Author Topic: Calculation of spin population  (Read 4030 times)


  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
Calculation of spin population
« on: August 31, 2010, 10:09:20 pm »
I'm studing a reaction coordinate for a bond dissociation (metal-carbon bond, closed shell system). I'm performing the reaction coordinate using cartessian coordinates (fixing the corresponding spatial coordinates of the atoms involved in the bond, and with an unrestricted-DFT method). The problem is that when I calculate the spin populations along the scan, they are always zero (using TMv5.9), but when I used the TMv5.8 I obtained spin populations different from zero and they are very reasonable (close to one at long bond distances). I am studying the spin populations calculated from NPA analysis. I calculated the geometries with both versions separatelly (obtaining two different wave functions for each point of the scan) but I calculated the spin population with TMv5.9 (because as long as I know this method is not implemented in TMv5.8 ), and I obtained the results described above.

As a test I tried to reoptimize the geometries obtained with TMv5.9 using TMv5.8 and then I recalculated the NPA population. For cases where the energy decreased significatively the spin population was different from zero (but they were not reasonable with values of the order of 10E-5 for every atom even at long bond distances) and for study cases when no change on the energy was observed after the reoptimization the spin population remained zero.

The actual problem is that I have lots of data using the TMv5.9 (reaction coordinates), so I need to know if there is something reported related to this problem (I have found here something that could be related on "Interpretation of NBO analysis: Bug or stupid user?", but I'm not sure if that applies to my problem). Why the data using this two different version is different both in energy and geometry (as a consequence I should asume that the calculated wave functions for the same point of the scan are different too). I want to avoid calculating everything from scratch using the TMv5.8, but I do not really know if the data obtained with TMv5.9 is reliable at all, also I cannot do the spin population analysis for those models. I am in badly need of helpful advices. Thanks in advance;

Sebastián Miranda
Pharmacist. Chemistry PhD student.
Theoretical Inorganic Chemistry Lab.
Universidad de Chile.


  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 258
  • Karma: +0/-0
    • Hattig's Group at the RUB
Re: Calculation of spin population
« Reply #1 on: November 18, 2010, 06:12:23 pm »
Can you send a bug report together with some input files to you TURBOMOLE reseller?