Module: Optimization (opt)
The Optimization (OPT) module performs geometry optimization to locate a stationary point on the potential energy surface. Starting from an initial molecular structure, the algorithm iteratively updates atomic positions to reduce the forces acting on the atoms until the predefined convergence criteria are satisfied. The optimization process minimizes the forces acting on the atoms, typically using algorithms such as LBFGS (Limited-memory Broyden-Fletcher-Goldfarb–Shanno), RFO (Rational Function Optimization), or SD (Steepest Descent).
I. Input Parameters
The optimization task is invoked with:
#opt(method=lbfgs, maxiter=200, convergence=1e-5)
The above represents the default settings. Adjustable parameters include:
method
Optimization algorithm:
maxiter
Maximum number of optimization steps.
convergence
Force convergence threshold (Hartree).
II. Output File Contents
Three categories of information are printed:
Updated geometry during the optimization
Example:
Iteration: 54
Coordinates
----------------------------------------------------------------------
0 C -0.8650 -0.4669 0.4322
1 C 1.4518 0.3063 -0.0433
...
21 H 2.6583 -0.7351 -1.4826
Final optimization results
The final section reports the optimized energy (in Hartree) and the four convergence metrics: Maximum Force; RMS Force; Maximum Displacement; RMS Displacement.
Each metric includes its value, the corresponding threshold, and a boolean indicator showing whether the criterion is satisfied. If all criteria are met, the program confirms successful convergence and reports the iteration number.
Example:
Energy: -383.655532
Convergence criteria Is converged
Maximum Force: 0.002118 0.002850 Yes
RMS Force: 0.001226 0.001900 Yes
Maximum Displacement: 0.002501 0.003150 Yes
RMS Displacement: 0.002010 0.002100 Yes
LBFGS optimization converged at iteration 54.
Output File Location
Example:
Final optimized structure written to:
/path/to/output/file.xyz