Skip to content

2.4 Optimize a water molecule and read all four criteria

These are unexecuted teaching inputs and starting models. Original diagrams are schematics, not calculated results. Validate version-specific syntax, licensed or authorized data, numerical convergence and the scientific model before using this workflow.

2.4.1 Model, units and provenance

Geometry in Å; electronic energy in hartree; vibrational wavenumbers in cm⁻¹. Check each printed field and keep thermal and standard-state terms distinct.

Shared inputs, conventions and evidence

Original schematic: Optimize a water molecule and read all four criteria. No numerical results are claimed.
Original schematic: Optimize a water molecule and read all four criteria. No numerical results are claimed.

2.4.2 Unexecuted inputs and explicit deltas

Use the accompanying instructions to identify the parent calculation and placement of every delta; a snippet is not automatically a standalone input. Preserve all blank-line and file-provenance requirements.

2.4.2.1 Input block 1

%chk=g08_water_opt.chk
%mem=2GB
%nprocshared=2
#p B3LYP/6-31G(d) Opt=(Tight,CalcFC) SCF=Tight Int=UltraFine

Water optimization from an intentionally distorted start

0 1
O  0.000000  0.000000  0.000000
H  0.820000  0.000000  0.650000
H -0.720000  0.000000  0.540000

2.4.3 Worked investigation

2.4.3.1 Intuition and prerequisites

Geometry optimization uses energies and gradients to move nuclei toward a stationary point. A low final energy alone is insufficient: residual forces and step sizes must also satisfy the stopping rules. Start from a deliberately distorted water structure so that optimization is visible, but keep the geometry chemically reasonable. The method and numerical grid match later frequency work. Prerequisites are the SCF and fixed-geometry lessons; this case creates g08_water_opt.chk, which is a useful course baseline.

See input block 1 above.

2.4.3.2 Workflow

  1. Visualize the input and measure its two O–H distances and H–O–H angle. Save those measurements as starting values, not predicted results.
  2. Run the optimization and separate outer geometry steps from inner SCF iterations. CalcFC requests an initial Hessian to guide the optimization.
  3. Read maximum force, RMS force, maximum displacement, and RMS displacement at the accepted endpoint. Confirm the explicit optimization-completion message.
  4. Extract the final coordinates from the correct final geometry or checkpoint. Compare bond lengths and angle with the starting structure.
  5. Repeat from a second distorted start with a new checkpoint. If both reach the same structure and energy within tolerance, you have evidence for local reproducibility, not proof of a global search.
  6. Reserve minimum certification for a matching frequency calculation in case 10.

2.4.3.3 Interpret check and limits

Gaussian prints force-related criteria in atomic units and displacement criteria in the units stated in the output; do not relabel raw values as eV Å−1. An energy decrease is useful evidence but not the formal stopping test, and every individual step need not lower energy under all optimization algorithms. Normal termination after a limited or constrained job must be interpreted in context. Symmetry or a constraint can hold a structure at a saddle; a small gradient alone cannot distinguish it from a minimum. If a new frequency job uses a different basis, functional, solvent, or grid, the old optimized structure is not guaranteed stationary on the new surface.

Exercise Produce a one-page geometry report with starting and final distances, all convergence criteria, software revision, and a statement of what remains unverified before frequency analysis.

2.4.5 Sources and further reading