ComQum-EXAFS

• QM+EXAFS, i.e. QM suplemented by EXAFS (or an automatic EXAFS-refinement protocol, where QM is used to suplement the EXAFS data to get a full set of coordinates, like MM is used in crystallographic refinement).
• QM/MM+EXAFS, i.e. QM/MM suplemented by EXAFS inside a full protein.
  

• QM+EXAFS: Y.-W. Hsiao, Y. Tao, J. E. Shokes, R. A. Scott, and U. Ryde (2006) "EXAFS structure refinement supplemented by computational chemistry", Phys. Rev. B, 74, 214101; DOI:  10.10.1103/PhysRevB.74.214101.
• QM/MM+EXAFS: U. Ryde, Y.-W. Hsiao, L. Rulíšek & E. I. Solomon (2007) "Identification of the peroxy adduct in multi-copper oxidases by a combination of computational chemistry and EXAFS measurements", J. Am. Chem. Soc., 129, 726-727; DOI: 10.1021/ja062594g; free reprints.

How to set up a ComQum-EXAFS calculation

1. Obtain the EXAFS data, preferably already treated with background removal, normalisation, splining, and conversion to k-space.
   The IFEFFIT program requires the k-interval to be 0.05 Å^-1 for the data, so one may need to do an interpolation --- See the IFEFFIT page.
   This file is called inter*.chi below.
   
   
2. Run the program turbo2feff to get the feff.inp file (you need to have a Turbomole calculation with coordinates to start with). Preferably, you should have first run a geometry optimisation and then a frequency calculation. turbo2feff reads force constants calculated by describe from a geo file and sets up also the spring.inp file (this will give Debye-Waller factors also).
   Consider to change the keywords TITLE, RPATH, NLEG, CRITERIA, HOLE, and DEBYE (cf. the FEFF page).
   Then, run FEFF by typing feff82
   The output will be lots of feffxxxx.dat files, which are the paths.
   The files files.dat and paths.dat give you more detailed information about the paths.
   
   NB: you need to have a separate directory for every absorbing atom (central metal) if you have several metals and you need to modify the feff.inp files by hand (by default the first heavy atom is given potential 0 and put first in the list). You must ensure that the scattering atom is always first in the list.
   
   The program swapabsorber may facilitate this renumbering of the atoms in files feff.inp and spring.inp. You just select the atoms to swap.
   

   The "testcriwmaxdosfit" (by Xichen Li) is used to accept the S02, the first 
   number of CRITERIA, the wmax, and the dosfit values from the 
   command line arguments, and then update feff.inp & spring.inp 
   and run "feff83", and finally collect the number of "WARNING" 
   and the number of check0 values larger than 10%. You could just 
   type "testcriwmaxdosfit -s 0.225 -c 0.1 -w 3.5 -d 0.1" or type
   "testcriwmaxdosfit -s0.225 -c0.1 -w3.5 -d0.1" or just type
   "testcriwmaxdosfit" or something in between. This utility is
   intended to be combined with the BASH for loop to get the 
   impression of a series of combination of FEFF parameters.

   
   
3. Run the program rmdegen to remove all degeneracies and set up the ComQum-EXAFS files.
   It will give you the two files comqum.dat and temp.iff.
   You need to have the control and coord file in each directory.
   The program may crash if you have ** in the file s2_em.dat. Remove them by hand.
   
   
4. Modify the file comqum.dat file (sample file).
   You should check the atoms for which you want to calculate the gradients in the keyword $exafs_gradients (see below for an explanation); by default internal gradients are calculated for all paths shorter than 3.0 Å.
   You may change $weight_exafs and perhaps also $weight_qm.
   
   You man also turn on the double-sided gradients (increases the time by a factor 2 in the EXAFS calculations):
   $double_sided_gradient on
   Finally, you may change the step size for the numerical EXAFS gradients (although the default is normally OK):
   $g_step
    1.0D-6
   
   
5. Set up the Turbomole geometry optimisation as usual (i.e. you can just copy the Turbomole files to the current directory)l.
   Always use internal or redundant coordinates (unless you will do a QM/MM calculation).
   
   For QM/MM-EXAFS, you need to copy the ComQum keywords from the old comqum.dat to the new one.
   
   
6. You need the following files in your directory (or will links), besides the Turbomole files
   *chi
   feff*.dat
   comqum.dat
   temp.iff
   
   mkdir Gz
   mv * Gz
   cp Gz/*chi Gz/feff*.dat Gz/comqum.dat Gz/temp.iff .
   gzip Gz/*&
   
   or
   
   mkdir Cqe
   cp * Cqe
   cp Exafs/*chi Cqe
   cp Exafs/feff*.dat Cqe
   cp Exafs/comqum.dat Cqe
   cp Exafs/temp.iff Cqe
   rm Cqe/*inp
   
   
7. If you intend to run with UFF, insert the following lines at the end of the control file:
   $uff
            1        -1          1 ! maxcycle,modus,nqeq
       111111                      ! iterm
     0.10D-07  0.10D-04            ! econv,gconv
         0.00  1.10                ! qtot,dfac
     0.10D+03  0.10D-04       0.30 ! epssteep,epssearch,dqmax
           25      0.10       0.00 ! mxls,dhls,ahls
         1.00      0.00       0.00 ! alpha,beta,gamma
            F         F          F ! transform,lnumhess,lm
   
   Change the yellow entry (qtot) so that it reflects the correct charge of the molecule.
   Then run
   uff>logu
   to get the topology file ufftopology.
   This file can now be modified before the program is run
   Finally change modus to 1 (marked in red).
   You can delete the alpha, beta, mos, auxbasis, and basis files and the corresponding keywords in control, but the $atoms keyword is still needed (for relax).
   
   
8. It is normally wise to first run (in a separate directory)
   ridft>logd
   rdgrad>logg
   cqefixenergy>loge
   cqefixgrad>logi
   And then move to the original comqum.dat file
   $norm_scale
    0.67531351E-09 (or whatever)
   Otherwise, the first energy will be different from the following ones.
   
   
9. Start ComQum-E by
   for QM+EXAFS
   jobexe -backup -ri -c 200
   or
   jobexe -backup -level uff -c 200
   or for QM/MM+EXAFS:
   comqume -backup -ri -c 200
   or
   comqume -backup -level uff -c 200
   
   
10. When finished, the results are in the files energy (the first energy is the ComQum-EXAFS energy, the third is the Turbomole energy, and the fourth is chi_squared EXAFS energy) and exafse.out:
    $energy      SCF               SCFKIN            SCFPOT
         2 -4043.63646887853  4024.47271869400 -4043.63702859200 22325.04453929200
    You may remove the comment (#) before the (new)plot lines in the exafs.inp and run
    ifeffit -x exafs.inp to get the final plots (exafs.ps).
    
    
11.  If you have problem with convergence, you may:

• reduce dqmax to 0.03 (or smaller; instead of 0.3) to improve the convergence, especially if w_Exafs is high.
  $coordinateupdate
     dqmax=0.03
     interpolate  on
     statistics    5
• Use steepest descent
  $forceupdate
     none
     allow=0.0 threig=0.005  reseig=0.005  thrbig=3.0  scale=1.00  damping=0.0
• Restart the forceapprox file
  $forceinit on
     diag=default
  
  
  

1. Finally, you can remove unnecessary files by typing purtur and then gzip everything (make sure that you have backup copies of the inter*.chi and feff????.dat files, because they are deleted).
   

• comqume
• cqefixenergy
• cqefixgrad
  

• turbo2feff
• feff2iff
  rmdegen
• changeexafs

• atom.dat   - Some information, including the atom number of the absorbers and atoms for which gradient should be calculated. Provided as input and read by xyz2r.
• chi.dat  - contains the EXAFS energies for the gradient calculations. Written by inp_gen in IFF.pe and read by IFF.pe and r2xyz.
• chisq.dat   - contains the EXAFS energy in each iteration. Written by IFF.pe and read by r2xyz.
  
• control.old  - temporary backup file of control. Written and read by d.pe and IFF.pe
• cycle.dat  - Contains only the iteration cycle number. Written by comqume and read by xyz2r (but only used by r2xyz, via the xyz2r.dat file).
• delr.dat  - Changes in all the path lengths in the gradient steps. Written by xyz2r.x and read by inp_gen, both within IFF.pe.
• e_diff.dat  - contains only the last energy change. Written by r2xyz and read by d.pe and IFF.pe
• exafs.log  - contains the current ifeffit results. Written by ifeffit and read by IFF.pe
• grad.dat  - contains the EXAFS gradients. Written by IFF.pe. Not used?
  
• inp.*.iff  - ifeffit input files for the gradient calculations. Written by inp_gen in IFF.pe and read by ifeffit.
• jndex.dat  - Contains the number of paths, the number of legs, and the atoms in the legs for all the paths. Provided as input and read by xyz2r.
• reff.dat  - Contains the R(effective) for all the paths. Provided as input and read by xyz2r.
• scale.dat  - Contains kappa (the rms ratio between Turbomole and EXAFS gradients. Written and read by r2xyz
  
• START.0 = job.1 for the first iteration (for information only). Written by comqume
• temp.iff  - A ifeffit template input file. Provided as input and read by inp_gen in IFF.pe.
• xyz2r.dat  - Almost the same information as in atom.dat (+ cycle iteration number). Written by xyz2r and read by r2xyz.