A small alphabet for a big language
You arrive at this final guide of the rung already carrying two powerful tools. You can bolt a carbon onto a metal and recognise the metal-carbon bond that defines organometallic chemistry, and — your reflex move on any new compound — you can run the valence electron count and read off whether the metal is coordinatively saturated near 18 or hungry at 16. Now we put those tools to work. The reason organometallic chemistry powers so much of modern catalysis is almost suspiciously simple: every catalytic cycle, however elaborate, is built from a tiny alphabet of elementary reaction steps. Learn maybe five or six of them and you can spell out, and read, an enormous range of chemistry. This guide teaches that alphabet. The next rung will use it to write whole words — actual catalytic cycles.
Here is the whole alphabet, in three pairs. Ligand association and its reverse, dissociation — a ligand simply walks onto, or off of, the metal. Oxidative addition and its reverse, reductive elimination — the metal breaks a bond into two new ligands, or stitches two ligands back into one bond. Migratory insertion and its reverse, beta-hydride elimination — two neighbouring ligands fuse into one, or one ligand splits while passing a hydrogen back to the metal. Three pairs, six moves. Notice they come in forward-and-reverse couples: this is no accident, because catalysis is fundamentally about lowering barriers, and a step that is easy in one direction is, by microscopic reversibility, available in the other too. We will take each pair in turn and watch two numbers like a hawk: the metal's oxidation state and its electron count.
Association and dissociation: opening and closing a seat
The gentlest pair first. In ligand dissociation a ligand lets go and floats away; in ligand association a fresh ligand wanders in and binds. Nothing is oxidised or reduced — no bonds within the ligands are made or broken — so the metal's oxidation state does not change at all. What changes is only the seat count. Lose a two-electron donor and the electron count drops by two; gain one and it rises by two. So dissociation takes an 18-electron complex down to a reactive 16, opening the vacant coordination site that the rest of the cycle needs; association fills a site back up, often 16 climbing to 18. These are exactly the two steps you already met under another name in the mechanisms rung, where ligand dissociation and association were the two limiting routes for substitution at a complex.
Why does this matter so much? Because catalysis lives on the 16-electron, coordinatively unsaturated species. A perfectly stable, saturated 18-electron complex is, in a sense, *too* contented to react — it has no empty seat for the substrate. The opening dissociation that creates a vacant site is therefore often the very first event of a cycle, and it is frequently rate-limiting. This is also why ligands that can swing on and off easily, or even slip from binding several atoms to binding fewer, are prized: they let the metal breathe between saturated rest and unsaturated readiness without the catalyst falling apart.
Oxidative addition and reductive elimination
Now the dramatic pair. In [[oxidative-addition|oxidative addition]] an incoming molecule X-Y — say H-H, or a carbon-halogen bond like CH3-I — lands on the metal and *splits in two*, so that both X and Y end up bonded to the metal as separate ligands. The metal has, in effect, used two of its own electrons to forge the two new bonds, so it is formally oxidised: its oxidation state rises by two. And because it has acquired two new ligands, its electron count also rises by two. A classic case is the late-metal centre Ir(I), a 16-electron complex, swallowing H2 to become a 7-coordinate Ir(III) dihydride at 18 electrons. The name is honest about both halves: *oxidative* (the metal's OS goes up) and *addition* (two ligands are added).
Run that film backwards and you have [[reductive-elimination|reductive elimination]]: two ligands that sit next to each other on the metal — a hydride and a methyl, two methyls, a hydride and a halide — couple together and leave as a single new molecule (CH4, ethane, HX). The metal recovers the two bonding electrons, so it is *reduced*: oxidation state falls by two, and having shed two ligands, its electron count falls by two as well. Reductive elimination is almost always how the freshly built product is finally expelled from the catalyst, regenerating a lower-oxidation-state, electron-poorer metal ready to start again. It is the workhorse that forges the new C-H or C-C bond in cross-coupling.
Migratory insertion and beta-hydride elimination
The third pair is where carbon frameworks actually grow. In [[migratory-insertion|migratory insertion]] two ligands already on the metal — typically an unsaturated one like a coordinated alkene or a carbonyl, sitting next to a sigma-bonded group like a hydride or an alkyl — fuse into a single, longer ligand. Picture a metal carrying both a hydride and a side-on bound ethylene: the hydride *migrates* onto one alkene carbon, the other carbon swings up to bond the metal, and where there were two ligands there is now one ethyl group. The metal forms no new bond into a different element, so the oxidation state does not change. But two ligands became one, vacating a seat, so the electron count drops by two — typically 18 sliding to 16, re-opening a site for the next molecule. Repeat this insertion over and over and you are growing a polymer chain one monomer at a time.
The reverse is [[beta-hydride-elimination|beta-hydride elimination]], and it is one of the most important — and sometimes most annoying — moves in the whole field. Take a metal alkyl, say a metal-CH2CH3. The carbon directly bonded to the metal is the alpha carbon; the next one along is the beta carbon. If that beta carbon carries a hydrogen, and an empty seat opens on the metal, the C-H bond can swing in close, the metal plucks the hydrogen off, and the alkyl departs as an alkene. So one ligand became two — a hydride and a coordinated alkene — the electron count rises by two (16 back up to 18), and again the oxidation state does not change. This is why a metal alkyl with beta-hydrogens and an open site is often fleeting: it bleeds away to metal hydride plus alkene. Catalyst designers either exploit this (it is a chain-terminating step in some polymerisations) or carefully block it when they need the alkyl to survive.
Notice the lovely symmetry in the ledger. Of our three pairs, only oxidative addition and reductive elimination touch the oxidation state, and they move it by plus or minus two. The other four moves — association, dissociation, migratory insertion, beta-hydride elimination — leave the oxidation state untouched and only shuffle the electron count by two. Memorise that single fact and you have already memorised half of what every cycle in the next rung will ask of you.
The bookkeeping at a glance
It is worth seeing all six moves side by side, with their effect on oxidation state and electron count laid out in two clean columns. The deltas below are the per-metal-centre values for the common two-electron versions; the exceptions noted earlier still apply.
step delta OS delta EC --------------------------------------------- ligand association 0 +2 ligand dissociation 0 -2 oxidative addition +2 +2 reductive elimination -2 -2 migratory insertion 0 -2 beta-hydride elimination 0 +2 rule of thumb: only OA / RE move the oxidation state. a valid cycle returns the metal to its starting OS and EC.
Read the table and a deep constraint jumps out: across a full catalytic cycle the metal must end exactly where it began, on *both* columns. So the plus-two of an oxidative addition has to be paid back by a reductive elimination somewhere later; an 18-down-to-16 dissociation must be answered by something that climbs back to 18. The cycle is a closed loop in this two-dimensional bookkeeping space, and the substrate's transformation is what the loop spits out each time around. Every elementary step is just a step short enough to have a clean, well-understood transition state — that is precisely why chemists insist on breaking cycles down into them rather than waving at one big arrow.
Assembling the steps into a cycle
Let us preview how the alphabet spells a word, using the most famous example: catalytic hydrogenation, turning an alkene plus H2 into an alkane. Walk it once and watch both ledgers close.
- Start from a coordinatively unsaturated metal centre with an open site — created by an opening ligand dissociation (OS unchanged, EC down to a reactive 16).
- Oxidative addition of H2 splits it into two hydrides on the metal (OS up by two, EC up by two).
- Ligand association: the alkene wanders in and binds side-on, filling the open seat (OS unchanged, EC up by two).
- Migratory insertion: a hydride migrates onto the bound alkene, fusing them into a metal alkyl (OS unchanged, EC down by two, re-opening a seat).
- Reductive elimination: the remaining hydride and the alkyl couple and leave as the alkane product (OS down by two, EC down by two) — and the metal is back exactly where it started, ready for the next turn.
Tally the oxidation-state column down that list — 0, +2, 0, 0, -2 — and it nets to zero; tally the electron count and it returns to its starting value. The loop closes on both numbers, which is exactly the consistency check the earlier tip promised. The substrate, meanwhile, has been quietly transformed from alkene to alkane, and the metal is untouched, free to do it all again thousands of times — that repeat-count is the turnover number that measures a catalyst's stamina. The next rung will dwell on real, named cycles built from exactly these moves — among them Wilkinson's hydrogenation catalyst and the cross-coupling cycles that hinge on the d8 16-electron square-planar resting state. You now hold the alphabet they are written in.