Skip to content

2.2 H₂ binding: an energy difference needs a reference

These are unexecuted teaching inputs and starting settings to test. Original figures are schematics, not computed results. Use licensed VASP and PAW data, replace every placeholder, record the executable version and validate convergence.

2.2.1 Model, units and provenance

Use eV for energy, Å for length and eV/Å for force; 1 kbar = 0.1 GPa. State normalization per atom, molecule, primitive cell or simulation cell. Record PAW identifiers, release, ZVAL, ENMAX and permitted hashes; never redistribute POTCAR. SCF convergence addresses the chosen electronic problem; convergence of the target property requires separate tests.

 H₂ binding: an energy difference needs a reference — original conceptual schematic

Original schematic. Curves explain concepts; blank data areas await verified learner results. No calculation is claimed.

\[ D_e=2E(\mathrm H)-E(\mathrm H_2) \]

2.2.2 Worked case: procedure, interpretation and checks

Goal and intuition

A negative total energy by itself does not prove a bond exists. Bonding is a statement about a defined reaction. For the electronic dissociation process H₂ → 2H, define Dₑ = 2E(H) − E(H₂ at its minimum). A positive Dₑ means the molecule lies below those separated atomic references within the chosen method. Alternatively define binding energy Ebind = E(H₂) − 2E(H); then favorable binding is negative. Both conventions are valid; mixing them is not.

Start with a bond-length scan so that the meaning of a minimum is visible before using an optimizer. A graph of energy versus H–H distance connects forces, equilibrium geometry and dissociation. This case also demonstrates why a chemically inappropriate atomic spin reference can contaminate a perfectly converged molecular result.

Workflow

  1. Reuse the validated H potential, cutoff and large box from 2.1 A spin-polarized isolated atom. For the near-equilibrium molecule use a non-spin-polarized closed-shell model as the introductory branch.

Unexecuted input changes

For a 15 Å cubic cell with a trial 0.75 Å bond along z, use Cartesian coordinates below. Substitute the actual converged box length when it differs from 15 Å.

Reading the calculation

Inspect forces on both atoms. On the stretched side of a bound minimum the force should tend to shorten the bond; on the compressed side it should tend to lengthen it. Verify signs using coordinates, not just magnitudes. The two atomic forces should approximately balance in an isolated neutral molecule, up to numerical error. Plot energies relative to the lowest computed point for visual clarity, but retain absolute raw energies for the reaction cycle. A smooth curve is necessary evidence of sensible sampling, not proof of an accurate functional.

Convergence and scientific limits

Test Dₑ as a combined quantity. If molecule and atom errors cancel, the difference may converge faster than either absolute energy, but demonstrate this rather than assume it. Dₑ excludes vibrational zero-point energy; a dissociation energy referenced to the vibrational ground state requires the appropriate zero-point correction. Finite-temperature enthalpies or free energies need additional thermal and standard-state terms. Do not call the raw electronic Dₑ a room-temperature bond free energy.

At large H–H separation, a restricted closed-shell description develops serious static-correlation limitations. An unrestricted broken-symmetry calculation may lower the energy, but is not automatically an exact spin-pure singlet. Use the far-dissociation region to discuss model failure, not to silently splice unrelated branches into a single “accurate” curve.

Exercise and original figure

Submit a reaction-energy ledger containing species, number of units, spin treatment, cell size, PAW identifier, energy label and coefficient. Then explain how replacing the spin-polarized H reference with a nonmagnetic one changes the interpretation, even if the H₂ input is untouched.

2.2.3 Unexecuted inputs and analysis scaffolds

These are unexecuted teaching inputs and starting settings to test. Original figures are schematics, not computed results. Use licensed VASP and PAW data, replace every placeholder, record the executable version and validate convergence.

2.2.3.1 Input block 1

H2 trial geometry
1.0
15.0 0.0 0.0
0.0 15.0 0.0
0.0 0.0 15.0
H
2
Cartesian
7.5 7.5 7.125
7.5 7.5 7.875

2.2.3.2 Input block 2

# Changes from atom to near-equilibrium H2 scan
ISPIN = 1
# Remove NUPDOWN and MAGMOM from the atomic template
ISMEAR = 0
SIGMA = 0.02
IBRION = -1
NSW = 0
# Keep the same explicit ENCUT, PREC, EDIFF, LREAL and Gamma mesh

2.2.5 Technical sources