skill Proprietary (Paramus) — guidance only; execution performed by Paramus tools under their own license gating.
Predict and characterize properties of a coordination complex (geometry, spin state, ligand-field splitting, stability trends) by building the complex, assigning oxidation/spin states, and computing electronic descriptors with Paramus QM/ML tools.When to use: a user asks about a metal complex's geometry, spin state, or relative stability.Multi-tool WORKFLOW with coordination-chemistry judgment.

Paramus — Coordination Complex Properties

Overview

Build a metal complex, assign oxidation/spin state, and compute geometry and electronic descriptors. The value is the assembled coordination-chemistry picture, not one calculation.

Guidance vs execution: geometries and electronic properties come from tested Paramus tools. Do not guess spin states or splittings from memory.

When to Use

  • Characterizing a coordination complex's geometry/spin/electronic structure
  • Comparing ligand effects or relative complex stability Do not use for: extended-solid band structure (use the semiconductor/bandgap screen).

Workflow

Metal + ligands + oxidation state (or "determine")
  ↓ 1. Build complex & validate     → coordination geometry enumeration
  ↓ 2. Assign spin/oxidation states  → candidate states to evaluate
  ↓ 3. Optimize / single point       → Paramus QM engine per state
  ↓ 4. Electronic descriptors        → splitting, HOMO-LUMO, charges
  ↓ 5. Report                        → preferred geometry/spin + descriptors + provenance

Procedure

  1. Discover tools via search/get_schema ("geometry optimization", "spin state", "QM single point").
  2. Build candidate geometries and enumerate plausible spin states.
  3. Compute energies per state with provenance; identify the ground state.
  4. Extract ligand-field splitting, frontier orbitals, and charges.
  5. Report the preferred geometry/spin state and descriptors, with confidence and method.

Domain Judgment

  • Spin-state ordering is method-sensitive (esp. DFT functional) — report the method and caveat it.
  • Compare states on the same method/basis, or the ordering is meaningless.
  • Note when multiple states are near-degenerate (spin-crossover candidates).

Fallbacks

  • QM engine unavailable → provide geometry/ligand-field qualitative analysis, clearly scoped.
  • Endpoint unreachable → stop and report; no local electronic-structure estimates.