This describes the importation and further processing of an output file from the program GAUSSIAN, specifically a calculation on benzene C6H6.  The main importation is of the formatted checkfile (.fchk or .fch).  The files needed for import are in the folder GAUSSIAN, and the sample VIBRATZ data files from the steps in the process are in the Completed subfolder.

1.  From the Startup Window, click on the Import button, after selecting GAUSSIAN fch from the list box to the right.  Click OK in the Import File dialog, then find the file c6h6.fch in the GAUSSIAN sub-folder.   Open (import) this file.  This creates the Cartesian force file c6h6.car, which contains values identical to those in the c6h6.fch file except for the factor to convert from  Hartree/Bohr to md/A.   The name of this file appears in the Cartesian Atomic Forces from File dialog, accessible with the Set File button in the Cartesian Forces group, or from the Forces menu

Now you should be in the VIBRATZ Control Window.  Use SaveAs to save the VIBRATZ data file as c6h6_1.vbr  (or benzene_1.vbr).  You can click on the Calculate button, and this should produce a list of 30 frequencies in the Vibratz Output Window - these frequencies should be within a few tenths of a wavenumber of the calculated frequencies in c6h6_freq.txt (extracted from the main GAUSSIAN output file, c6h6.out) - open this file with any text editor or word processor).

2.  (Optional) Now read in some observed frequencies.  In the Input menu, select Observed.  In the Read File section to the lower right, select browse and open (select) the file benzene_nosym.frq - this is a list of known modes for benzene, arranged in order of frequency.  The observed frequencies should appear.  Exit the dialog with OK.  Click on Calculate and the observed frequencies should appear in the final list in the Output Window.  These observed frequencies are much lower than the calculated frequencies, because of bias in the GAUSSIAN calculation.  This discrepancy can be reduced by refining a scale factor which multiplies all the elements of the Cartesian force matrix.

In the Least-squares box in the Control Window, select Step-by-step and then click on Calculate.  A message box should appear with the query "Derived scale factor is 0.6903 - keep this?".  Select OK.  The new scale factor can be seen in the Cartesian Atomic Forces from File dialog, accessible with the Set File button in the Cartesian Forces group, or from the Forces menu.  Save the file as C6h6_2.vbr.  Change the Least-squares setting to None before clicking on Calculate again.

3.  In order to convert to valence or internal force constants, you must have a complete set of specifications in the various dialogs of the Forces menu.  Such a set of forces has been saved from an independent standard valence calculation in the file benz_val_complete.frc.  You can read in these specifications with the Read Force File option in the Forces menu.  Now you can reproduce the valence calculation, if desired, by changing the setting (Choice) in the Cartesian Atomic Forces from File dialog to "Do not use Cartesian forces". 

Of course you could enter your own specifications for forces through the dialogs in the Forces menu.  The specifications could be "Manual" instead of "Automatic", that is they could use actual generated atom numbers instead of bond and angle limits.  However, automatic specifications are usually preferred because they can be used even if the symmetry changes (changing the symmetry may change the order of generated atoms), and may be transferrable to other similar compounds.

If you are not familiar with specifying forces in VIBRATZ it may be a useful exercise to try to reproduce the forces as given in the Internal Coordinate Types section of the output (Output Window), by entering them in the dialogs of the Forces menu.

In order to convert the GAUSSIAN forces (in the file c6h6.car) to valence forces according to the specifications, select the the option Convert Cartesian to Internal in the Forces menu.  Reply Yes when asked if you want to replace the old values.  Click on the Calculate button, and the Output window will show the results of calculation with the valence forces derived from the GAUSSIAN calculation.  Save this file as c6h6_3.vbr.

4.  (Optional) As the GAUSSIAN output seems to lack symmetry, symmetry has not been used so far.  It is best to use symmetry, if observed frequencies can be assigned to symmetry species.  Some changes have to be made to the symmetry setting and the input atoms.

Change the Symmetry (Input menu) to point group D6h.  Delete all but the first C and first H atoms (Input Atoms in the Input menu).  Set the z coordinates to zero for both atoms (these coordinates should be very small anyway).   With the "Rotate all input atoms" option, rotate by -90 degrees on z, so that both atoms lie on the x axis (y and z coordinates zero).

The Observed frequencies are not arranged according to species in D6h.  In the Observed dialog (Input menu), you can read in the benzene_sym.frq file.  This has the same observed frequencies as the benzene_nosym.frq file, but arranged according to species and with degenerate mode removed.

Save this file as c6h6_4.vbr.  Now a calculation has vibrations properly assigned to species.  The agreement with observed is not as good as in step 3, indicating that a simple list of modes in order of frequency does not lead to a proper match.

5.  (Optional) The force constants can be refined by least-squares. Select Step-by-step in the Least-squares box, and 0.001 for both the End criterion and Increment.  A calculation refines the force constants back to nearly the same values as originally obtained with the valence calculation.

-------------------------------------------

Note that the imported force specifications (step 3) have some manual bond-bond interactions.  These are dependent on the specific order of generated atoms and bonds.  Fortunately, the arrangement in GAUSSIAN and that generated by VIBRATZ from the symmetry-unique atoms in c6h6_4.vbr give the same result, as would any arrangement of the atoms in sequence either clockwise or counter-clockwise.  These interactions can be omitted, but the agreement between calculated and observed is distinctly worse.

Importation from other programs.  It is possible to do essentially the same thing as above with output from other programs which produce a Cartesian force-constant matrix.  What is covered in step 1 above will have to be done in two steps.  First the atomic coordinate information can be imported with the Import Free-form option.  The atomic coordinates should be saved in a separate file with extension .inp.  If the units are not Angstroms, it may be necessary to use the FACTOR keyword in the .inp file (a factor of 0.529177 converts from Bohrs to Angstroms).  The Cartesian force matrix should also be in a separate file with the .car extension, which is linked to in the Cartesian Atomic Forces from File dialog.  If the forces are not in md/A, a scale factor may need to be entered in this dialog (use 15.56914 to convert from Hartress/Bohr to md/A).  The Cartesian force matrix should be in lower diagonal form - the format is irrelevant.