Reduction of Nitriles

Nitrile-to-primary-amine reduction (via LiAlH₄, borane, or catalytic hydrogenation) is the industrial backbone of aliphatic diamine manufacture: hexanedinitrile (adiponitrile) is hydrogenated at scale over cobalt or nickel catalysts, with excess ammonia added specifically to suppress the secondary-amine condensation pathway discussed above, to produce 1,6-hexanediamine, the diamine half of nylon-6,6. The same nitrile-reduction logic underlies pharmaceutical routes to primary amine side chains — many β-blocker and antihistamine syntheses install a nitrile early (cheap, robust to a wide range of other reaction conditions) and reduce it only at a late stage specifically because nitriles survive Grignard, Wittig, and acidic/basic hydrolysis conditions that would destroy a free amine or aldehyde carried through the same sequence. DIBAL-H's chelation-controlled stop at the aldehyde oxidation level is exploited constantly in total synthesis whenever a chain needs to be extended by exactly one carbon while preserving a sensitive aldehyde for a subsequent Wittig, aldol, or Grignard step; because the DIBAL route to an aldehyde proceeds through the same electrophilic-nitrile-carbon logic developed for the parent Structure and Reactivity of Nitriles chemistry, it is routinely chosen over direct ester-to-aldehyde reduction (which faces an analogous but less forgiving over-reduction problem) when a nitrile precursor is available. In flavor and fragrance chemistry, controlled nitrile reduction to aldehydes is used industrially to access aromatic and aliphatic aldehydes that would otherwise require harsher, less selective oxidative routes from alcohols. Finally, in biochemistry and enzymology, nitrile hydratases and nitrilases perform closely related C≡N functionalization biocatalytically, and synthetic nitrile reduction chemistry is frequently used as the benchmark against which these biocatalytic alternatives — prized for their aqueous, mild-temperature operation — are compared for green-chemistry process design; nitriles reduced this way are also a common masked form of the carboxylic acid oxidation level exploited synthetically before a Decarboxylation Reactions step is ever needed downstream, since a nitrile survives conditions that would decarboxylate a free acid prematurely.

Ask the tutor: Why does the Lewis acidity of aluminum specifically (versus boron in NaBH₄) make chelation-controlled partial reduction of nitriles possible?

A nitrile carbon is sp-hybridized, linearly bonded to R and triply bonded to N. Two orthogonal π orbitals of the C≡N system are available as the LUMO pair, exactly analogous to the two π systems of an alkyne, but polarized: nitrogen is more electronegative than carbon, so the carbon end of each π orbital has the larger coefficient. Hydride (from AlH₄⁻ or from an Al–H bond in DIBAL) attacks along a trajectory roughly perpendicular to the C≡N axis, consistent with Bürgi–Dunitz-type attack on a linear π system, delivering H⁻ to carbon and pushing the electron pair onto nitrogen. The first hydride addition converts the linear, sp nitrile carbon into a trigonal, sp² imine carbon: R–CH=N⁻(–M) — a metalated aldimine, formally the conjugate base of an imine, isoelectronic with an enolate. This nitrogen anion is basic and only weakly electrophilic at carbon (the negative charge is stabilized on the electronegative nitrogen, exactly as an enolate is stabilized on oxygen). Whether this species reacts further depends entirely on how tightly the metal holds the nitrogen and how much free hydride is still around. With <strong>LiAlH₄</strong>, three more hydrides remain on the aluminum after the first delivery, the medium is a coordinating ether solvent (Et₂O or THF) that solvates Li⁺ and keeps the aluminate hydrides reactive, and the reaction is normally run at reflux with excess reagent. The metalated imine, though a poorer electrophile than the nitrile, is still attacked by a second equivalent of hydride, collapsing the C=N π bond entirely and giving a dianionic species, R–CH₂–N²⁻(AlH₃)⁻-type aluminum amide, which upon aqueous quench simply protonates twice to the primary amine. With <strong>DIBAL-H</strong> [(iBu)₂AlH], the picture changes for two reasons. First, DIBAL is normally used as exactly one equivalent at low temperature, so there is no large excess of unreacted hydride sitting nearby waiting to attack a second time. Second — and mechanistically the more important point — the aluminum in the metalated-imine intermediate is trivalent and strongly Lewis acidic*, and it chelates directly onto the newly formed imine nitrogen lone pair. This produces a four-membered or tightly associated N→Al adduct that is both sterically shielded (two isobutyl groups still on aluminum) and electronically deactivated: the nitrogen lone pair, now donated into the empty aluminum p-orbital, is no longer available to push extra electron density toward carbon, and the resulting aza-aluminate is thermodynamically stable at −78 °C. It simply does not react with a second hydride under the reaction conditions; it survives, as a bench-stable-looking salt in solution, until the chemist adds aqueous acid on workup.

The oxidation-level bookkeeping for a nitrile carbon, treating each bond to N as equivalent to a bond to O for the purposes of oxidation-state counting (both are more electronegative than C), places the nitrile at the same level as a carboxylic acid derivative:

where: each hydride delivery is a two-electron reduction, exactly matching the acid → aldehyde → alcohol ladder for esters and acids, except the nitrogen retains its substituents (eventually two H's) instead of gaining an oxygen back. The stoichiometric requirement for full reduction, reading off the two-step ladder:

where: the aldehyde pathway requires exactly one hydride delivered to carbon; the second "half" of the reduction is never a hydride at all — it is a proton transfer/hydrolysis of the resulting imine on workup, which is why controlling hydride equivalents delivered to carbon, not total reagent added, is the real handle on selectivity. The thermodynamic condition for DIBAL to stop cleanly at the metalated-imine stage is that the rate of intramolecular N→Al chelation and the resulting deactivation vastly exceeds the rate of intermolecular hydride transfer from a second DIBAL (or, worse, from the same aluminum after ligand reorganization):

where: $k_{\text{chelate}}$ is fast because it is an intramolecular, entropically cheap Lewis acid–base pairing between the newly formed nitrogen lone pair and the aluminum already tethered to that same nitrogen (it does not even need to diffuse), while $k_{\text{2nd H}^- \text{transfer}}$ is slow at −78 °C both because it is intermolecular and because the nitrogen lone pair that would otherwise activate the adjacent carbon toward further reduction is now tied up donating into aluminum.

Mechanism 1 — LiAlH₄ reduction of a nitrile to a primary amine

Goal: show, with full curved arrows, how R–C≡N becomes R–CH₂–NH₂ via two discrete hydride deliveries and an aqueous workup, and why excess reagent and warm conditions guarantee the amine rather than the aldehyde.

Step 1 (first hydride delivery). AlH₄⁻ presents one of its four hydridic Al–H bonds to the electrophilic nitrile carbon. A curved arrow runs from the Al–H σ bond to the nitrile carbon; simultaneously, one pair of π electrons from the C≡N triple bond curls onto nitrogen. The nitrile's linear sp carbon rehybridizes toward sp² as the C=N imine forms, and the nitrogen becomes a formal anion, now coordinated (through that same lone pair) to the aluminum that just delivered the hydride, giving a four-coordinate aluminum amide/imide complex: R–CH=N–AlH₃⁻ (charge-balanced by Li⁺).

Step 2 (second hydride delivery). In THF or ether at reflux, with three Al–H bonds still present on that same aluminum (or on a fresh equivalent of LiAlH₄ in solution — both pathways are documented, and the reaction is normally run with several equivalents of hydride so this step is never starved), a second hydride is delivered to the same imine carbon. The curved arrow runs from a second Al–H bond to the sp² imine carbon; the C=N π electrons move fully onto nitrogen, converting the imine into a saturated, doubly-anionic aluminum amide: R–CH₂–N²⁻(AlHₓ). This step is viable specifically because, unlike in the DIBAL case, the aluminum delivering hydride is not simultaneously chelating and deactivating the nitrogen of this substrate molecule — with excess AlH₄⁻ present, the reactive hydride donor and the metalated nitrogen need not be the same aluminum center, and the anionic amide nitrogen, while poorer than the neutral imine at delivering electron density back to carbon, is still attacked because so much hydride excess is present and the reaction is driven at reflux.

Step 3 (aqueous workup, first protonation). Dilute acid (or Fieser workup, sequential H₂O/NaOH/H�2O for aluminum salts) delivers a proton to nitrogen. A curved arrow runs from the N–H bond being formed (proton source) is drawn conventionally the other way: the nitrogen lone pair attacks H⁺, giving R–CH₂–NH⁻(AlHₓ) still bound to aluminum, or, once aluminum–nitrogen bonds are hydrolyzed by excess water, a free amide-nitrogen anion R–CH₂–NH⁻.

Step 4 (second protonation, release of amine). A second equivalent of water or acid protonates the remaining nitrogen lone pair, giving the neutral primary amine R–CH₂–NH₂ and releasing aluminum as Al(OH)₃/aluminate salts, which are removed by filtration (Fieser workup) or aqueous extraction.

Net transformation:

where: the two hydrides come from AlH₄⁻ (delivered in the reaction step) and the two protons come from the aqueous workup (not from the hydride reagent), so the mechanism cleanly separates "reduction chemistry" from "acid–base workup chemistry" — a distinction worth keeping sharp because it is exactly the same separation that fails to matter for the amine pathway but matters enormously for DIBAL.

Mechanism 2 — DIBAL-H reduction of a nitrile to an aldehyde

Goal: show why exactly one hydride delivery, trapped as a stable aluminum-chelated imine, survives until hydrolytic workup, releasing the aldehyde rather than over-reducing to the amine.

Step 1 (coordination). DIBAL-H, (iBu)₂Al–H, is itself a three-coordinate, Lewis-acidic aluminum with an empty p-orbital. Before hydride transfer, this empty orbital coordinates to the nitrile nitrogen lone pair (nitrogen's sp lone pair is oriented exactly along the C≡N axis and is stereoelectronically well-placed to donate into an approaching Lewis acid). This Lewis acid–base adduct, R–C≡N→Al(iBu)₂H, both activates the nitrile carbon toward hydride attack (the aluminum withdraws electron density through the σ framework and lowers the π* energy) and pre-organizes the aluminum for what comes next.

Step 2 (intramolecular hydride delivery). With the aluminum already tethered to nitrogen, the Al–H bond delivers hydride to the adjacent, now more electrophilic, nitrile carbon in what is effectively an intramolecular (or tightly ion-paired) hydride transfer. Curved arrows: Al–H σ bond → nitrile carbon; one C≡N π pair → nitrogen. This is mechanistically continuous with Step 1's coordination — the same aluminum that activated the substrate delivers the hydride, so no diffusion-limited search for a second reactive center is needed and the whole event is fast even at −78 °C.

Step 3 (formation of the stable aza-aluminate, and why it stops here). The immediate product is a four-membered chelate: R–CH=N–Al(iBu)₂, where the nitrogen lone pair generated by the reduction (formally the anion left after hydride addition) sits directly on the aluminum that just delivered the hydride. This N→Al bond is thermodynamically strong (aluminum is oxophilic/azaphilic and strongly Lewis acidic; the isobutyl groups are poor donors so aluminum "wants" this extra ligand) and kinetically fast to form (it is intramolecular). Once formed, the nitrogen lone pair is no longer free to push electron density toward carbon or to be basic toward another proton or hydride source; the carbon is now sp², flanked by R and H, and shielded by the bulky diisobutyl groups on aluminum sitting right at nitrogen. A second equivalent of DIBAL-H (present in only catalytic excess at best, since exactly 1.0–1.1 equiv is used deliberately) essentially never finds this deactivated, hindered imine carbon before the reaction is quenched. This aluminum aza-enolate is the isolable "stopping point" — the entire selectivity of the DIBAL reaction is that this species is a thermodynamic and kinetic trap relative to further reduction.

Step 4 (quench and hydrolysis). The reaction is quenched at low temperature (this is essential — warming before quench risks slow over-reduction or side reactions), typically with dilute aqueous acid (HCl) or with the Weinreb-style Rochelle's salt (potassium sodium tartrate) workup to break up aluminum aggregates cleanly. Water attacks aluminum first, displacing the chelate: a curved arrow from an O–H bond of water to aluminum, with the Al–N bond breaking and its electron pair moving onto nitrogen, liberates the free imine R–CH=NH along with an aluminum hydroxide/alkoxide byproduct.

Step 5 (imine hydrolysis, the real "second half" of the reduction). The free primary imine R–CH=NH is itself hydrolytically unstable — a proton transfer protonates the imine nitrogen (curved arrow: N lone pair attacks H⁺) giving an iminium ion R–CH=NH₂⁺, whose carbon is now strongly electrophilic (positively charged nitrogen withdraws through the π system, exactly as in the corresponding Structure and Reactivity of Nitriles hydrolysis logic run one oxidation level down). A water molecule's oxygen lone pair attacks this iminium carbon (curved arrow: O lone pair → C, with the C=N⁺ π pair moving fully onto nitrogen), giving a tetrahedral hemiaminal-type intermediate R–CH(OH)–NH₂. Proton transfers (deprotonate the new O–H-turned-oxonium as needed, protonate nitrogen) followed by collapse of the C–O bond's partner — actually here it is loss of ammonia — proceed as: the nitrogen leaving group is protonated to NH₃⁺ character, and the oxygen lone pair (now as a neutral or anionic alkoxide/alcohol oxygen) pushes its lone pair to re-form the C=O π bond as NH₃ departs, exactly mirroring hemiaminal-to-carbonyl collapse in ordinary imine hydrolysis chemistry. The result is the free aldehyde R–CHO plus ammonia.

Net transformation:

where: only one hydride is ever delivered by aluminum; the "second reduction step" that would be needed to reach the amine simply never happens, and instead an ordinary imine hydrolysis (a hydration/elimination sequence, not a redox step) converts the metal-free imine into the aldehyde on aqueous workup.

Example 1 — foundational: LiAlH₄ reduction of propanenitrile. Propanenitrile, CH₃CH₂–C≡N, is treated with excess LiAlH₄ in dry Et₂O at reflux, then worked up with Fieser conditions (sequential dilute NaOH/H₂O). Follow the carbon count and hydrogen count explicitly, since this is the step most students get wrong by forgetting the workup protons. Starting material: CH₃CH₂–C≡N has 3 carbons, and the nitrile carbon bears zero hydrogens. First hydride delivery converts C≡N to CH=N⁻(AlH₃⁻): the nitrile carbon now bears one hydrogen (from the delivered hydride) and is doubly bonded to nitrogen. Second hydride delivery (excess LiAlH₄, reflux, so this is not rate-limiting) converts CH=N⁻ to CH₂–N²⁻: the same carbon now bears two hydrogens, single-bonded to a doubly-anionic (aluminum-bound) nitrogen. Aqueous workup delivers two protons to nitrogen (Step 3 and Step 4 of Mechanism 1), converting N²⁻ to neutral NH₂. Final product: CH₃CH₂CH₂–NH₂, propan-1-amine (n-propylamine). Check the atom economy explicitly:

Left side hydrogen count on the relevant carbon/nitrogen fragment: nitrile carbon has 0 H, nitrogen has 0 H (lone pair only); right side: that carbon has 2 H, nitrogen has 2 H — a net gain of 4 H atoms across carbon+nitrogen, supplied as 2 hydride equivalents (contributing 2 H, each as H⁻, i.e., H atom plus 2 electrons) and 2 proton equivalents (contributing 2 H⁺). This 2+2 split (not "4 hydrides") is the detail that most reliably separates students who have memorized the outcome from students who understand the mechanism — the reagent is only ever responsible for the hydride half.

Example 2 — applied: converting a Grignard-incompatible substrate into an aldehyde via nitrile homologation with DIBAL-H. Suppose the synthetic target is 3-phenylpropanal, PhCH₂CH₂CHO, and the available starting material is (2-bromoethyl)benzene, PhCH₂CH�2Br. A direct route to the aldehyde via a Grignard reagent (PhCH₂CH₂MgBr) attacking a formate ester or DMF is plausible but competes with elimination and requires strictly anhydrous handling of an unstabilized primary Grignard; the nitrile route is often more forgiving. Step 1: Displace bromide with cyanide. PhCH₂CH₂Br + NaCN (DMSO or aqueous ethanol, SN2) gives PhCH₂CH₂–C≡N, 3-phenylpropanenitrile, extending the chain by exactly one carbon (the nitrile carbon becomes the future aldehyde carbon — this is the key homologation bookkeeping: cyanide's carbon is the new carbonyl carbon). Step 2: Reduce with exactly 1.05 equivalents of DIBAL-H in toluene at −78 °C, added slowly, keeping the internal temperature below −70 °C throughout the addition (the low temperature is what prevents the aluminum aza-enolate from being attacked by any adventitious excess hydride, and it also slows any competing direct reduction pathway that bypasses chelation). Step 3: Quench at −78 °C by adding the cold reaction mixture into a vigorously stirred mixture of dilute HCl (or into methanol first, to destroy excess DIBAL-H as isobutane + aluminum methoxide species, followed by aqueous acid to hydrolyze the metalated imine). Warming is deliberately delayed until after the quenching acid is present, so that the imine intermediate is never exposed to warm conditions in the presence of any remaining hydride. Following Mechanism 2 exactly: PhCH₂CH₂–C≡N + DIBAL-H → PhCH₂CH₂–CH=N–Al(iBu)₂ (stable chelate, isolable as a solution-phase species at −78 °C) → (H₃O⁺ workup) → PhCH₂CH₂–CH=NH (free imine) → (hydrolysis, Step 5 of Mechanism 2) → PhCH₂CH₂–CHO + NH₃. Final product: 3-phenylpropanal, PhCH₂CH₂CHO — the chain has been extended by one carbon relative to the original alkyl bromide, and the oxidation level has been set at the aldehyde exactly because the DIBAL-mediated single hydride delivery was trapped by chelation rather than driven to completion.

Example 3 — exam/competition level: chemoselective differentiation of a dinitrile, and predicting the outcome of a stoichiometry-limited LiAlH₄ reduction. Consider hexanedinitrile (adiponitrile), N≡C–(CH₂)₄–C≡N, treated with only 1.0 equivalent of LiAlH₄ (i.e., 4 hydride equivalents total per molecule of reagent, but only one molecule of reagent per molecule of dinitrile) at 0 °C in THF, followed immediately by a rapid, cold aqueous quench before the mixture is allowed to warm. First, recognize the stoichiometric trap this problem is designed to test: LiAlH₄ carries four hydrides per aluminum, and the dinitrile has two nitrile groups, each of which needs two hydrides to reach the amine. A naive "one equivalent of LiAlH₄ per one equivalent of substrate" reading might suggest exactly enough hydride (4 total) for full double reduction at both nitrile termini (2 + 2 = 4). This is numerically true — but the mechanistic point being tested is whether that stoichiometric coincidence guarantees clean full reduction, or whether kinetics intervenes. In practice, at 0 °C (deliberately cold, not reflux) with rapid quenching, the first hydride delivery at each nitrile terminus (converting C≡N to the metalated imine, the fast, highly favorable step because the nitrile carbon is a superb electrophile) outcompetes the second, slower hydride delivery to the already-anionic, less electrophilic metalated-imine carbon at either terminus. This is exactly the same electronic argument used in Mechanism 1, Step 2: the second hydride delivery is intrinsically slower than the first because it attacks a poorer electrophile (an anionic, nitrogen-stabilized sp² carbon rather than a neutral, linear sp carbon). Under kinetically controlled, cold, rapidly-quenched conditions, the statistically and electronically favored outcome is that most molecules receive one hydride at each end (using all 4 hydrides, 1 per nitrile terminus, across the molecule) rather than two hydrides fully consumed at one terminus while the other remains untouched — because the first hydride addition at any nitrile is always faster than the second hydride addition at any imine, regardless of which specific nitrile group in the molecule is reacting. Quenching immediately at 0 °C before the (slower) second reduction step can occur at either terminus traps both nitrile groups as their respective metalated imines, and hydrolysis on workup (Step 5-type imine hydrolysis, exactly as in the DIBAL mechanism, since a metalated/free imine hydrolyzes the same way regardless of which reagent generated it) converts each C=NH into C=O. Predicted product: OHC–(CH₂)₄–CHO, hexanedial, isolated instead of the "expected" diamine H₂N–(CH₂)₆–NH₂ that naive stoichiometric bookkeeping (4 hydrides = enough for full reduction) would predict. The lesson: total hydride count is necessary but never sufficient information; the rate of each individual hydride-delivery step, and the temperature/quench protocol that determines which steps actually get to occur, governs the real product distribution. (In practice, controlled partial reduction of dinitriles to dialdehydes is difficult to execute cleanly and is rarely used preparatively for exactly this reason — the two steps' rate difference is real but not always large enough for perfectly clean stopping, which is precisely why DIBAL-H, with its chelation-based thermodynamic trap rather than a purely kinetic one, is the reagent of choice whenever a clean nitrile-to-aldehyde reduction is actually required.)

Hans Meerwein, Vladimir Ipatieff, and later, most directly relevant here, Herbert C. Brown and the organoaluminum chemists who developed DIBAL-H's synthetic use, form the intellectual lineage behind controlled nitrile reduction — but the figure most tightly associated with the specific chemistry of complex metal hydrides (LiAlH₄ itself) is Hermann I. Schlesinger, whose wartime and postwar boron/aluminum hydride work at the University of Chicago produced both LiAlH₄ and NaBH₄ as practical laboratory reagents, transforming carbonyl and nitrile reduction from a stoichiometric metal/acid or catalytic hydrogenation problem into a clean, homogeneous, low-temperature hydride-transfer problem that could be understood and controlled mechanistically for the first time.