I did the calculation previously where I optimised the reactant and product using b3lyp/6-31+g(d) followed by a refinement using b3lyp/6-311++g(d) and then ran a qst2 calculation at b3lyp/6-31+g(d) which worked, however I am required to use the B3LYP/aug-cc-pVTZ for optimisations, so I first optimsed the reactant and product using b3lyp/6-31+g(d) and then refined at B3LYP/aug-cc-pVTZ and when I ran the qst2 using these optimised structures I am getting the error: Old curvilinear step not converged, using linear step:
SCX= 2.13D+01 DXMaxT= 6.95-310 SCLim= 0.00D+00 Fact= 0.00D+00
RedCar/ORedCr failed for GTrans.
I dont think there is an issue with atoms not matching since I was sure they were and the qst2 worked using a different basis set. My current input is # opt=(calcfc,tight,qst2) freq b3lyp/6-31+g(d) nosymm scf=(qc,xqc,maxcyc=1024) however I also tried variants without xqc and variants with geom=connectivty and the error above still persists
I attached the log file if anyone may be willing to take a look, I would greatly appreciate help with this.
Hi, please check out the common errors and solutions.
https://wongzit.github.io/gaussian-common-errors-and-solutions/
Hallo, Joshua!
Your problem originates in geometries (sometimes small change in structure causes big problems). I have this type problem usually with rotation barriers.
In such case I use the 'qst3' keyword instead of the 'qst2' one. You can use the b3lyp/6-31g+(d) barrier structure as the 3rd geometry (=approximate barrier estimation).
Best wishes
mb
Some transition states are not that easy to find and qst2 just won't work, no matter what you do. Try to first freeze some coordinates to get a reasonable guess for a transition state (do frequency calculations to see of there is one vibration with imaginary frequency that resembles the one you're looking for) and then use it as a guess for qst3. That obviously requires some chemical intuition to know what you're looking for.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis