A new kind of ligand — and a counting problem
So far in this rung the ligands have behaved themselves. A carbonyl, a phosphine, a hydride, a methyl — each grips the metal through one atom, so when you do the valence electron count you simply ask how many electrons that one atom hands over. But now meet a family that refuses to be so tidy. A flat carbon ring can lie *face-on* against a metal, touching it through several carbons at once; a plain alkene, with no lone pair anywhere, can still clamp on sideways. These are the pi-ligands, and the obvious first question is brutal: if a ligand touches the metal through five atoms, how many electrons is it donating, and where do you write that down?
The fix is a piece of notation so useful it has become reflex for every organometallic chemist: hapticity, written with the Greek letter eta and a number. Hapticity, or eta-n, simply counts how many *contiguous* atoms of the ligand are bonded to the metal at once. An alkene gripping through both its carbons is eta-2; an allyl group lying sideways through all three carbons is eta-3; a five-membered ring lying flat through all five carbons is eta-5; benzene face-on is eta-6. Think of your own hand: pinching with one fingertip is low hapticity, pressing down with the whole flat palm is high hapticity. The eta number is the chemist's shorthand for how much of the ligand is actually touching.
How an alkene with no lone pair holds on
Start with the simplest pi-ligand, an alkene like ethylene. It has no lone pair to offer — every electron is locked in a bond — so how does it stick to a metal at all? The answer is the Dewar-Chatt-Duncanson model, and it is the exact same two-way handshake you met one guide ago for carbon monoxide. Picture ethylene lying sideways across the metal, its carbon-carbon axis parallel to the metal surface. First, the filled pi bond *between* the two carbons donates its electron pair into an empty metal orbital — a sigma-type gift coming from a pi source. Then a filled metal d orbital of the right shape reaches up and pushes electron density back into the alkene's empty pi-star antibonding orbital. That return flow is pi back-donation, and the two directions reinforce one another into a single strong, synergic bond.
This picture is not just hand-waving; it leaves a fingerprint you can measure. Pouring electrons into the pi-star orbital *weakens* the carbon-carbon bond, so a bound alkene shows a longer, looser C=C than a free one, and its two carbons bend their hydrogens back away from the metal, losing a little of their flat sp2 character. In the classic example, Zeise's salt K[PtCl3(eta-2-C2H4)], an X-ray photograph catches exactly that: the C-C distance stretched, the hydrogens tilted off. When the back-donation is mild the alkene stays electron-poor and primed for a nucleophile to attack it; when back-donation is heavy the alkene goes electron-rich and resists nucleophiles, the bond shading toward a three-membered metallacyclopropane ring. It is a continuum, not an either/or — and being honest, this is a molecular-orbital *model* of where electrons go, predictive and trustworthy, but not a literal photograph of the electrons themselves.
The Dewar-Chatt-Duncanson model matters far beyond one quirky platinum salt, because an alkene binding this way is the *opening move* of an enormous swathe of catalysis. Hydrogenation, polymerization, hydroformylation — they nearly all begin with an alkene sliding onto a metal exactly like this, before it inserts or gets attacked. You will see that first step recur in the final guide of this rung when the catalytic cycle is assembled. For now, hold onto the lesson: a molecule with no lone pair can still be a ligand, as long as it has a filled bonding pi to give and an empty pi-star to receive.
Allyl, cyclopentadienyl, and reading off the electron count
Climb up the family. The allyl ligand, CH2-CH-CH2, can perch on a single end carbon as eta-1 — a plain one-carbon X-type anchor — or lie sideways through all three carbons as eta-3, sharing a delocalized pi cloud spread across the trio. The same skeleton, two different hapticities, two different electron counts. And then comes the star of the whole show: the cyclopentadienyl ligand, written Cp, the C5H5 ring. Picture a tiny, perfectly flat five-pointed coin, five carbons each carrying a hydrogen, with a thin even film of shared electrons over both faces. A metal sticks to one face like a magnet to a fridge door.
Where does Cp's even electron film come from? Cyclopentadiene loses a proton to give the anion C5H5-, which carries six electrons spread right around the five-membered ring — a genuine aromatic, delocalized pi system, the same 4n+2 closed count that makes benzene aromatic. Because the cloud is smeared evenly over the whole ring, the metal does not grab individual carbons; it sits face-on with all five equivalent, an eta-5 attachment. The cyclopentadienyl ligand is generous and bulky: it blankets a large chunk of the metal's coordination sphere and pours in a lot of electron density, which makes it superb at both protecting a metal and filling the count toward eighteen. Dress every carbon with a methyl and you get pentamethylcyclopentadienyl, Cp-star — bulkier, more electron-rich, and giving tougher, more soluble complexes.
Here is the payoff that makes hapticity worth all the notation: the eta number feeds *straight* into the electron count. In the neutral convention, an eta-2 alkene donates 2 electrons, an eta-3 allyl donates 3, an eta-5 Cp donates 5 — the donation just equals the eta number. (In the ionic convention you instead hand the ligand its anion charge first, so Cp goes in as Cp-minus giving 6; remember from the previous guides that the two conventions take different routes but must land on the same total.) Get the hapticity wrong and the whole count and interpretation collapse — which is exactly why eta is not decoration.
CpMn(CO)3 (neutral convention) Mn (group 7) ........ 7 e- Cp (eta-5) .......... 5 e- 3 x CO (eta-1) ...... 6 e- ------------------------ total ............... 18 e- stable, saturated
Ring slippage: a hidden door
Hapticity is not frozen — and that turns out to be quietly powerful. A Cp ring sitting comfortably as eta-5 can *slip* to eta-3, peeling two of its five carbons away from the metal so that only three still bond. Watch what that does to the count: an eta-5 donor giving 5 electrons becomes an eta-3 donor giving 3, so the metal sheds two electrons' worth of crowding without any ligand actually leaving. A saturated 18-electron complex, which by the rules should be unreactive, can therefore behave as if it had an empty seat — a vacant coordination site conjured from nowhere. Chemists call this trick ring slippage.
This is why the eta number earns its keep beyond bookkeeping. It is a control knob the molecule itself can turn. A ligand that can change its hapticity gives the metal a way to open and close a site on demand, which is precisely the kind of breathing room a catalyst needs as it grabs a substrate, transforms it, and lets go. Keep this idea in your pocket for the final guide of the rung: many of the elementary steps of a catalytic cycle lean on a metal having, or making, an open coordination site.
Ferrocene: the sandwich that started everything
Now put two Cp rings together with a metal filling between them and you build the most famous molecule in the field. Ferrocene, Fe(C5H5)2, is an iron atom sandwiched between two flat cyclopentadienyl rings, one above and one below, each bound eta-5 with all five carbons engaged. It is orange, it is air-stable, it can be heated and even distilled — and when it turned up in the early 1950s it detonated the whole modern field of organometallic chemistry. Nobody had expected a metal to sit so contentedly clamped between two aromatic rings, and working out exactly *how* it bonds earned a Nobel Prize.
The bonding is not iron clutching ten separate carbons; it is the iron's d orbitals overlapping with the two rings' delocalized pi clouds, knitting the whole sandwich into one piece. And the electron count comes out gorgeously either way you do it. Neutral convention: iron is group 8, so 8 electrons, and each neutral eta-5 Cp gives 5, totalling 8 + 5 + 5 = 18. Ionic convention: the iron is Fe2+ with six d electrons, and each Cp-minus gives 6, totalling 6 + 6 + 6 = 18. Same answer, two routes, and that closed 18-electron shell is the deep reason for ferrocene's astonishing stability. One charming detail: the two rings spin almost freely against each other, like the two lids of a jar, so it is simply wrong to imagine them locked rigidly in one orientation.
What truly made ferrocene a landmark, though, was not its stability but its *reactivity*. The rings behave chemically much like benzene rings: treat ferrocene with an acyl chloride and a Lewis acid and it undergoes electrophilic substitution — acylation — to give acetylferrocene, just like a Friedel-Crafts reaction on an aromatic compound. Chemists could decorate the sandwich with new groups exactly as they decorate ordinary aromatics. That single discovery married rich organic reactivity to a metal centre and proved that organometallics were a whole fertile new chemistry, not a freak curiosity. Today ferrocene and its derivatives serve as catalysts and ligands, as reversible redox markers (the iron flips cleanly between Fe2+ and Fe3+), as fuel additives, and as building blocks reaching into materials and even medicinal chemistry — the textbook icon of the metallocene sandwich, and the reason this whole rung exists.