Skip to content

Gaussian Input File

A Transition State (TS) Search uses an input file that is very similar to an Optimization + Frequency calculation. The major difference lies in the route section, where additional keywords instruct Gaussian to search for a first-order saddle point instead of a minimum on the Potential Energy Surface.

The input file used in this tutorial is shown below.


Complete Input File

%nprocshared=1
%mem=2GB
%chk=formicacid-ts.chk
# opt=(calcfc,ts,noeigen) freq rb3lyp/6-31g geom=connectivity

Title Card Required

0 1
...

Similarity with Previous Calculations

Most sections of the input file are identical to those discussed previously.

These include

  • %nprocshared
  • %mem
  • %chk
  • rb3lyp/6-31g
  • freq
  • geom=connectivity
  • Title section
  • Charge and multiplicity
  • Cartesian coordinates
  • Connectivity table

For detailed explanations of these sections, please refer to

See also

  • Optimization + Frequency → Input File
  • Geometry Optimization → Input File
  • Frequency Calculation → Input File

The only new feature in this calculation is the transition-state optimization keyword.


Route Section

# opt=(calcfc,ts,noeigen) freq rb3lyp/6-31g geom=connectivity

The route section instructs Gaussian to

  1. locate a transition state,
  2. verify it using a frequency calculation.

The most important keyword is

opt=(calcfc,ts,noeigen)

Understanding the opt Options

Unlike a normal optimization,

opt

a Transition State Search requires several additional options.


ts

opt=(ts)

The ts keyword tells Gaussian that the desired stationary point is a transition state rather than an energy minimum.

Instead of minimizing the molecular energy,

Gaussian searches for a structure that satisfies

  • zero energy gradient,
  • exactly one direction of negative curvature.

This corresponds to a first-order saddle point on the Potential Energy Surface.

Without this keyword,

Gaussian would simply optimize the structure to the nearest local minimum.


calcfc

opt=(calcfc)

The calcfc keyword requests that Gaussian calculate the force constant matrix (Hessian) at the very beginning of the optimization.

The initial Hessian provides

  • information about the local curvature,
  • improved optimization directions,
  • faster convergence.

Although calculating the Hessian increases the cost of the first optimization step,

it usually makes Transition State searches significantly more reliable.

This keyword is strongly recommended for most TS optimizations.


noeigen

opt=(noeigen)

During a Transition State Search,

Gaussian normally performs an eigenvalue analysis of the initial Hessian.

For some difficult systems,

this analysis may fail or lead to poor optimization steps.

The noeigen keyword tells Gaussian to skip this initial eigenvalue analysis.

This option is often useful when

  • the initial geometry is already a good approximation to the transition state,
  • convergence problems occur,
  • the Hessian contains numerical instabilities.

Although optional,

it is commonly used in practical transition-state calculations.


Frequency Verification

The keyword

freq

appears immediately after the optimization keywords.

Once the transition state has been located,

Gaussian automatically performs a Frequency Calculation.

This step is essential because it verifies the nature of the optimized structure.

A valid transition state must exhibit

  • one and only one imaginary vibrational frequency.

If

  • zero imaginary frequencies are found,

the structure is a local minimum.

If

  • two or more imaginary frequencies are present,

the optimization has converged to a higher-order saddle point rather than the desired transition state.


Workflow

The calculation proceeds as

Read Initial Guess
Calculate Initial Hessian
Transition State Optimization
First-Order Saddle Point
Frequency Calculation
Verify One Imaginary Frequency

Summary

The input file for a Transition State Search differs from a standard Optimization + Frequency calculation mainly through the use of the opt=(calcfc,ts,noeigen) keyword. The ts option instructs Gaussian to locate a first-order saddle point, calcfc computes the initial Hessian to improve convergence, and noeigen suppresses the initial eigenvalue analysis when appropriate. After the optimization converges, the freq keyword automatically performs a vibrational analysis to confirm that the optimized structure possesses exactly one imaginary frequency, providing definitive evidence that a true transition state has been located.