The Reaction Coordinate Approximation and Its Limitations¶
A relaxed scan minimizes the energy over every coordinate except the one being scanned, it follows a path constrained to move along a single, hand-chosen coordinate \(q\). The rigorous reaction path, found by an Intrinsic Reaction Coordinate (IRC) calculation, instead follows the steepest-descent direction in the full mass-weighted coordinate space starting from a true transition state, with no coordinate singled out in advance.
The two agree closely whenever \(q\) happens to be an excellent proxy for the true reaction coordinate, a simple bond-length scan of a proton mid-transfer, as in this tutorial, is often a good approximation, since the O–H distance dominates the reaction's progress. They can diverge sharply, however, when:
- the true reaction path curves significantly away from the fixed coordinate \(q\) in the space of the other coordinates,
- the electronic structure changes character partway along the scan (for example, an orbital crossing or a change in spin state), producing a discontinuous jump in the energy curve rather than a smooth one,
- the reaction genuinely involves simultaneous, coupled motion along more than one coordinate that no single scan variable can capture.
For this reason, the highest-energy point of a relaxed scan is best treated as an approximate transition-state guess, a good starting geometry to hand to a full Transition State Search, rather than as a rigorous barrier height in its own right.