g94 compound freq+opt

["Michael Probst" <Michael.Probst { *at * } uibk.ac.at>]

 Dear Gaussian - experts:
 The following input:
 -----------------------------------------------
  #P BLYP/aug-cc-pVTZ SCF=DIRECT OPT=Z-MATRIX FREQ MAXDISK=13000000
  NITRAT:aug-cc-pVTZ
 -1 1
  N1
  O2    1     RONN
  O3    1     RONN      2     AOON
  O4    1     RONN      2     AOON      3     TORS      0
 Variables:
  RONN=1.271
 Constants:
  AOON=120.
  TORS=180.
 ------------------------------------
 seems to work well but upon cloeser looking there is a large energy
 difference between the energy at the end of the optimization
 (-280.470491943) and the energy calculated in the frequency step
 (-280.446822509).
 I would not expect the NO3- ion to behave that way and further, if I
 make a independent frequency calculation with the optimized N-O
 distance (1.280612 Angstroem) I get the same energy as in the
 optimization job (and, of course, different frequencies as in the
 compound job)
 Can somebody tell me the reason for the differences in the job with
 freq and opt together ?
 Thanks a lot !
 Michael
 ---------------------------------------------------------------------------
 Michael Probst                          Internet: michael.probst { *at * }
 uibk.ac.at
 Department of Inorganic Chemistry       Fax:      ++43(512)507 2934
 Innsbruck University                    Tel:      ++43(512)507 5153
 Innrain 52a
 Austria