Hessian does not have the desired local structure

I was trying a search at B3LYP/^-31G* for a TS directly from the reactant using an appropriate frequency (3rd) in N2CO, diazirionone (three-membered ring, C2V). The job failed with the following error (last portion of the output):

Attempting to generate delocalized internal coordinates

Transforming Cartesian Hessian to Internal Coordinates
Hessian Transformation does not Include Derivative of B-matrix

3 Hessian modes will be used to form the next step
Hessian Eigenvalues:
0.307752 0.409839 0.946956
WARNING Hessian does not have the desired local structure
WARNING Hessian does not have the desired local structure

Transition state search - taking P-RFO step

Q-Chem fatal error occurred in module progman/main.C, line 185:

Problem running Q-Chem: mode != 0

Thank you for your advice in advance!

Sincerely,
Anatoli Korkin

When in doubt with a TS search, start by computing the exact Hessian (JOBTYPE=FREQ) and then reading it into the subsequent JOBTYPE=TS job (GEOM_OPT_HESSIAN=READ).

I use IQmol to build q-chem input for the purpose of using it as a educational tool. I did calculated frequencies in the first job and then requested a TS search following the frequency mode 3 (the one which should break the three-membered ring. Here is my input:


$molecule
0 1
O 0.0000000 0.0000000 1.4622351
C 0.0000000 0.0000000 0.2698952
N 0.6557750 0.0000000 -0.9512323
N -0.6557750 0.0000000 -0.9512323
$end

$rem
BASIS = 6-31G*
GUI = 2
JOB_TYPE = Frequency
METHOD = B3LYP
SCF_CONVERGENCE = 8
$end

@@@

$molecule
read
$end

$rem
BASIS = 6-31G*
GEOM_OPT_MODE = 3
GUI = 2
JOB_TYPE = TS
METHOD = B3LYP
SCF_CONVERGENCE = 8
$end

@@@

$molecule
read
$end

$rem
BASIS = 6-31G*
GUI = 2
JOB_TYPE = Frequency
METHOD = B3LYP
SCF_CONVERGENCE = 8
$end