3.3 Keeping thermochemical corrections and standard states consistent
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.
3.3.1 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
3.3.2 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.
3.3.2.1 Input block 1
%oldchk=case10_water_min.chk
%chk=case11_water_298.chk
%mem=4GB
%nprocshared=2
#p B3LYP/6-31G(d) Freq=ReadFC Geom=AllCheck Guess=Read
Temperature=298.15 Pressure=1.0
3.3.2.2 Input block 2
%oldchk=case10_water_min.chk
%chk=case11_water_350.chk
%mem=4GB
%nprocshared=2
#p B3LYP/6-31G(d) Freq=ReadFC Geom=AllCheck Guess=Read
Temperature=350.0 Pressure=1.0
3.3.3 Worked investigation
3.3.3.1 Question and intuition
Why do electronic energy, zero-point-corrected energy, enthalpy, and Gibbs energy differ for the same optimized molecule? Electronic energy is one part of the bookkeeping. Nuclear zero-point motion and temperature-dependent rotational, translational, and vibrational terms add further contributions. Gibbs energy also contains the entropy term. Think of several carefully labeled totals in a ledger: each total contains different entries, so adding two totals can count some entries twice. This case uses the validated water minimum from Case 10 to learn that accounting without introducing a chemical reaction.
Prerequisite: case10_water_min.chk must contain the completed same-level frequency calculation, not merely an SCF wavefunction or an optimization checkpoint with an approximate Hessian. Reusing the computed Hessian changes the statistical-mechanical analysis conditions without recomputing the electronic potential-energy surface. Changing the temperature keyword does not perform molecular dynamics, introduce a liquid, or optimize a finite-temperature free-energy surface.
3.3.3.2 Checkpoint based inputs
Run the two inputs separately. Each copies the verified source into a different output checkpoint. With Geom=AllCheck, omit the title, charge/multiplicity, and coordinate sections entirely; a blank line terminates the route section. The source remains available for future cases.
See input block 1 above.
See input block 2 above.
3.3.3.3 Workflow
- Confirm the source minimum has no imaginary internal modes. Read the mass/isotope information and rotational symmetry number from the original output; these affect thermochemistry even if the electronic energy is unchanged.
- Run both analyses and verify the printed temperature in kelvin and pressure in atmospheres. Do not assume that an input file was parsed correctly simply because its name includes “350K”.
- Transcribe Eelec, the zero-point correction, thermal corrections to energy/enthalpy/Gibbs energy, and the corresponding sums. Retain full precision until taking differences. Most correction and sum lines are in hartree; entropy in the thermochemistry table is commonly cal mol−1 K−1, so read the table header rather than guessing.
- Reconstruct H = Eelec + Hcorr and G = Eelec + Gcorr, then check G = H − TS after converting units. The listed thermal corrections already include zero-point energy; do not add the zero-point correction a second time.
- Compare the two temperatures. With an unchanged geometry and Hessian, changes come from the thermochemical treatment. Record the exact temperature, pressure, isotopes, symmetry convention, and scaling convention beside every reported G value.
3.3.3.4 Standard state audit
Gaussian's ordinary frequency thermochemistry is an ideal-gas RRHO treatment, with defaults of 298.15 K and 1 atm. A solution calculation elsewhere in a project does not automatically replace the translational partition function by a liquid-state model. If the chosen thermodynamic cycle requires converting an ideal-gas 1 atm standard chemical potential to a 1 mol L−1 standard chemical potential, the per-species correction is ΔG° = RT ln(C°RT/p°), with units inside the logarithm made dimensionless. At 298.15 K this is approximately +7.93 kJ mol−1. This number is a standard-state conversion derived from constants, not a Gaussian molecular result.
For a balanced reaction, the net conversion is Δν RT ln(C°RT/p°), where Δν is the sum of product stoichiometric coefficients minus the reactant sum for the species being converted. A one-molecule-to-one-molecule isomerization has cancellation; an association does not. Do not add this correction automatically to a solvation free energy already defined for the target standard states. Write down the gas and solution reference states on both sides of the thermodynamic cycle first. Also distinguish 1 atm from 1 bar if the experimental reference uses the latter.
3.3.3.5 Checks pitfalls and exercise
Acceptance requires successful reconstruction of the reported sums and an explicit standard-state statement. Frequent errors are combining a 350 K enthalpy with a 298 K entropy, mixing calories and joules, calling a thermal correction the total energy, or interpreting the molecular Gibbs energy as a standard Gibbs energy of formation from the elements. Low-frequency torsions can make RRHO entropy unreliable; water avoids most of that complexity but does not make the approximation exact.
EN exercise: Derive the sign of the 1 atm→1 M correction from μ = μ° + RT ln a, and predict its net sign for A+B→AB. Then explain why the same correction cancels for the ammonia inversion used in Cases 14–16. No molecular energy values need to be invented to answer either question.
3.3.4 Related calculations
- 3.2 Proving that an optimized water structure is a minimum
- 3.4 Comparing gas phase and SMD solution consistently