Stationary Points & Optimization
Finding maxima, minima and transition states
Finding peaks and valleys
Set the derivative to zero and let the second derivative decide
A curve is perfectly flat at its maxima and minima — the tangent line is horizontal and the slope is zero. That is why we find them by solving . To tell which type of stationary point we have, we check the second derivative:
If the second-derivative test is inconclusive — you need to check the sign of on either side (the first-derivative test). A point where and changes sign is a point of inflection — neither a peak nor a valley.
This procedure is optimization — finding the best (lowest-energy, fastest, most probable) value of a quantity. Nearly every time a chemist writes “minimise the energy” or “most probable speed”, they are setting a derivative to zero.
The Lennard-Jones potential provides an important worked example of the complete optimisation workflow. Setting gives the equilibrium separation — the van der Waals contact distance. The force on the two atoms is , and at this force is exactly zero. The second derivative confirms a true minimum (restoring force), and its value is proportional to the harmonic vibrational frequency of the dimer — linking optimisation directly to spectroscopy.
A subtler point: inflection points on kinetic curves carry chemical information. An autocatalytic reaction with rate produces a sigmoidal concentration–time curve for . The inflection point — where — marks the moment the rate of reaction is increasing most rapidly and corresponds to the maximum rate of rate change. After the inflection, substrate depletion decelerates the acceleration. Identifying this point requires the second-derivative test (set , check sign change), not just looking for .
- Not classifying the stationary point. Finding only locates the point — you must then use the second-derivative test (or the first-derivative sign test) to decide whether it is a maximum, minimum, or inflection point.
- Missing endpoint extrema. On a closed interval, the overall extremum might occur at the boundary or even if interior stationary points exist. Always evaluate at both endpoints and compare.
- Assuming f″ = 0 means an inflection point. If you must check that actually changes sign at — if it doesn't, the second-derivative test is inconclusive and the point could still be a maximum or minimum (e.g. at ).
- Confusing the force with the potential minimum. In a potential-energy well, at the equilibrium position. This is not where is zero, but where the slope of is zero.
Stationary-point finder
Adjust the parameter — the red dot is the maximum (transition state analogue) and the green dot the minimum (stable intermediate).
A reaction-coordinate energy model gives kJ mol⁻¹. Find all stationary points and classify each.
The Maxwell–Boltzmann speed distribution for a gas of molar mass at temperature is . Find the most-probable speed by differentiating and setting equal to zero.
Worked examples
Optimization across kinetics, molecular structure and thermodynamics.
The Lennard-Jones potential is . Find the equilibrium separation where is a minimum.
In consecutive first-order reactions with and , the intermediate B reaches its maximum at time . Evaluate this time.
For a simple reaction , the Gibbs free energy is near equilibrium, where depends on . The minimum is where . Show that this condition gives .
An autocatalytic reaction gives the logistic concentration profile where is the maximum concentration, is a growth rate constant, and is a reference time. Find the inflection point and state what it represents physically.
The Lennard-Jones pair potential is . The pair force is . Show that at and confirm the second derivative confirms a minimum of there.
In (consecutive, first order) with , the intermediate concentration is:
(a) Find the time at which is maximised.
(b) Show that .
(c) Evaluate both for , .
Check yourself
Four quick questions on stationary points and their chemical meaning.
A stationary point of f(x) occurs where: