Flipping the Carbon's Polarity
Every addition you met earlier in this rung sent an oxygen or nitrogen nucleophile at the carbonyl — water, an alcohol, an amine. They build hydrates, acetals, and imines, but they never touch the carbon backbone; they only hang a heteroatom off it. This guide introduces a different kind of guest entirely: nucleophiles whose attacking atom is CARBON. When a carbon nucleophile adds to a C=O, it forges a brand-new carbon-carbon bond, and that single fact is why these reactions sit at the center of synthesis. Decorating a molecule is useful; growing its skeleton is power.
How do you make carbon, the patient peacemaker of organic chemistry, behave as a nucleophile? You bolt it to a metal that craves the positive charge even more than carbon does. Bond carbon to magnesium or lithium and the shared electrons slump back toward carbon, leaving it electron-RICH — a carbon with a partial negative charge, practically a carbanion. This is the heart of the trick: the carbon's usual polarity is reversed. Where a carbonyl carbon is δ+ and hungry for electrons, this metal-bound carbon is δ− and ready to give. Put the two together and they snap shut like a magnet finding its opposite.
Grignard and Organolithium: One Carbon at a Time
The classic carbon nucleophile is the Grignard reagent, written R-MgBr. You make it by stirring an alkyl or aryl halide R-X with magnesium metal in scrupulously dry ether; the metal inserts into the C-X bond and the carbon emerges nucleophilic. Its fiercer cousin is the organolithium reagent, R-Li, made the same way with lithium metal. Both deliver an R group that behaves as R−, but the lithium reagent is even more reactive and more basic — useful when a stubborn or hindered carbonyl shrugs off the milder Grignard.
- The metal-bound carbon, electron-rich, attacks the δ+ carbonyl carbon; the C=O π bond breaks and its electrons fold up onto oxygen — this is the Grignard addition step proper.
- The carbonyl carbon, once flat and sp2, rehybridizes to sp3, turning tetrahedral; oxygen now carries the negative charge as a magnesium alkoxide.
- Add dilute acid or water in a separate workup step; the alkoxide grabs a proton and you isolate an alcohol that is now one R group bigger than the carbonyl you started with.
The class of alcohol reads straight off the carbonyl partner, like a dial. Grignard plus formaldehyde (H2C=O) gives a primary alcohol; plus any other aldehyde gives a secondary alcohol; plus a ketone gives a tertiary alcohol. That ladder — formaldehyde, aldehyde, ketone yielding 1°, 2°, 3° — is the single most useful pattern here, because it is how you stitch small fragments into a bigger carbon framework. One honest limit worth stating plainly: a Grignard is a ferocious base, so it will rip a proton off ANY -OH, -NH, or -COOH in the flask and die as a harmless alkane before it ever reaches the carbonyl. That is why everything must be bone dry, and why a molecule carrying its own acidic group needs a protecting group first.
Hydride: When You Only Want an H
Sometimes you do not want to grow the skeleton — you already have the right carbon framework and merely wish to soften a sharp C=O into a gentle C-OH. For that, swap the carbon nucleophile for a hydrogen one. A hydride reagent carries an H together with its bonding pair, effectively delivering H−, and that hydride attacks the carbonyl carbon by exactly the same move you just saw: nucleophile in, π electrons onto oxygen, then a workup protonates the alkoxide. Because you added only an H and no carbon, the skeleton is untouched — an aldehyde becomes a primary alcohol, a ketone a secondary one.
Two hydride reagents matter, and their personalities are opposite. Sodium borohydride, NaBH4, is mild and forgiving: it is happy in water or alcohol, reduces aldehydes and ketones cleanly, and mostly leaves esters and carboxylic acids alone. Lithium aluminium hydride, LiAlH4, is fierce: it reacts violently with water, must be used in dry ether, and reduces almost every carbonyl in sight — ketones, aldehydes, esters, even carboxylic acids. So the choice of hydride is itself a control knob. Reach for NaBH4 when you want to touch only the easy carbonyls and spare the rest; reach for LiAlH4 when you want to flatten everything to the alcohol level.
The Wittig Reaction: Trading C=O for C=C
All the additions so far leave an oxygen behind, parked on the carbon as an -OH. The Wittig reaction does something more daring: it carries a carbon nucleophile in, then carries the oxygen back out, replacing the entire C=O with a carbon-carbon double bond. You hand it an aldehyde or ketone and it hands you back an alkene with a new piece of skeleton grafted exactly where the oxygen used to be. For making a double bond at a precisely chosen position, nothing is more direct.
The carbon nucleophile here is a phosphorus ylide: a carbon bearing a negative charge sitting right next to a positively charged phosphorus, written R2C=PPh3 (the two charges nearly cancel, which tames the carbanion). That ylide carbon attacks the carbonyl carbon — same opening move as a Grignard. But then the phosphorus, which loves oxygen dearly, reaches over and grabs it: a four-membered ring forms briefly and collapses, ejecting a very stable phosphorus-oxygen byproduct (triphenylphosphine oxide) and leaving the two carbons joined by a fresh double bond. The thermodynamic prize — that strong P=O bond — is the engine that drives the whole reaction forward.
R2C=PPh3 + O=CR'2 -> R2C=CR'2 + Ph3P=O (ylide) (carbonyl) (new alkene) (driving-force byproduct)
Erasing the Carbonyl Entirely
Reduction with hydride stops at the alcohol — it leaves an oxygen on the carbon. But sometimes you want the carbonyl GONE: not C=O, not C-OH, but a plain CH2 methylene group, as if the oxygen had never been there. Two classic reactions do exactly this, deoxygenating a C=O all the way down to CH2, and they exist as a matched pair because they work in opposite conditions. Whichever your molecule can survive, the other partner usually can.
The Clemmensen reduction uses zinc amalgam in strong hydrochloric acid — it is the ACIDIC route, so it is the one to pick when your molecule carries acid-sensitive groups it must keep, but would be ruined by base. The Wolff-Kishner reduction takes the opposite tack: it first converts the carbonyl to a hydrazone with hydrazine (H2N-NH2), then heats it with strong base, expelling nitrogen gas as the driving force and leaving CH2 behind. It is the BASIC route, the one for molecules that tolerate base but would not survive hot acid. Choosing between them is rarely about yield — it is about which harsh condition the rest of your molecule can endure.
Reading the Carbon-Nucleophile Toolkit Backwards
Lay the tools side by side and a strategy appears. Need a new C-C bond AND an alcohol? Grignard or organolithium. Need a new C-C bond as a double bond, with the oxygen gone? Wittig. Need to keep the skeleton and just drop C=O to C-OH? Hydride — NaBH4 if you must spare other groups, LiAlH4 if you want to reduce everything. Need the carbonyl erased to CH2? Clemmensen if the molecule tolerates acid, Wolff-Kishner if it tolerates base. Each reaction is chosen not by habit but by the exact bond you want to end up holding.
Notice the quiet unity beneath all of it. The very first step of the Grignard, the organolithium, the hydride reduction, and even the Wittig is one and the same: a nucleophilic addition to the carbonyl carbon, electrons settling onto oxygen, the flat sp2 carbon going tetrahedral. What happens AFTER that shared opening — protonate to an alcohol, or let phosphorus haul the oxygen away to an alkene — is what gives each reaction its different product. Learn the one opening move deeply, and these four named reactions stop being four things to memorize and become four endings to a single, familiar story.