Common Problems - Troubleshooting
Previous  Top  Next

Whenever problems are encountered, the first thing to check is usually the generation of atoms by symmetry. The generated atoms are in the list All Atoms in Molecule/Unit Cell section in the printed Output. If the number of generated atoms is wrong, or if their coordinates are wrong, check the Input Atoms, the Atom Types and the Symmetry.

It may be that the input atom coordinates are not consistent with the symmetry matrices. If the misorientation is caused by a rotation about the z-axis, this can be corrected in the Input Atoms dialog.

Atom coordinates are often simple fractions by symmetry - that is 1/2, 1/3, 1/6, etc. Atoms with such coordinates are in special positions, and may be generated by more than one symmetry matrix. VIBRATZ must apply numerical tests to determine whether atoms are duplicated, and the decimal coordinates entered must be precise enough for these tests. They should be given to the 5th decimal place - e.g. enter 0.66667 for 2/3, not 0.667.

Wrong number of modes in individual species or overall. Be sure that all boxes are checked in the Species dialog before the initial calculation. For crystals, it may be that space group is not oriented correctly with respect to the basis function used for symmetry analysis. This is most likely if the Custom Symmetry option is used, and this is also possible if this symmetry option is used for molecules. See Orientation and/or Space-Group Symmetry for details. This situation is corrected by supplying rotations in the Pre-Calculation Rotation dialog (Input menu). Actually, there is only a small number of possible rotations of this type. Monoclinic crystals (or molecules) must have the unique axis parallel to z (or the mirror plane perpendicular to z), and for crystals in the second setting (unique axis = b) the required rotation is -90 degrees on x. Some tetragonal crystals may need to be rotated 45 degrees on z, and some trigonal and hexagonal crystals may need to be rotated 30 degrees on z.

All frequencies are zero. See Wrong number... above. Be sure Atom Types are correct, i.e. that the numbers given for Input Atoms match those in the Atom Types list.

Some frequencies are zero. Normally this is a result of an insufficient number of force constants. First, check the output to see if the forces specified in the dialogs in the Forces menu have actually been located correctly. If zero internal coordinates are located for a given specification, an error message is printed, but if the number located is incorrect there is no special indication. If you are sure that there is a sufficient number of bonds and 3-atom angles, then it is probably necessary to add out-of-plane 3-atom angle coordinates, Tau forces or Psi forces. Generally, if a given atom is central atom to only one 3-atom angle, it is necessary to have some other type of force involved. It is easy to add 3-atom bending forces perpendicular to the plane of the angle in the Angle Force Data dialog; simply select the Both... radio button. If this solves the problem, a separate force constant can be defined if desired. 180 degree angles should probably always use the Both... option, that is have bending coordinates in two perpendicular directions. In some cases the second coordinate is superfluous, but its presence does no harm.

Calculated frequencies differ from published (calculated) values by a large amount. The choice of units for angle force constants must be consistent. It is easy to change this in the Control Window.

Calculated frequencies differ from published (calculated) values by a small amount. If the choice of units for angle force constants is md/A (not md-A), there may still be small differences because of different choices of bond lengths used to convert unit. If the bond lengths used are specified in a published calculation, it is safer to convert the published force-constant values to md-A. Small differences may also arise from different choices for atomic weights, which are seldom given in published work. If natural isotopes are used and two or more are present in significant abundance, some observed absorption or reflectance bands may show a complete splitting while others may represent an average (see Atom Types).