I am optimizing a transition state using the following command. It doesn't converge and the geometry moves back to the starting point (reactant).
%chk=M2-TS1.chk
%mem=5GB
%nprocshared=8
# opt=(calcfc,ts,modredundant,noeigen) b3lyp extrabasis
Title Card Required
0 1
C -4.54305200 3.59144000 -2.45311300
C -4.45047500 2.06860700 -2.27111000
C -3.03794400 1.59493300 -2.11696600
N -2.00311400 1.79297600 -3.02743200
C -2.47134300 0.87539200 -1.10855200
C -0.88092400 1.19461400 -2.54348500
N -1.14296700 0.63117600 -1.38062700
C -5.79852600 -0.92471200 1.86179300
C -4.32195100 -0.88082000 1.42713700
C -3.40027200 -1.95699400 2.01914000
C -1.92172400 -1.82124500 1.61737300
O -1.67446400 -0.96468100 0.69902100
O -1.05750300 -2.52561300 2.19101300
C 4.73054800 3.12710100 -1.98470300
C 3.65752000 2.11839900 -1.52714000
C 3.62420000 0.76579500 -2.25966200
C 2.72038000 -0.25315100 -1.55260300
O 1.49516000 0.05933600 -1.40131100
O 3.26607100 -1.33123300 -1.17659100
C 1.03597400 5.09029900 3.41895800
C 2.13658500 4.18879900 2.81093400
C 1.59207600 2.85644200 2.38808800
N 1.39702800 1.80376900 3.27164900
C 1.08978900 2.38757600 1.19739000
C 0.81945600 0.77559700 2.58315800
N 0.62210700 1.09782600 1.32469000
Ni -0.06767800 -0.56119000 -0.29807500
H -4.99821400 1.74396600 -1.37986500
H -4.92929400 1.56950200 -3.12652600
H -2.91847500 0.50167900 -0.20833400
H 0.07726300 1.17805700 -3.03846400
H -5.57712700 3.92733100 -2.57987100
H -3.91339000 0.10063100 1.70033800
H -4.27471200 -0.94360700 0.33487900
H -3.44186800 -1.95459700 3.11593600
H -3.72072800 -2.96229300 1.71073500
H -6.35029300 -1.75730500 1.41166100
H 3.80251600 1.93410600 -0.45467000
H 2.66823600 2.58252700 -1.61520400
H 4.62645700 0.33041900 -2.32107700
H 3.25958300 0.90627400 -3.28705600
H 5.74632700 2.78187100 -1.76246300
H 2.95318400 4.05013500 3.53367000
H 2.58193000 4.67928600 1.93741800
H 1.05377600 2.90203600 0.24697500
H 0.58086200 -0.17605700 3.03805100
H 1.44494200 6.03670000 3.79075300
H -2.06078200 2.28476300 -3.90877500
H 1.67703300 1.78116300 4.24349800
H 4.67250600 3.37063400 -3.05584100
H 4.53212900 4.05132600 -1.42213500
H -5.88192900 -1.00408800 2.95576600
H -6.28192300 0.02482100 1.58839800
H 0.30755500 5.30052200 2.62195400
H 0.58419500 4.53992900 4.25741200
H -4.09391700 4.10614100 -1.59091500
H -3.98637000 3.81362900 -3.37547400
O -0.29720400 -2.62694500 -1.39712100
C 0.63687900 -3.34153300 -1.02473500
C 1.58340200 -2.81138600 -0.00989200
H 2.45378800 -2.18120000 -0.62182900
O 1.01238000 -1.75888100 0.78900000
H 0.31529300 -2.16089200 1.41793500
C 0.93167900 -4.65606800 -1.71075900
H 0.69477300 -5.49380500 -1.04385800
H 0.32332700 -4.72891300 -2.61573900
H 1.99555800 -4.73853400 -1.95817500
S 2.36875500 -4.07749300 1.07187900
H 3.61212000 -3.47357000 1.13937700
X 53 F
X 8 F
X 52 F
X 57 F
X 1 F
X 56 F
X 55 F
X 20 F
X 54 F
X 51 F
X 14 F
X 50 F
C O H N 0
6-31+g(d)
****
Ni S 0
lanl2dz
****
As you are reading a checkpoint the Opt=RCFC keyword, that reds the FC used to its generation, could be a solution.
As the Gaussian manual said, " Normally it is preferable to pick up the force constants already converted to internal coordinates as described above (CalcFC). However, a frequency calculation occasionally reveals that a molecule needs to distort to lower symmetry. In this case, the computed force constants in terms of the old internal coordinates cannot be used, and instead Opt=RCFC is used to read the Cartesian force constants and transform them"
And the last option, that I could imagine now, if you are using the Berny algorithm, the best could be use Opt=VeryTight. You could try different combinations, tight, verytight, extreme...
Hope it helps, but as always you have to try different convergence options, as there is no a unique and magic way to treat this kind of problems.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry