JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Carbonylation & Polymerization

Two of the largest-scale homogeneous catalysts on Earth — the cycles that turn methanol into acetic acid, and the metal sites that knit ethylene and propylene into the plastics we live among. Watch how the elementary steps you already know assemble into industry, and how design tunes rate, selectivity, and even handedness.

The same four steps, two giant industries

By now the elementary steps of a catalytic cycle should feel like old friends. A metal can swallow a small molecule and split it across two new bonds in an oxidative addition; it can slide a ligand from the metal onto a neighbouring group already bound to it in a migratory insertion; it can stitch two fragments back together and spit them out in a reductive elimination; and it can let a beta hydrogen hop back onto the metal in a beta-hydride elimination. This guide does something satisfying with those steps: it shows you that two of the largest-scale chemical processes on the planet are nothing more than these same moves, looped over and over around a single metal atom that is never used up.

The first industry is carbonylation: bolting a carbon monoxide onto an organic molecule to grow its carbon chain by one and cap it with a carbonyl group. Its flagship is the conversion of methanol, CH3OH, plus CO into acetic acid, CH3COOH — the vinegar acid, made not by fermentation for industry but by the millions of tons a year for solvents, plastics, and a hundred downstream chemicals. The second industry is polymerization: clipping small alkene molecules end to end into chains tens of thousands of atoms long. Ethylene, CH2=CH2, becomes polyethylene; propylene, CH3-CH=CH2, becomes polypropylene; together these polyolefins are made on a scale of hundreds of millions of tons a year, the most-produced synthetic materials in human history.

Monsanto and Cativa: methanol becomes acid

The Monsanto process hangs the whole job on a single rhodium atom carried in a flat, square-planar complex, the anion [Rh(CO)2I2]-. Rhodium here is in oxidation state +1 with a d8 count, so it sits at a comfortable sixteen electrons with a vacant site — exactly the coordinatively unsaturated, primed-to-react species the previous rung taught you to expect. The cycle does not act on methanol directly. First an iodide promoter converts the methanol into methyl iodide, CH3I, a far more reactive carbon electrophile, and it is CH3I that the rhodium attacks.

  1. Oxidative addition: the square-planar Rh(I) inserts itself into the C-I bond of CH3I, becoming an octahedral Rh(III) that now carries both a methyl group and an iodide. The metal's oxidation state climbs by two and its electron count rises from sixteen to eighteen.
  2. Migratory insertion: a CO already bound to the rhodium slips between the metal and the methyl, fusing them into an acetyl group, CH3CO-, still anchored to the metal. No atoms enter or leave; the metal simply rearranges what it holds, and a vacant site opens up.
  3. A fresh CO from solution fills that empty site, reloading the complex for the final step.
  4. Reductive elimination: the metal releases an acetyl iodide, CH3COI, dropping back to the starting Rh(I). Water then hydrolyses the acetyl iodide into acetic acid and regenerates the iodide promoter, closing the loop with no net consumption of rhodium or iodine.

Decades of measurement showed that the slowest, rate-determining step is the very first one — the oxidative addition of CH3I to rhodium. That single fact is the lever the industry pulled. The Cativa process, which has now largely replaced Monsanto, swaps rhodium for its heavier neighbour iridium. Iridium adds CH3I much faster, so the bottleneck moves to a later step; with the right promoters (ruthenium or rhenium species and far less iodide), the iridium cycle runs faster, tolerates less water, wastes less methanol to side reactions, and forms fewer byproducts. Same four steps, same logic — a different metal chosen because it shifts which step is rate-determining. That is catalyst design in one clean comparison.

Ziegler-Natta: stitching ethylene and propylene

Polymerization runs on an even shorter loop — really just one elementary step, repeated. The classic Ziegler-Natta catalyst is born when a titanium chloride is treated with an aluminium alkyl such as triethylaluminium. The aluminium hands the titanium an alkyl group, putting a titanium-carbon bond and a vacant coordination site side by side on the metal. That pairing — a growing chain and an empty seat right next to it — is the whole engine.

An ethylene molecule drifts into the empty site and binds sideways through its C=C double bond. Then comes the key move you already know: migratory insertion. The growing chain migrates onto the bound ethylene, the double bond becomes a single bond, and the chain is now two carbons longer — with a fresh empty site exactly where the ethylene used to sit. Another ethylene drops in, inserts, lengthens the chain by two more carbons, and the cycle simply repeats, thousands of times per chain, adding two carbons each pass. There is no oxidation-state change at all; titanium stays put while the chain grows out of it like thread off a spool.

Chain growth (one turn of the loop):

  Ti-CH2CH2~~~   +   CH2=CH2
         |              (binds in empty site)
         v   migratory insertion
  Ti-CH2CH2-CH2CH2~~~      <- chain is 2 carbons longer,
         |                    empty site restored
         v   ... repeat thousands of times ...

  Chain release: beta-hydride elimination
  Ti-CH2CH2~~~  ->  Ti-H  +  CH2=CH~~~ (dead chain)
Insertion grows the chain two carbons at a time; a beta-hydride elimination cuts it loose and resets the metal.

What stops a chain and sets its length? Mostly a competing beta-hydride elimination: now and then, instead of inserting another monomer, the metal pulls a hydrogen off the second carbon of its own chain, releasing a finished polymer molecule with a double bond at its tail and leaving a metal hydride that promptly starts a brand-new chain. The race between insertion (grow) and beta-hydride elimination (terminate) sets the average chain length — and therefore the molecular weight, the very property that decides whether you get a flexible film or a rigid bottle. Adding a little hydrogen gas to the reactor is the standard industrial knob: it terminates chains and trims the molecular weight on demand.

Single-site metallocenes and the gift of stereocontrol

Propylene exposes a problem ethylene never had. Each propylene carries a dangling methyl group, so as the chain grows, every inserted unit must choose which face of the prochiral alkene presents to the metal — and that choice fixes whether each methyl points the same way as its neighbours or alternates. If the methyls land in a regular, all-on-one-side pattern (isotactic) the chains pack into a strong, high-melting crystalline plastic; if they scatter at random (atactic) you get a soft, tacky gum. The original heterogeneous Ziegler-Natta solid had a mixture of slightly different active sites, so it made mostly-isotactic material with an inevitable spread of behaviour.

The modern answer is the single-site metallocene catalyst: one molecularly defined complex in which a metal like zirconium is gripped between two flat aromatic rings — the same Ziegler-Natta chemistry, but now the cyclopentadienyl-type rings are tied together by a bridge and decorated to make a precisely shaped, chiral pocket. Because every metal atom is identical, every chain grows under identical rules. The shape of that pocket physically forces each incoming propylene to present the same face, so the methyls march down the chain in lockstep. By choosing the symmetry of the ligand cage, a chemist selects the outcome on purpose: a chiral, C2-symmetric pocket gives clean isotactic polypropylene, while a different mirror-asymmetric cage gives the strictly alternating syndiotactic pattern.

What design buys you, and what to keep honest

Step back and the unifying lesson is clear. In both industries a single metal atom turns over its cycle again and again, untouched at the end — the very definition of a catalyst, captured by its turnover number, the count of product molecules one metal centre makes before it dies. The art is choosing the metal and tailoring the ligands so that the right step is fast, the unwanted side step is slow, and the product comes out with the rate, the selectivity, and the shape you want. Rhodium versus iridium moved the rate-determining step; the metallocene's chiral pocket moved the stereochemistry. Same toolbox, different dials.

Now a few honest cautions, because the clean cartoons hide real subtlety. The arrows we draw for a cycle are a model: many proposed intermediates are too short-lived to bottle, and the mechanisms above were pieced together from kinetics, labelling, and spectroscopy, not seen directly. Oxidation states, climbing by two in the Monsanto oxidative addition, are bookkeeping that helps us track electrons, not literal charges sitting on the metal. The Ziegler-Natta cycle is genuinely simpler than Monsanto's — it is essentially one repeated insertion — but the real industrial solid is a messy surface with many site types, which is exactly why the molecularly defined metallocene was such a leap. And isotactic versus atactic is a spectrum of regularity, not a clean on-off switch.

One last honest note about scale and worth. Acetic-acid carbonylation is prized partly for its atom economy: methanol and CO assemble almost entirely into product, with little waste — a green-chemistry virtue baked into the stoichiometry. Polyolefins, by contrast, owe their dominance to making strong, cheap, light materials from the simplest possible feedstocks. Both rest on the same quiet miracle you have now seen twice: a handful of elementary organometallic steps, looped around a single carefully chosen metal, doing work at a scale that literally shapes the material world.