Everything you know about substitution on a benzene ring so far runs one way: the ring is electron-rich, so it attacks an electrophile. That is electrophilic aromatic substitution, and it is the whole of the aromatic chapter up to here.
The reverse — a nucleophile displacing a leaving group on the ring — looks impossible at first, and for a plain aryl halide it is. Chlorobenzene does not do SN2, because the backside of that carbon is behind the ring and the carbon is sp². It does not do SN1 either, because an aryl cation is badly unstable.
But there are two ways round it, and which one runs depends entirely on what else is on the ring.
Route one: SNAr, when the ring is electron-poor
Put strong electron-withdrawing groups on the ring, ortho or para to the leaving group, and the picture changes. The nucleophile adds to the carbon bearing the leaving group, giving a negatively charged intermediate called a Meisenheimer complex, and then the leaving group departs.
addition first, elimination second — the opposite order from every substitution you have seen
Three consequences fall out of that mechanism, and all three are testable:
- The withdrawing group must be ortho or para. Only from those positions can the negative charge reach it by resonance. A meta nitro group still withdraws inductively and does speed the reaction up measurably, but induction is the small effect — far too little to make SNAr practical. Resonance is what the reaction needs.
- Fluorine is the best leaving group here — the reverse of every other substitution. The rate-determining step is the nucleophile's attack, not the C–X bond breaking, and fluorine's electronegativity makes that carbon the most electrophilic. F > Cl > Br > I.
- No rearrangement. The nucleophile arrives exactly where the leaving group was.
Route two: benzyne, when the ring is not activated
Take chlorobenzene with no activating groups and force it with a very strong base — NaNH2 in liquid ammonia. Something quite different happens. The base removes a hydrogen ortho to the chlorine, chloride leaves, and what remains is benzyne: a benzene ring with an extra bond made from two sp² orbitals in the plane of the ring.
That bond is a terrible one. The two orbitals point away from each other rather than overlapping face to face, so benzyne is strained, extremely reactive, and lasts only long enough to be attacked.
That scrambling is how benzyne was proved to exist. Label the carbon bearing the chlorine with 14C, run the reaction, and the label comes out split roughly evenly between two positions — which no direct displacement could produce.
Telling them apart
| SNAr | Benzyne | |
|---|---|---|
| Ring needs | EWG ortho or para | Nothing; works on plain rings |
| Conditions | Mild nucleophile, warm | Very strong base, NaNH2 |
| Order | Add, then eliminate | Eliminate, then add |
| Intermediate | Meisenheimer complex | Benzyne |
| Best halide | F, by a wide margin | All work; rates differ, and an ortho H is the real requirement |
| Product | One, at the original position | Two, at adjacent positions |
Read the substrate first. Nitro groups ortho or para to a halide mean SNAr; a bare aryl halide plus NaNH2 means benzyne; and a substrate with no hydrogen ortho to the halide cannot go through benzyne at all, because there is nothing to eliminate.
What carries forward
Two ways to put a nucleophile on a ring, and they are opposites. SNAr is addition–elimination through a stabilized anion, needs withdrawing groups in the right places, and gives one product. Benzyne is elimination–addition through a strained intermediate, needs no activation and a brutal base, and gives two.