinach can run calculations in both Zeeman and spherical tensor basis sets, you just flick a switch. The operators in the Zeeman basis are built in the usual way with direct products of Pauli matrices followed by the transformation into Liouville space.
 
3.       Accuracy of quantum chemical CSAs is a bit of a black art, unfortunately. See the last two lectures in the Quantum Chemistry course at http://spindynamics.org – the question of “what works” is basically empirical, and many methods produce right answers for wrong reasons due to error compensation.

faq_user

  • Member
  • *
  • Posts: 23
    • View Profile
Re: questions about new spinach version
« Reply #2 on: November 15, 2013, 06:01:55 PM »
Thank you very much for your reply! I am trying your latest published version now, 1.2.1437 but let me make sure I understand it right first.
 
1. I am trying to see spin locking effects on the singlet states. So for example after the M2S sequence I need to quench the chemical shift difference so that no leakage happens from the singlet to the triplet due to the coherent mixing. Please let me know how to implement that in spinach.
 
2. I see the "bas.formalism" input in the new version, I have not tried yet, but this is not in the 1.2.1437 version right? In that case, I will wait for the new version officially coming out. I guess if I can read out the zeeman base density matrix at the end of the evolution it will be very useful. As to calculation itself, I am not sure, does the change into zeeman basis slow down the calculation or increase the demand for RAM since you have an excellent way to store tensors as numbers, if I remembered right.
 
3. For the DFT, I guess I will run some training set to see if they make a systematic difference.
 
Again, thank you very much! I will for sure keep the version you gave me confidential, I notice also that the package you gave me probably is the core not including everything since it is much smaller than the 1.2.1437, but anyway, look forward to your new version!

kuprov

  • Administrator
  • Member
  • *****
  • Posts: 58
    • View Profile
Re: questions about new spinach version
« Reply #3 on: November 15, 2013, 06:02:13 PM »
1.       You can just remove that chemical shift difference from the Hamiltonian manually. There’s an option to the assume() function that allows you to drop all Zeeman interactions. Just make two Hamiltonians – one with Zeeman interactions and one without, and use them as needed.
 
2.       Yes, bas.formalism is not in the public release yet. It switches Spinach between spherical tensor basis and Zeeman basis. Redfield theory should not be any slower, we use complete basis sets for those calculations anyway.

faq_user

  • Member
  • *
  • Posts: 23
    • View Profile
Re: questions about new spinach version
« Reply #4 on: November 15, 2013, 06:02:30 PM »
Thank you very much for your last email. If I want to watch the power dependence of spin locking, I assume there maybe another to do it instead of wiping out zeeman terms, which is an infinite power solution, is there a way to do that in spinach? Also, I really like your lecture materials online, I will be very interested of course in Module III, where the relaxation theory and singlet states will be discussed, do you have a timeline for the course or when probably I will see the rest of module 3 and 4 online? Thanks a lot!

kuprov

  • Administrator
  • Member
  • *****
  • Posts: 58
    • View Profile
Re: questions about new spinach version
« Reply #5 on: November 15, 2013, 06:02:47 PM »
I’ll cook you a template for the spin-lock simulation, give me a couple of days.
 
Module III is starting on Wednesday, I will do the first four lectures and Giuseppe Pileio will do the other four. So we’ll get to those singlet lectures in about a month from now.