CCL:G: Reproducing the Gaussian reported frequencies

 Sent to CCL by: "Aron Rogier Vandeputte" [wachtebeekse%x%hotmail.com]
 To obtain SCT-factors for a specific reaction, I need the frequencies and
 eigenvectors at different points along the reaction path. To do this, I
 performed an IRC calculation with the calcall option. In this way I got the
 second derivative matrix at every point. I wrote the matrix (after I made it
 symmetric) to Maple where I diagonalized it. The eigenvalues correspond to those
 reported by Gaussian, but when I try to convert them to frequencies something
 goes wrong.
 Gaussian output:
 ----------------
 All quantities printed in internal units (Hartrees-Bohrs-Radians)
 The second derivative matrix:
                           cc2       hc3      hcc3       hc4      hcc4
            cc2          0.05230
            hc3         -0.00001   0.34915
            hcc3        -0.00775  -0.00970   0.11675
            hc4         -0.00001   0.00088  -0.00807   0.34923
            hcc4        -0.00775  -0.00807   0.02445  -0.00967   0.11679
            dih4         0.00000   0.00000   0.00000  -0.01504   0.04380
            hc5         -0.00001   0.00088  -0.00807   0.00089  -0.00807
            hcc5        -0.00775  -0.00313   0.06337  -0.00313   0.06332
            ....
 Maple worksheet:
 ----------------
 hartree:=4.35975E-18; Joule/Hartree
 bohr:=0.52917725E-10: meter/bohr
 amu:=1.660538E-27:    kg/amu
 hbar:=1.0545E-34:
 for j from 1 to 21 do
     if (Re(v[j])>0) then      v[j] = j_th eigenvalue
     FREQUENCY[j]:=evalf(Re(1/2/Pi*sqrt(v[j]*hartree/bohr**2/amu)/100/3E8));
     fi;
 od:
 The strange thing is that there is no correlation between the frequencies that I
 found and those produced by Gaussian. So, the fault can't be only caused by
 wrong conversions. I wonder what I'm doing wrong: isn't this the correct matrix
 for frequency calculation and how can I get the right one?
 Thanks in advance,
 Aron
 wachtebeekse%%hotmail.com