Heat butanenitrile, CH3CH2CH2C≡N, in aqueous acid and it turns into butanoic acid. No oxidizing agent is needed. A nitrile has no carbonyl and no leaving group, so it looks nothing like the RCO–X compounds the rest of this chapter is built on. It belongs here anyway, and the reason is a simple count of bonds.
Structure first
Acetonitrile, CH3C≡N, is the simplest example. Its nitrile carbon is bonded to two atoms, the methyl carbon and the nitrogen, so it is sp hybridized, just like an alkyne carbon (Hybridization). It makes two sigma bonds and two pi bonds, and the C≡N bond is 116 pm long. Its bond angle is 180°, so the C–C≡N unit is a straight line.
The nitrogen keeps one lone pair, in an sp orbital that points straight out, away from the carbon. A nitrile can donate that pair to an electron-poor atom, so it is a weak Lewis base (Lewis acids and bases). The same lone pair lets a nitrile bind to a metal.
Nitrogen is more electronegative than carbon, so the triple bond is strongly polarized toward nitrogen. Nitriles therefore have large dipole moments, and they boil high for their size. Acetonitrile (41 g/mol) boils at 82 °C, while propyne, CH3C≡CH (40 g/mol), boils at −23 °C.
A polar aprotic solvent has a large dipole and no O–H or N–H to donate a hydrogen bond. Acetonitrile fits, and it is one of the standard ones from Substitution & Elimination. It also mixes with water, which makes it an easy solvent for running reactions and for chromatography, the lab method that separates a mixture by how fast each part moves through a column.
The polarization also reaches the next carbon. A hydrogen on the carbon beside the C≡N is weakly acidic: CH3CN has a pKa of about 25. When that proton leaves, the carbon's lone pair can move into the C≡N, and one pair of pi electrons moves onto the nitrogen. Resonance spreads the charge onto the more electronegative atom, and that stabilizes the anion. A hydrogen next to a C=O is acidic for the same reason (Factors affecting acidity).
To name a nitrile, count the nitrile carbon as C1 and add -nitrile to the name of the parent alkane. CH3CH2CH2C≡N has four carbons, one of them the carbon bonded to nitrogen, so it is butanenitrile (Functional group priority).
Same rung as a carboxylic acid
Count the bonds from the nitrile carbon to a heteroatom: three, all of them to the same nitrogen. A carboxylic acid carbon also has three, to oxygen. So the two sit on the same rung of the oxidation ladder from Oxidation levels: the +3 level, which the acid shares with the ester, the amide and the acid chloride.
That count tells you which reagents a nitrile needs. Turning it into any other member of the rung is redox-neutral, so no oxidant is required. Leaving the rung, down to an aldehyde or an amine, takes a reducing agent, as it does for every other member.
Where nitriles come from
- SN2 on a primary alkyl halide. RCH2X + NaCN → RCH2CN. Cyanide is a strong nucleophile and only a weak base (HCN has a pKa of 9.2), which is the combination SN2 wants. The cyanide carbon stays, so the chain grows by one.
- Dehydration of a primary amide with SOCl2 or P2O5. RCONH2 → RCN, and water is removed. This is a sideways move along the rung.
- From a cyanohydrin. Cyanide adds to an aldehyde or ketone, and the product carries an OH on the carbon next to the C≡N (Hydrates & cyanohydrins).
The SN2 route needs a primary substrate. Secondary halides react slowly and give some elimination product as well. Tertiary halides give no nitrile at all, because a tertiary carbon cannot undergo SN2 (SN2).
Cyanide is too weak a base to force a fast E2, so the solvent decides what a tertiary halide gives instead. In a protic solvent such as aqueous ethanol, the halide mostly solvolyzes: the solvent replaces the bromide, and the main product is the alcohol. In a polar aprotic solvent such as DMSO, with heat, elimination takes over and gives the Zaitsev alkene. Either way, plan a chain extension with cyanide around a primary substrate.
SN2 cannot put a nitrile on a benzene ring, because a ring carbon cannot be attacked from behind. Aryl nitriles need other routes, which Diazonium salts and Sandmeyer covers (a preview; nothing here depends on it).
What a nitrile does
| Reagent | Product | Through |
|---|---|---|
| H3O+ or HO−, heat | RCOOH | the amide |
| LiAlH4, then H2O | RCH2NH2 | an aluminum-bound imine, then a second hydride |
| DIBAL-H, 1 equiv, −78 °C, then H3O+ | RCHO | an aluminum-bound imine |
| R′MgX or R′Li, then H3O+ | RCOR′ | an imine anion, then an imine |
Hydrolysis runs in aqueous acid or base with heat, and it passes through the amide on the way. That amide is a real intermediate: milder conditions stop the reaction at RCONH2. In base the acid comes out as its carboxylate salt, and adding acid at the end gives RCOOH.
LiAlH4 adds a hydride to the nitrile carbon, and one pair of pi electrons moves onto the nitrogen. The product is an aluminum-bound imine, R–CH=N–Al: an imine whose nitrogen carries the metal where a hydrogen would be. LiAlH4 is reactive enough to add a second hydride to that C=N, and water then gives a primary amine, RCH2NH2. The nitrile carbon becomes the CH2, so no carbon is gained or lost. An amide also reduces to an amine with LiAlH4, for the same reason: the nitrogen stays attached (Reducing carbonyls).
DIBAL-H is diisobutylaluminum hydride, (i-Bu)2AlH: one hydride on a neutral aluminum that also carries two bulky groups. It is a much weaker hydride donor than LiAlH4. One equivalent at −78 °C adds one hydride and gives the same kind of aluminum-bound imine. With the metal on the nitrogen, the C=N is a poor electrophile, too poor for DIBAL-H to attack. That imine waits in the flask until water is added, and aqueous acid then hydrolyzes it to the aldehyde.
So the two reagents differ in strength, not in amount. Using less LiAlH4 does not stop the reaction at the aldehyde; it gives a mixture of amine and unreacted nitrile. Only the reagent's strength stops it at the aldehyde. Reducing carbonyls called this crippling a reagent rather than rationing it.
A Grignard reagent or an organolithium, the carbon nucleophiles from Nucleophilic addition, adds to the nitrile carbon in the same way. Its carbon bonds to the nitrile carbon, and one pair of pi electrons moves onto the nitrogen. The product is an imine anion, RR′C=N−, paired with MgBr+. It is too poor an electrophile for a second Grignard to attack, so the reaction stops after one addition. Aqueous acid at the end protonates it to an imine, and the imine hydrolyzes to a ketone.
An ester behaves differently. After the first Grignard adds, the intermediate expels alkoxide, and a ketone forms while Grignard reagent is still in the flask. The ketone is more reactive than the ester was, so it is attacked again, and the product is a tertiary alcohol. A nitrile makes no carbonyl until water is added at the end, and by then the Grignard reagent is gone. Grignard reagents set out these reagents in full.
One carbon, four destinations
Put the four reactions together, and a nitrile is a one-carbon extension that can end up as an acid, an amine, an aldehyde or a ketone. With the SN2 step in front, RX → RCN → RCOOH is the standard way to lengthen a chain by one carbon and finish with an acid.
You have 1-bromopropane, CH3CH2CH2Br, and you need pentan-2-one, CH3CH2CH2COCH3.
Count first. Three carbons go in and five come out, so two must be added, and they come from different reagents. Cyanide brings the one that becomes the carbonyl carbon. A methyl Grignard brings the other, which ends up on the far side of the carbonyl.
Step 1. NaCN. SN2 on the primary bromide gives butanenitrile, CH3CH2CH2C≡N. It has four carbons, and the new one is the nitrile carbon.
Step 2. CH3MgBr, then H3O+. The Grignard adds once to the nitrile carbon and gives the magnesium imine anion. Aqueous workup hydrolyzes the imine to the ketone, CH3CH2CH2–CO–CH3: pentan-2-one.
Check the alternative. Hydrolyzing the nitrile instead gives butanoic acid, and a methyl Grignard fails on that. The acidic O–H destroys the reagent, and the carboxylate left behind does not react with it.
Two routes to the same acid
RX → RCN → RCOOH is not the only way to add one carbon and finish with an acid. A Grignard reagent does it too. RX and magnesium give RMgX. The Grignard adds to CO2 and gives a carboxylate salt, and aqueous acid then gives RCOOH (Carboxylic acids). The two routes do not duplicate each other. Each one works where the other fails.
The cyanide route needs a primary, unhindered substrate. In return, it tolerates an alcohol O–H elsewhere in the molecule, because that proton is not acidic enough to protonate cyanide. A carboxylic acid O–H is a different case: its pKa of about 5 is lower than HCN's, so the acid protonates cyanide.
Cyanide is still a nucleophile, so it does react with some groups. A ketone or aldehyde elsewhere in the molecule turns into a cyanohydrin, the third route into a nitrile listed earlier on this page.
The Grignard route works on a hindered or tertiary substrate. But any O–H or N–H in the molecule destroys the Grignard reagent, and it attacks any carbonyl it finds (Nucleophilic addition). Both routes reach the same target, so look at the substrate, not at the target, to choose.
A nitrile is also easy to spot by infrared spectroscopy; IR showed how.