VSEPR Geometry Predictor

Inspectable records

Ledger state

The current ledger marks this instrument verified clean within the source-bound verification corpus.

Area

Chemistry & Materials

Tranche

tranche-6

Evidence ID

t6-135

Ledger rows

1

No ledger note is attached.

Review findings

The core steric-number concept is standard, but the instrument needs an explicit, auditable lookup table and an unambiguous definition of bonding domain. It must also support multiple idealized angles or clearly define which angle is returned. Because these requirements are unspecified, the instrument cannot yet be verified as a complete geometry-and-angle predictor.

Formula and scope findings

• The steric-number formula is standard only when “bonding domains” means bonded-atom electron domains; a single, double, or triple bond each counts as one domain. The input description instead says “bonding pairs,” which is ambiguous and could incorrectly count multiple bonds as multiple pairs.
• Steric number alone is insufficient to determine every molecular geometry and bond angle unless the lookup table also specifies lone-pair placement and domain arrangement, especially for steric numbers 5 and 6.
• Many cases do not have one bond angle: for example, square-planar species have 90° and 180° angles, while trigonal-bipyramidal derivatives can have both approximately 90° and 120° angles.
• The specification does not define the supported steric-number range or the complete geometry lookup table, so “etc.” is not an auditable algorithm.
• The output must distinguish idealized VSEPR angles from measured molecular angles; lone-pair repulsion commonly makes actual angles differ from idealized values.

Unit assumptions

• Bond angles are reported in degrees, not radians.
• Bonding-domain and lone-pair counts are dimensionless nonnegative integers.
• Each bonded atom contributes one bonding domain, regardless of single, double, or triple bond multiplicity.
• The listed angles are idealized VSEPR angles; they are not experimental bond angles unless explicitly labeled otherwise.

Boundary hardening

BREAK_CLAIMED · SELF_FLAGGED

The specification exposes invalid-input hazards at zero and at overflowing magnitudes, where a robust implementation must reject the input rather than perform an unchecked lookup. It also has a material convention hazard because VSEPR counts bonding domains, whereas the label 'bonding pairs' can be interpreted as individual shared electron pairs.

Errata and corrections

No applied erratum is attached to this unique-name record.

Inspectable example

4 review vectors are recorded. One source-bound sample is shown below.

{
  "inputs": {
    "central_atom": "C",
    "ligands": "O=C=O",
    "bonding_domains": 2,
    "lone_pairs": 0
  },
  "expected_outputs": {
    "steric_number": 2,
    "electron_geometry": "linear",
    "molecular_geometry": "linear",
    "idealized_bond_angles_deg": [
      180
    ]
  },
  "working": "Step 1: Count bonding domains: the two C–O directions give 2 domains; each C=O double bond still counts as one domain. Step 2: Add lone pairs: steric number = 2 + 0 = 2. Step 3: Steric number 2 maps to linear electron geometry. With no lone pairs, molecular geometry is also linear. Step 4: The ideal linear angle is 180°."
}

Related workflows

No source-defined multi-instrument workflow membership is recorded.

Sources and evidence

These pointers bind the page to evidence. They do not grant authority or collapse verification into deployment.

• VERIFIED_LEDGER.json · VERIFIED_CLEAN
• evidence-tranche6/tranche6-results.json
• BOUNDARY_HAZARD_CATALOG.json