All real runs

The outputs below are not illustrative — they come straight from ORCA 6.1.1. Running one is a single line, orca water.inp > water.out. Clicking a file opens it as text or downloads it.

Worked examples — input + real output

CalculationKey resultFiles
Single point energy
water · B3LYP/def2-SVP
FINAL SPE −76.321274 Eh input · output
Opt + frequencies
water · with thermochemistry
Gibbs −76.422869 Eh · 0 imaginary modes input · output
DLPNO-CCSD(T)
water · cc-pVTZ
−76.331959 Eh input · output
Solvation
water · SMD(water) single point
−76.332972 Eh input · output
CASSCF(2,2)
H2 · stretched bond (multireference)
E(CAS) −1.056125 Eh · N(occ) 1.81/0.19 input · output
Fragment merge → CASSCF(5,5)
[Mn(H2O)]2+ · high-spin d5
N(occ) 1.00×5 (merge places d5) · −1225.0207 Eh (default-guess run) input · merge log

The dissociation_{rhf,cas}.dat files under 05-casscf-h2/ are the real RHF and CASSCF curves scanned over H–H = 0.4–3.0 Å (the source for h2-dissociation.png). RHF diverges at dissociation, CASSCF stays well-behaved — the classic single-reference breakdown.

Fragment-merge example — 06-fragment-merge/

A real run of the split-and-merge procedure: solve the metal and the ligand separately, then merge their orbitals with orca_mergefrag. The full walkthrough and the gotchas are in Chapter 13, the “fragment merge” section.

FileRole
fragMn.inpMetal fragment — Mn2+ high-spin (sextet)
fragH2O.inpLigand fragment — H2O (UHF, to match HFTyp)
mergefrag.logReal merge output — Dim 31+24=55, NEl 23+10=33
super-cas-merged.inpCASSCF(5,5) reading the merged guess
cas-default.inpReference — CASSCF from the default guess
cas-default.outConverged result — E −1225.0207 Eh, N(occ) 1.0×5

Templates — ready-to-edit skeletons

All verified to run on ORCA 6.1.1. Swap mol.xyz for your own coordinate file.

FileUse
single-point.inpSingle point energy
opt-freq.inpOptimisation + frequencies + thermochemistry
solvation-smd.inpSMD solvation single point
dlpno-composite-step1.inpDFT opt+freq (D4) → writes step1.xyz
dlpno-composite-step2.inpDLPNO-CCSD(T)/CBS single point (AutoAux, reads step1.xyz)
Two traps found while testing (folded into the templates)
  • DLPNO-CCSD(T) Extrapolate(...) dies with an “auxiliary basis required” error unless you add AutoAux.
  • Bundling a DFT opt (D4) and CCSD(T) in one input via $new_job lets D4 carry into the CC step and trips a “dispersion + correlated method” abort → split the steps into separate files.

ORCA English User Guide · Based on the ORCA 6.0.0 manual · This guide is unofficial.