From the alkane lab bench to real selectivity
Back in the alkanes you met the radical chain and its three beats — initiation lights a fuse with light or heat, two propagation steps pass the radical around like a baton, and termination ends a strand when two radicals meet. You also met its great flaw: chlorine is so reactive it grabs almost any hydrogen and gives a mess, while bromine is fussier. This guide does not re-derive that chain. Instead it asks the grown-up question — *how do chemists actually use radicals on purpose?* — and the answer turns entirely on selectivity: getting one bond to break out of the dozens on offer.
Two threads control everything that follows. The first is which radical is *most stable*, because a slow, choosy reagent steers toward the easiest, most stabilising hydrogen. The roughly familiar ladder — tertiary above secondary above primary above methyl — is here joined by two new champions that sit even higher: a radical next to a double bond (allylic) or next to a benzene ring (benzylic). Both are extra-stable because the lonely electron is no longer stuck on one carbon; it spreads across a delocalized pi system by resonance. The second thread is sheer reactivity: less reactive reagents are more selective. Hold those two ideas and the rest of this guide unfolds.
NBS: striking only the allylic and benzylic spot
Here is a real problem. You have an alkene — say cyclohexene — and you want to brominate the carbon *next to* the double bond (the allylic position) without touching the double bond itself. But molecular bromine, Br2, loves double bonds: it adds across them in a flash to give a dibromide, the ionic reaction you saw with alkenes. If you pour Br2 on cyclohexene you get addition, not the allylic product you wanted. The trick is to keep the bromine concentration absurdly *low* — too low for the fast ionic addition, but just enough to feed a slow radical chain. The reagent that does this is NBS (N-bromosuccinimide), a tame, solid source that drips out Br2 a trace at a time.
Why does the radical chain pick the allylic hydrogen and ignore the others? Pure stability. A bromine radical abstracts the allylic hydrogen because the carbon radical left behind is allylic — resonance-stabilized, the lonely electron shared across both ends of the system. That is a far easier, far more downhill hydrogen to pull than an ordinary one, so the choosy bromine radical takes it almost exclusively. The same logic makes NBS the go-to for benzylic bromination: a hydrogen on the carbon attached to a benzene ring gives a benzylic radical, equally spread out over the ring, equally favoured. NBS is the practical, selective sibling of the crude alkane halogenation you already know.
Anti-Markovnikov HBr: addition that runs backwards
Now flip the role of the alkene. Instead of taking a hydrogen *off* a carbon next to a double bond, we *add* across the double bond — but by a radical route, and the result is startling. When HBr adds to an alkene the ordinary, ionic way, it follows Markovnikov's rule: the proton goes onto the carbon that already has more hydrogens, so the bromine lands on the more substituted carbon. The real reason, recall, is that the rule is about the most stable carbocation intermediate — the H+ adds wherever it builds the more stable positive charge. Add a trace of a peroxide as a radical initiator, though, and the bromine lands on the *opposite* carbon. Same reagents, reversed regiochemistry.
The reversal is not magic; it is the same stability rule applied to a *radical* instead of a cation, and the order of the two pieces is swapped. In the ionic path, H+ adds first and the carbocation forms. In the radical path, a bromine radical adds *first*. So now we ask which carbon the bromine should attack so as to leave the more stable *carbon radical* behind — and a radical, like a cation, is most stable on the more substituted carbon. The Br therefore adds to the *less* substituted carbon (leaving the radical on the more substituted one), and the hydrogen ends up on the more substituted carbon. That is the mirror image of Markovnikov, so we call it anti-Markovnikov addition.
- Initiation. The weak O–O bond of a peroxide snaps by homolysis into two oxygen radicals, which then pull a hydrogen off HBr to make a bromine radical, Br. This is the only step that needs the peroxide.
- Propagation, step one. The bromine radical adds to the less substituted end of the C=C, building the more stable (more substituted) carbon radical. This single choice is what sets the anti-Markovnikov outcome.
- Propagation, step two. That carbon radical abstracts a hydrogen from a fresh HBr, giving the finished bromoalkane and regenerating a bromine radical to start step one again. The chain turns over.
When radicals run wild: autoxidation and combustion
The very feature that makes the chain so productive — one radical regenerating another, lap after lap — is also what makes radicals dangerous when nobody is steering. The classic slow example is autoxidation: ordinary molecular oxygen, O2, which happens to be a stable double radical (it carries two unpaired electrons), quietly attacks the weak bonds in fats, oils, and even the lipids of your cell membranes. It plucks an allylic-type hydrogen from an unsaturated fatty acid, the carbon radical grabs O2 to form a peroxide, and that peroxide passes the chain on to the next molecule. This is the chemistry of butter going rancid, cooking oil turning sour, and paint hardening in air — a radical chain creeping through your pantry at room temperature.
Combustion is the same family of chemistry with the brakes off. Burning a hydrocarbon in air is a runaway radical chain in which initiation, propagation, and a wealth of branching steps cascade so fast that the energy released as heat and light keeps generating fresh radicals — that self-amplifying flood is what we experience as a flame. It is worth seeing autoxidation and combustion as two speeds of one underlying process: oxygen, a radical, chewing through C–H bonds. One creeps over months in a butter dish; the other roars through a log in minutes. The difference is rate and chain branching, not a different kind of chemistry.
Antioxidants: how to stop a chain in food and in you
If a chain survives because every radical makes a new one, the way to kill it is to feed it a radical that *cannot* carry on. That is exactly what an antioxidant does. A molecule like vitamin E (in vegetable oils) or BHT (the preservative listed on cereal boxes) donates a hydrogen to the marauding chain-carrying radical, quenching it — and the new radical left on the antioxidant is so stabilized by resonance into its aromatic ring that it is content to sit still rather than rip a hydrogen from the next fatty acid. The chain dead-ends. One antioxidant molecule can mop up a propagating radical and halt a whole strand, which is why a pinch goes a long way.
radical stability ladder (most stable on top): benzylic ~ allylic > 3deg > 2deg > 1deg > methyl (resonance-delocalized) (hyperconjugation only) chain killer: R-O-O* + Antioxidant-H -> R-O-O-H + Antioxidant* (the Antioxidant* radical is too stable to continue -> chain ends)
The same drama plays out inside you. Reactive oxygen radicals are an unavoidable by-product of breathing, and left unchecked they would oxidize the lipids in your membranes by exactly this chain. Your body answers with its own antioxidants — vitamin E embedded in the membrane, vitamin C in the watery cytoplasm, and enzymes that defuse peroxides. One honest caution, though: the popular leap from "antioxidants stop a radical chain in a test tube" to "antioxidant megadoses prevent disease" is not well supported, and some large trials of high-dose supplements found no benefit or even harm. The mechanism is real and beautiful; the health claims built on top of it are a separate question that chemistry alone does not settle.