Finding transition states is crucial to understanding chemical reactivity and predicting reaction rates and selectivity.
Unlike conventional optimizations, which minimize a structure's energy, transition-state optimizations use an eigenvector-following algorithm to locate first-order saddle points (i.e. transition states) on the potential-energy surface. There are many more transition states than ground states, so it's important to start from a good guess before commencing a TS optimization. The most common way to do this is to first carry out a scan along one of the reaction coordinates that's changing in the TS, and then to resubmit the highest-energy structure from this scan as a transition state optimization.
It's important to always run a frequency calculation after locating a transition state, to verify that the transition state located is indeed a true first-order saddle point and that it doesn't have any additional vibrational frequencies. Rowan will automatically display a warning if a subsequent frequency job shows that a transition state is invalid, thus ensuring that users' results are accurate.
With Rowan, it's easy to find transition states! We employ the restricted-step partitioned rational function optimization algorithm through the geomeTRIC library for all levels of theory, so robust transition-state optimizations can be run with any of our underlying engines, even those which don't come with transition-state optimizers. This makes it faster and easier to predict barriers, selectivities, and reactivity. Rowan's web interface enables quick visualization of the key imaginary vibrational mode, so users can quickly see which bonds are breaking or forming:
Rowan also makes it easy to go from scans to transition states: you can resubmit from a scan directly to a transition-state optimization with just a few clicks. The scan doesn't even need to be finished; this can all be done from Rowan's web interface while the scan is still running!
For high-accuracy results, it's possible to run a conformational search on a transition-state structure by restricting the key bonds or angles that move in the transition state. This generates lots of excellent near-TS structures which can then be subjected to additional transition-state searches, vastly speeding up the laborious but important process of finding different conformers of a transition state. Here's one such search:
