Metal Carbonyl Bonding and Structure Explained Simply

Carbon monoxide is infamous as a poison, yet bound to metals it becomes one of the most important ligands in chemistry. Metal carbonyls are compounds where CO molecules attach directly to a transition metal through their carbon atom, such as Ni(CO)4 (a volatile liquid), Fe(CO)5 (a yellow liquid) and Cr(CO)6 (a white solid). The bonding is special: CO donates its carbon lone pair to the metal and accepts electron density back from the metal into its empty pi orbitals, a two-way exchange called synergic or backbonding.

Carbonyls matter because they are the workhorses of industrial catalysis. Hydroformylation turns alkenes into aldehydes on cobalt carbonyl catalysts at massive scale, acetic acid is manufactured via rhodium and iridium carbonyl cycles (the Monsanto and Cativa processes), and the Mond process purifies nickel by shuttling it as volatile Ni(CO)4. Understanding their bonding unlocks the whole of organometallic catalysis.

Key Concepts

• Sigma donation: The carbon lone pair (HOMO of CO) donates into an empty metal orbital, forming the M-C sigma bond. CO always binds through carbon, never oxygen, because the lone pair lives mainly on carbon.
• Pi backbonding: Filled metal d orbitals overlap with CO's empty pi* antibonding orbitals, pushing electron density back onto the ligand. The two effects strengthen each other: more sigma donation makes the metal richer, which increases backdonation, which makes the metal accept more sigma donation.
• Effect on the C-O bond: Backdonation populates the CO pi* orbital, weakening the C-O bond. Free CO stretches at 2143 cm-1 in the infrared, while terminal carbonyls appear around 2000-2100 cm-1 and bridging carbonyls (bound to two metals) drop to 1700-1860 cm-1.
• Bonding classification: CO is a strong-field ligand high in the spectrochemical series, which is why carbonyl complexes are typically low spin and often obey the 18-electron rule.
• Structures: Ni(CO)4 is tetrahedral, Fe(CO)5 is trigonal bipyramidal, Cr(CO)6 is octahedral. Mn2(CO)10 has a direct Mn-Mn bond with no bridging COs, while Fe2(CO)9 and Co2(CO)8 contain bridging carbonyls shared between two metals.
• Bridging vs terminal: A bridging CO (mu-CO) donates 2 electrons total shared between two metals (1 each in counting), and its lowered IR stretch is the standard diagnostic for the bridging mode.
• Preparation: Direct combination works for the most reactive metals: Ni + 4 CO gives Ni(CO)4 at just 80 C, the basis of the Mond refining process.

Formulas

• Electron counting (neutral method): Total valence electrons = group number of metal + 2 x (number of CO ligands) + adjustments for other ligands, charges and metal-metal bonds. Each terminal CO donates 2 electrons; each bridging CO donates 1 electron to each metal it bridges.
• Fe(CO)5 check: Iron is group 8, five CO ligands give 10 electrons: 8 + 10 = 18. The 18-electron rule is satisfied, consistent with its stability.
• Mn2(CO)10 check: Each manganese is group 7 with five CO ligands (10 electrons) plus one Mn-Mn bond (1 electron): 7 + 10 + 1 = 18 per metal. The metal-metal bond supplies the missing electron.
• IR diagnostic rule: nu(CO) above about 2000 cm-1 indicates terminal CO; 1700-1860 cm-1 indicates bridging CO; stronger backbonding always shifts the stretch to lower wavenumber.

Worked Example

Question: Predict the structure type and verify the electron count for Cr(CO)6, and explain what its IR spectrum shows about backbonding.

1. Count electrons: Chromium is group 6; six CO ligands donate 12 electrons: 6 + 12 = 18. The 18-electron rule holds.
2. Predict geometry: Six identical ligands around a d6 metal give an octahedral complex with all CO ligands terminal and equivalent.
3. Read the IR: Cr(CO)6 shows its C-O stretch near 2000 cm-1, well below free CO at 2143 cm-1. The shift proves pi backdonation is weakening the C-O bond.
4. Connect to stability: The strong backbonding both strengthens the Cr-C bonds and satisfies the 18-electron count, which is why Cr(CO)6 is a stable white solid that sublimes unchanged, unlike the fragile carbonyls of metals that cannot reach 18 electrons.

Common Misconceptions

• CO binds through oxygen because oxygen is more electronegative: It binds through carbon. The HOMO lone pair of CO is concentrated on carbon, and the pi* acceptor orbitals also have larger carbon coefficients, so the carbon end does both jobs.
• Backbonding weakens the metal-carbon bond: The opposite. Backdonation adds pi character to the M-C bond, strengthening and shortening it, while the bond that weakens is C-O, as the IR shift proves.
• All carbonyls are extremely toxic gases: Ni(CO)4 and Fe(CO)5 are indeed volatile and highly toxic, but Cr(CO)6, Mo(CO)6 and W(CO)6 are stable solids. Toxicity tracks volatility and lability, not the mere presence of CO ligands.

Real-World Applications

• Acetic acid manufacture: The Monsanto (rhodium) and Cativa (iridium) processes carbonylate methanol using anionic carbonyl catalysts, producing millions of tonnes of acetic acid yearly.
• Hydroformylation (oxo process): Cobalt and rhodium carbonyl catalysts add CO and H2 to alkenes to make aldehydes, the gateway to plasticizers, detergents and solvents.
• Nickel refining: The Mond process converts impure nickel to volatile Ni(CO)4 at 80 C and decomposes it at 230 C to deposit 99.99 percent pure nickel.
• Precursors in materials science: Volatile carbonyls deposit pure metal films by chemical vapour deposition for electronics, and Fe(CO)5 was once an antiknock petrol additive.

Metal carbonyls bind CO through carbon with a two-way synergic interaction: sigma donation from the ligand and pi backdonation from the metal, each strengthening the other. This backbonding weakens the C-O bond (visible in the IR spectrum), enforces low-spin 18-electron configurations, and creates the stable yet reactive complexes behind hydroformylation, acetic acid synthesis and nickel refining. Master the electron count and the IR diagnostic, and carbonyl chemistry becomes predictable.