Page 1 of 1

Error in orca2easyspin - rotation matrix

Posted: Wed Mar 18, 2020 4:44 am
by Urwald
Dear all,

I am using ORCA v.4.2.1 for calculating the g-tensors and I am importing the results to EasySpin using the orca2easyspin-function. When I am using DFT everything works fine. However, when I calculate the g-tensor with CAS-SCF pepper terminates with an error:

Error using eulang
eulang: Determinant of rotation matrix is +0.000, deviates too much from +1!
Rescale argument to R/det(R)^(1/3) if result wanted.

Error in orca2easyspin

Is there any possibility to fix this issue by hand?

Thank you and all the best,
Urwald

Re: Error in orca2easyspin - rotation matrix

Posted: Wed Mar 18, 2020 9:42 am
by Stefan Stoll
Which EasySpin version are you using?
Also, please post the .prop file from your calculation, as well as the relevant sections from the ORCA output file.

Re: Error in orca2easyspin - rotation matrix

Posted: Thu Mar 19, 2020 1:23 am
by Urwald
Hi Stefan,

I am using EasySpin 5.2.28. As an example I am using the [FeFCl5]3- complex, which has a sixtet ground state. I attached the corresponding .prop file of the CAS(5,5) calculation.
Some relevant informations from the output:

Code: Select all

--------------------------------------------
           ZERO-FIELD SPLITTING
(2ND ORDER SPIN-ORBIT COUPLING CONTRIBUTION)
--------------------------------------------

Second order D-tensor: component '0'
    0.000000     0.000000     0.000000 
    0.000000     0.000000     0.000000 
    0.000000     0.000000     0.000000 

Second order D-tensor: component '-'
    2.520570     0.000209     0.023864 
    0.000209     2.534509    -0.000344 
    0.023864    -0.000344     2.493718 

Second order D-tensor: component '+'
    0.000000     0.000000     0.000000 
    0.000000     0.000000     0.000000 
    0.000000     0.000000     0.000000 


Raw matrix (cm-1):
    2.520570     0.000209     0.023864 
    0.000209     2.534509    -0.000344 
    0.023864    -0.000344     2.493718 

Eigenvalues (cm-1):
    2.534509     2.534528     2.479760 

Eigenvalues (traceless) (cm-1):
    0.018243     0.018262    -0.036506 

Eigenvectors:
    0.319590     0.801899    -0.504797 
   -0.927316     0.374206     0.007359 
    0.194799     0.465755     0.863207 

Euler angles w.r.t. molecular frame (degrees):
     67.3032      30.3214       0.8352

D   =   -0.054759  cm-1
E/D =    0.000171

Individual contributions to D-tensor:
Block  Mult  Root          D          E
    0     6     0      0.000      0.000
    1     4     0      0.854      0.067
    1     4     1     -0.324      0.456
    1     4     2     -0.364     -0.364
    1     4     3     -0.042     -0.042
    1     4     4     -0.000     -0.000
    1     4     5     -0.000      0.000
    1     4     6     -0.000      0.000
    1     4     7     -0.004      0.009
    1     4     8      0.014      0.002
    1     4     9     -0.000     -0.000
    1     4    10     -0.000     -0.000
    1     4    11     -0.087      0.104
    1     4    12      0.192      0.009
    1     4    13     -0.000      0.000
    1     4    14     -0.834     -0.834
    1     4    15     -0.414      0.548
    1     4    16      1.031      0.067
    1     4    17     -0.000     -0.000
    1     4    18      0.004      0.001
    1     4    19     -0.001      0.002
    1     4    20     -0.000     -0.000
    1     4    21      0.006      0.000
    1     4    22     -0.003      0.003
    1     4    23     -0.000     -0.000
Norm of projected state 0 of the effective Hamiltonian: 0.999293
Norm of projected state 1 of the effective Hamiltonian: 0.999293
Norm of projected state 2 of the effective Hamiltonian: 0.999319
Norm of projected state 3 of the effective Hamiltonian: 0.999319
Norm of projected state 4 of the effective Hamiltonian: 0.999332
Norm of projected state 5 of the effective Hamiltonian: 0.999332
WARNING: If the projections have norms significantly smaller than 1, the effective Hamiltonian might not be very meaningful!

--------------------------------------------------------
                  ZERO-FIELD SPLITTING
(EFFECTIVE HAMILTONIAN SPIN-ORBIT COUPLING CONTRIBUTION)
--------------------------------------------------------

Effective Hamiltonian from projected relativistic states and relativistic energies:

Real part:  
                  0          1          2          3          4          5    
      0       0.084604   0.129415  -0.027168  -0.000093  -0.000127  -0.000000
      1       0.129415   0.249295   0.081872  -0.036367   0.000000  -0.000127
      2      -0.027168   0.081872   0.331613  -0.000000  -0.036367   0.000093
      3      -0.000093  -0.036367  -0.000000   0.331613  -0.081872  -0.027168
      4      -0.000127   0.000000  -0.036367  -0.081872   0.249295  -0.129415
      5      -0.000000  -0.000127   0.000093  -0.027168  -0.129415   0.084604

Image part: 
                  0          1          2          3          4          5    
      0      -0.000000   0.002080  -0.000724   0.000005   0.000006   0.000000
      1      -0.002080  -0.000000   0.001311  -0.000978   0.000000   0.000006
      2       0.000724  -0.001311  -0.000000   0.000000  -0.000978  -0.000005
      3      -0.000005   0.000978  -0.000000   0.000000  -0.001311  -0.000724
      4      -0.000006  -0.000000   0.000978   0.001311   0.000000  -0.002080
      5      -0.000000  -0.000006   0.000005   0.000724   0.002080  -0.000000

Raw matrix (cm-1):
    0.030497     0.000230     0.028940 
    0.000230     0.047654    -0.000465 
    0.028940    -0.000465    -0.002093 

Eigenvalues (cm-1):
    0.047409     0.047664    -0.019014 

Eigenvalues (traceless) (cm-1):
    0.022057     0.022311    -0.044367 

Eigenvectors:
   -0.855050     0.119263    -0.504645 
   -0.142629    -0.989746     0.007758 
   -0.498545     0.078610     0.863292 

Euler angles w.r.t. molecular frame (degrees):
     -8.9606      30.3117       0.8807

D   =   -0.066551  cm-1
E/D =    0.001910

Individual contributions to D-tensor:
Block  Mult  Root          D          E
    0     6     0      0.000      0.000
    1     4     0     -0.503      0.519
    1     4     1      1.021      0.008
    1     4     2     -0.364     -0.364
    1     4     3     -0.042     -0.042
    1     4     4     -0.000     -0.000
    1     4     5      0.000      0.000
    1     4     6      0.000     -0.000
    1     4     7      0.021      0.001
    1     4     8     -0.009      0.010
    1     4     9     -0.000     -0.000
    1     4    10      0.000     -0.000
    1     4    11      0.224      0.000
    1     4    12     -0.108      0.108
    1     4    13      0.000     -0.000
    1     4    14     -0.834     -0.834
    1     4    15      1.214      0.006
    1     4    16     -0.599      0.610
    1     4    17      0.000      0.000
    1     4    18     -0.002      0.002
    1     4    19      0.005      0.000
    1     4    20     -0.000     -0.000
    1     4    21     -0.003      0.003
    1     4    22      0.006      0.000
    1     4    23     -0.000     -0.000

  *** Storing geometry in DTensor axis frame as fecl6.CAS_DTensor_HEFF_rotated.xyz  ***

--------------
KRAMERS PAIR 1
--------------

Matrix elements Re<1|S|1>   -1.074197   0.016558   1.839078 
Matrix elements Re<1|S|2>   -0.639785   0.010116   1.092145 
Matrix elements Im<1|S|2>    0.165944  -0.001177  -0.283411 
Matrix elements Re<1|L|1>   -0.000285   0.000004   0.000488 
Matrix elements Re<1|L|2>   -0.000170   0.000003   0.000290 
Matrix elements Im<1|L|2>    0.000044  -0.000000  -0.000075 

-------------------
ELECTRONIC G-MATRIX
-------------------

g-matrix:
   -2.562446    -0.664632    -4.302340 
    0.040516     0.004712     0.066319 
    4.374227     1.135108     7.365818 

g-factors:
    0.005620     0.005635    10.010075 iso =     3.340443

g-shifts:
   -1.996699    -1.996685     8.007756 iso =     1.338124
 Orientation: 
  X          -0.1143051    0.8557266    0.5046447
  Y           0.9905559    0.1368901   -0.0077577
  Z          -0.0757193    0.4989920   -0.8632923

Notes:  (1) The g-matrix conforms to the "BgS" spin Hamiltonian convention.
        (2) The principal values are square roots of the eigenvalues of g*gT
        (3) Orientations are eigenvectors of g*gT written as column vectors
        (4) Tensor is right-handed


-----------------------------------
ELECTRONIC G-MATRIX: S contribution
-----------------------------------

g-matrix:
   -2.562107    -0.664544    -4.301770 
    0.040510     0.004712     0.066310 
    4.373648     1.134958     7.364842 

g-factors:
    0.005620     0.005634    10.008749 iso =     3.340001

g-shifts:
   -1.996700    -1.996685     8.006430 iso =     1.337682
 Orientation: 
  X          -0.1143097    0.8557260    0.5046447
  Y           0.9905551    0.1368954   -0.0077577
  Z          -0.0757220    0.4989916   -0.8632923

Notes:  (1) The g-matrix conforms to the "BgS" spin Hamiltonian convention.
        (2) The principal values are square roots of the eigenvalues of g*gT
        (3) Orientations are eigenvectors of g*gT written as column vectors
        (4) Tensor is right-handed


-----------------------------------
ELECTRONIC G-MATRIX: L contribution
-----------------------------------

g-matrix:
   -0.000339    -0.000088    -0.000570 
    0.000005     0.000001     0.000009 
    0.000579     0.000150     0.000976 

g-factors:
    0.000001     0.000001     0.001326 iso =     0.000442

g-shifts:
   -2.002319    -2.002319    -2.000993 iso =    -2.001877
 Orientation: 
  X          -0.1009205    0.8574389    0.5045925
  Y           0.9925648    0.1214623   -0.0078802
  Z          -0.0680458    0.5000455   -0.8633217

Notes:  (1) The g-matrix conforms to the "BgS" spin Hamiltonian convention.
        (2) The principal values are square roots of the eigenvalues of g*gT
        (3) Orientations are eigenvectors of g*gT written as column vectors
        (4) Tensor is right-handed


-------------------------------------------------------------------------------------------
                       TRANSITION MAGNETIC DIPOLE CONTRIBUTIONS  
       (theta is the angle between the gz component of the lowest and the nth Kramers pairs)
-------------------------------------------------------------------------------------------
Ms States     Energy        M2            MX             MY            MZ            theta  
              (cm-1)      (au**2)        (au)           (au)          (au)           (Deg)  
-------------------------------------------------------------------------------------------
 0  0         0.00     0.00000032     0.00028495     0.00000446     0.00048788     180.00
 0  1         0.00     0.00000012     0.00017529     0.00000274     0.00029934     180.00
 1  1         0.00     0.00000032     0.00028495     0.00000446     0.00048788     180.00

----------------------------------------------
ELECTRONIC G-MATRIX FROM EFFECTIVE HAMILTONIAN
----------------------------------------------
Spin multiplicity = 6

g-matrix:
    2.002018     0.000000     0.000000 
    0.000000     2.002018     0.000000 
    0.000000     0.000000     2.002018 

g-factors:
    2.002018     2.002018     2.002019 iso =     2.002018

g-shifts:
   -0.000301    -0.000301    -0.000301 iso =    -0.000301
 Orientation: 
  X          -0.5108070   -0.2188438   -0.8313745
  Y          -0.1425724    0.9752302   -0.1691129
  Z           0.8477909    0.0321470   -0.5293555

Notes:  (1) The g-matrix conforms to the "BgS" spin Hamiltonian convention.
        (2) The principal values are square roots of the eigenvalues of g*gT
        (3) Orientations are eigenvectors of g*gT written as column vectors
        (4) Tensor is right-handed
I just did some tests and it might be a problem with the .prop file, since also certain ORCA stand-alone programms do not recognize g- or D-tensors written to the .prop file.

Best,
Urwald

Re: Error in orca2easyspin - rotation matrix

Posted: Thu Mar 19, 2020 2:36 pm
by Stefan Stoll
Indeed, the CAS-SCF g-tensor is not stored in the .prop file, even though it should. The .prop file contains a list of entries with IDs. The standard-SCF-based g tensor has ID 5, and the CAS-SCF g-tensor has ID 23. In your file, there is no entry with ID 5, and there is an entry with ID 23 that has all zeroes.

This looks like an ORCA bug, so it's best to submit this issue on the ORCA forum.

We'll improve the error message in EasySpin for this case.

Re: Error in orca2easyspin - rotation matrix

Posted: Fri Mar 20, 2020 12:09 am
by Urwald
Dear Stefan,

thank you very much for your helpful suggestions.

Best,
Urwald