Cyanohydrins in Synthesis
A single cyanide ion, delivered to a carbonyl carbon, turns a two-carbon problem into a three-carbon solution. That is the quiet power of the cyanohydrin. It is not just an equilibrium addition product studied for its thermodynamics. Instead, it is a masked, umpolung-free one-carbon-extended synthon. Chemists have exploited it for over a century. They use it to build α-hydroxy acids, α-amino alcohols, and homologated aldehydes. These syntheses achieve genuine stereochemical control. Cyanohydrin Formation establishes how and why cyanide adds reversibly across a carbonyl. This article asks the synthetic question instead: once you have R₂C(OH)CN sitting in your flask, what can you actually do with it? The answer touches nearly every functional-group interconversion a synthetic chemist needs. Hydrolysis of the nitrile under acid or base converts the cyanohydrin directly into an α-hydroxy carboxylic acid. This is the route by which mandelic acid, citramalic acid, and many pharmaceutical intermediates are made industrially. Partial reduction of the nitrile to an imine, followed by further reduction, delivers a 1,2-amino alcohol. This motif is found in adrenaline, ephedrine, and countless β-blockers. Full reduction to the primary amine, or hydrolysis-decarboxylation sequences, extend carbon chains by exactly one carbon. These extensions are difficult to achieve by other means. Each of these transformations is mechanistically transparent. No step requires anything beyond nucleophilic addition to a nitrile or carbonyl, proton transfers, and tetrahedral-intermediate collapse. This is chemistry you already know from ester and amide hydrolysis, dressed in new clothing. What makes cyanohydrin chemistry a genuinely strategic tool, rather than a curiosity, is stereochemistry. The initial cyanide addition creates a new stereocenter at the former carbonyl carbon. Enzymes called hydroxynitrile lyases (oxynitrilases) catalyze this addition with exquisite enantioselectivity. This single stereocenter is then carried through — untouched — into the acid or amino-alcohol product. That happens because none of the downstream steps (nitrile hydrolysis, nitrile reduction) touches the C–O bond. A cyanohydrin formed with 99% ee therefore delivers an α-hydroxy acid or amino alcohol with the same 99% ee. This is one of organic chemistry's cleanest examples of "chirality transfer by non-interference." You set the stereocenter once, early. You simply protect it by not disturbing it for the rest of the synthesis. This article works through the three canonical destinations — α-hydroxy acids, amino alcohols, and chain-extended carbonyl compounds. It gives full curved-arrow mechanisms for every hydrolysis, reduction, and decarboxylation step. It then follows with three worked syntheses. These are mandelic acid, an adrenaline-type amino alcohol, and a Corey–Seebach-flavored competition problem. Finally, it surveys where these reactions sit in industrial and medicinal chemistry today.
- A cyanohydrin R₂C(OH)CN is a masked α-hydroxy carbon nucleophile: one carbon added to a carbonyl, ready for three different downstream fates.
- Full hydrolysis of the C≡N (acid or base, via amide intermediate) gives an α-hydroxy carboxylic acid — the mandelic acid route.
- Partial reduction (DIBAL-H or Stephen conditions) stops at the imine, which hydrolyzes to an α-hydroxy aldehyde: net one-carbon chain extension of the original carbonyl.
- Full reduction (LiAlH₄ or H₂/catalyst) converts the nitrile all the way to a primary amine, giving a 1,2-amino alcohol directly.
- The stereocenter created at cyanide addition is never touched by any of these steps, so enantioselective cyanohydrin formation (via hydroxynitrile lyases) transfers ee cleanly into the final product.
- Every step is ordinary nucleophilic addition/tetrahedral-intermediate chemistry — no new reaction type is required, only careful bookkeeping of oxidation state.
- Key idea: cyanide addition converts a carbonyl into a protected, chain-extended stereocenter whose downstream chemistry (hydrolysis, reduction) never re-approaches the C–O bond, so whatever stereochemistry you set at addition, you keep.
The cyanohydrin carbon is set once and never revisited: every useful transformation downstream (hydrolysis to acid, reduction to amine, partial reduction to aldehyde) operates exclusively on the nitrile carbon, leaving the adjacent stereocenter — and its configuration — completely untouched.
Where this lives today
Mandelic acid, made industrially by exactly the hydrolysis of racemic or enantiopure mandelonitrile described in Example 1, is a large-scale chemical used both directly (as a topical dermatological agent and urinary antiseptic) and as a chiral resolving agent and a precursor for cephalosporin antibiotic side chains. The enantiopure (R)- or (S)-mandelic acid needed for these applications is obtained either by classical resolution of the racemic acid or, increasingly, by using an enantioselective hydroxynitrile lyase to set the stereocenter at the cyanohydrin stage before hydrolysis — a direct industrial application of the chirality-transfer principle developed in this article. Amino alcohols derived from cyanohydrin reduction sit at the heart of several drug classes. The 1,2-amino alcohol motif — a hydroxyl and an amine on adjacent carbons — appears in adrenergic agonists and antagonists (adrenaline, salbutamol-type bronchodilators, many β-blockers such as propranolol-class scaffolds built by related routes), and cyanohydrin-based chain extension followed by reduction is one of the standard synthetic disconnections taught for constructing this motif with control over the newly formed stereocenter. The chain-extension logic of Mechanism II — cyanohydrin formation followed by controlled partial reduction to give a homologated aldehyde bearing a new α-hydroxyl — is a recognized alternative to other one-carbon homologation strategies (such as the Arndt–Eistert reaction for carboxylic acids, or Wittig-based approaches), valued specifically because it installs a stereocenter with a defined, often enzymatically controlled configuration, which many alternative homologation methods do not offer as directly. Beyond isolated small-molecule targets, cyanohydrin hydrolysis chemistry (nitrile → amide → acid) is mechanistically identical to, and pedagogically inseparable from, the broader family of nitrile hydrolysis reactions used throughout industrial process chemistry — for example, in the manufacture of acrylamide from acrylonitrile, and in numerous carboxylic acid syntheses from readily available nitrile precursors — making the mechanism taught here a genuinely general tool, not a reaction confined to cyanohydrins alone.
Ask the tutor: How does the choice of hydroxynitrile lyase enzyme (R-selective vs. S-selective) actually discriminate between the two prochiral faces of an aldehyde during cyanohydrin formation?
The physical picture
Start from the structure everyone already knows how to make: cyanide adds to a ketone or aldehyde carbonyl, generating a tetrahedral alkoxide that is protonated to give the cyanohydrin.
The product has two functional handles: a tertiary (or secondary) alcohol, and a nitrile. The alcohol is chemically inert to everything that follows — it does not react with hydrolysis reagents, reducing agents, or acids under the conditions used, because it is simply an alcohol. The nitrile, by contrast, is a reactive electrophilic carbon (the carbon of C≡N is sp-hybridized, polarized C^{δ+}≡N^{δ-}, isoelectronic with CO and highly electrophilic toward nucleophilic addition). This electrophilicity is what lets us drive three different downstream reactions, depending on what nucleophile/reagent we throw at the nitrile carbon and how far we let the reaction proceed. The key oxidation-state ladder to keep in your head is this. A nitrile carbon (oxidation state of a carboxylic acid carbon) can be:
- Fully hydrolyzed (two sequential additions of water, through an amide) to a carboxylic acid — no change in oxidation state relative to nitrile, just oxygen for nitrogen.
- Partially reduced (one hydride equivalent) to an imine, which on aqueous workup hydrolyzes to an aldehyde — a two-electron reduction.
- Fully reduced (two hydride equivalents, or catalytic H₂) to a primary amine — a four-electron reduction.
Each destination shares the identical first mechanistic step: nucleophilic addition to the nitrile carbon, generating a metalated (or protonated) imine/imidate intermediate. The divergence happens entirely in what happens after that addition — hydrolytic workup versus further reduction versus tautomeric collapse.
The middle branch is the one students most often mismanage: DIBAL-H (diisobutylaluminum hydride) delivers exactly one hydride to the nitrile carbon at low temperature, generating a stable aluminum-chelated imine (metalated imine) that survives until aqueous acidic workup, at which point it hydrolyzes — imine plus water, not reduction again — to the aldehyde. If you instead used LiAlH₄ (a far more reactive, non-selective hydride source) at room temperature, the intermediate imine is itself reduced by a second hydride before you ever get to isolate it, and you land on the amine instead. The chemoselectivity is entirely about controlling how many hydride equivalents reach the substrate and at what temperature — the same logic that lets DIBAL-H stop ester reduction at the aldehyde stage.
Notice the orbital logic: a nitrile has two mutually perpendicular π systems around the linear C≡N axis, both empty of nucleophile character on carbon (both are π-LUMO-like in reactivity terms), so nucleophilic attack can occur from any angle perpendicular to the C≡N axis, exactly analogous to attack on the π of a carbonyl but with an sp→sp² rehybridization at carbon instead of sp²→sp³. The nitrogen lone pair generated is basic and will pick up a proton (or stay as a metal amide/aluminum-nitrogen bond) until workup.
The core ideas
Three governing relationships organize everything that follows.
1. Chain extension by exactly one carbon.
where: the aldehyde carbon count increases by one, and a new hydroxyl-bearing stereocenter appears adjacent to the new aldehyde.
2. Oxidation-state accounting for nitrile → acid vs. nitrile → amine.
where: hydrolysis replaces both π bonds of C≡N with C=O and departing ammonia, no net change in carbon oxidation level relative to a carboxylic acid; reduction instead adds hydrogen atoms, dropping the carbon two full oxidation levels (nitrile → imine → amine).
3. Stereochemical fidelity.
where: because no bond to the stereocenter (C–OH, C–R, C–R′) is ever broken in hydrolysis or reduction of the nitrile, the enantiomeric excess set at cyanide addition propagates unchanged into the final α-hydroxy acid or amino alcohol. The reading of these three lines is the entire strategic argument of this article: cyanohydrin chemistry lets you (1) add exactly one carbon, (2) choose the oxidation level of that new carbon post hoc by choosing hydrolysis vs. reduction conditions, and (3) do all of this without ever perturbing the stereocenter you worked hard to set enantioselectively at the addition step.
Goal. Convert R₂C(OH)–C≡N into R₂C(OH)–COOH under aqueous acid, showing every proton transfer and every tetrahedral intermediate, and explain why the reaction proceeds all the way to the carboxylic acid rather than stopping at the amide.
Step 1 (protonation of the nitrile nitrogen). Aqueous H₃O⁺ protonates the nitrile nitrogen lone pair, generating a nitrilium-type cation, R₂C(OH)–C≡N–H⁺. Protonation dramatically increases the electrophilicity of the nitrile carbon: the positive charge is delocalized between N and C by resonance,
with substantial positive character sitting on carbon in the second resonance structure, making it a much better electrophile than the neutral nitrile.
Step 2 (nucleophilic addition of water). A water molecule attacks the activated nitrile carbon with a lone pair, forming a new C–O bond and pushing electron density into the C=N π bond, which collects on nitrogen. The result, after a proton transfer from the incipient oxocarbenium-like oxygen to a nearby water (acting as base) and back-protonation of nitrogen, is the neutral imidic acid tautomer:
This is exactly analogous to hydration of a protonated carbonyl — the same nucleophilic-addition-to-an-electrophilic-π-system logic as Cyanohydrin Formation itself, just with water instead of cyanide as nucleophile, and a nitrile instead of a carbonyl as electrophile.
Step 3 (tautomerization: imidic acid → amide). The imidic acid tautomerizes to the far more stable amide by a formal 1,3-proton shift (mechanistically, two sequential acid/base proton transfers via solvent water): the O–H proton is removed and an N–H proton is installed, converting C(OH)=NH into C(=O)–NH₂. This tautomerization is strongly downhill — amide C=O is much more stable than the corresponding imidic acid tautomer, by direct analogy to keto–enol tautomerism (amide-imidic acid equilibrium favors amide by roughly 20 kcal/mol, comparable to keto-enol energetics but larger because of amide resonance).
Step 4 (protonation of the amide oxygen). The amide carbonyl oxygen, though far less basic than an amine, is protonated under the strongly acidic, often forcing conditions used (typically aqueous H₂SO₄ or HCl at reflux). Protonation converts the amide carbonyl into a much better electrophile, exactly as in ester/amide hydrolysis mechanisms.
Step 5 (second nucleophilic addition of water, tetrahedral intermediate). A second water molecule adds to the protonated amide carbonyl carbon, generating a tetrahedral intermediate bearing OH, OH₂⁺ (protonated original amide oxygen becomes an -OH after proton shuffling), and NH₂:
Step 6 (collapse of the tetrahedral intermediate, expulsion of ammonia). Proton transfer converts the amine nitrogen into a protonated ammonium leaving group (–NH₃⁺), which is a vastly better leaving group than –NH₂⁻. The tetrahedral intermediate collapses, re-forming the C=O π bond and expelling ammonia (as NH₄⁺ under the acidic conditions):
Result. The nitrile carbon has been carried through nitrilium → imidic acid → amide → protonated amide → tetrahedral intermediate → carboxylic acid, using only nucleophilic additions of water and proton transfers — the identical mechanistic toolkit as ester and amide hydrolysis. Crucially, at no point does any step touch the R₂C(OH) stereocenter; the entire six-step sequence occurs at the nitrile-derived carbon, three bonds away.
Go deeper · Derivation B · 2 minMechanism II — DIBAL-H partial reduction, then hydrolysis, to give the chain-extended aldehyde
Goal. Show why DIBAL-H at −78 °C delivers exactly one hydride to the nitrile and stops there, and why aqueous acidic workup converts the resulting metalated imine into an aldehyde rather than reducing further.
Step 1 (coordination of DIBAL-H aluminum to the nitrile nitrogen). DIBAL-H, iBu₂Al–H, is a Lewis acidic aluminum center bearing one hydridic Al–H bond. The nitrile nitrogen lone pair coordinates to the empty p-orbital on aluminum, forming a Lewis acid–base adduct:
This coordination polarizes and activates the nitrile carbon toward intramolecular hydride delivery, exactly as it activates an ester carbonyl in the analogous DIBAL-H ester-to-aldehyde reduction.
Step 2 (intramolecular hydride delivery to the nitrile carbon). The activated hydride on aluminum is delivered through a four-centered transition state to the adjacent, now more electrophilic, nitrile carbon. The C≡N π bond accepts the hydride and collapses to a C=N π bond, with the developing negative charge on nitrogen stabilized immediately by coordination to the aluminum center:
This is the crucial metalated imine (aza-enolate/aluminum amide) intermediate. It is thermodynamically and kinetically stable at −78 °C: the nitrogen lone pair is tied up in a strong, tetrahedral Al–N dative bond, which removes the nitrogen's ability to act as a nucleophile or leaving group, and the low temperature keeps the second equivalent of DIBAL-H (present in excess in some protocols) from reacting further before workup.
Step 3 (why no second hydride is delivered). In LiAlH₄ (highly reactive, "naked" hydride, high effective negative charge on H), an imine intermediate is itself electrophilic enough to be attacked by a second hydride at any accessible temperature, driving reduction all the way to the amine. DIBAL-H, in contrast, is a bulkier, less reactive hydride donor whose aluminum center is now coordinatively saturated by the newly formed aluminum-nitrogen bond in the intermediate — there is no accessible open coordination site left for a second hydride-delivery event at low temperature. The reaction is quenched (aqueous acid added) before warming can allow ligand exchange and a second reduction.
Step 4 (aqueous acidic hydrolysis of the metalated imine). Upon addition of dilute aqueous acid at low temperature, water protonates the aluminum-bound nitrogen and hydrolyzes the aluminum-nitrogen and aluminum-oxygen bonds, releasing the free imine:
Step 5 (imine hydrolysis via nucleophilic addition of water and loss of ammonia). The free imine, protonated at nitrogen to give an iminium ion, is attacked by water at carbon (the identical nucleophilic-addition-to-a-protonated-π-system logic used throughout this article), generating a hemiaminal, which collapses with loss of ammonia to give the aldehyde:
Result. Net two-electron (one hydride) reduction of nitrile to imine, then simple hydrolysis (not reduction) of the imine to the aldehyde — chemically the same imine hydrolysis that terminates reductive amination reactions. The original R₂C(OH) stereocenter is, again, never approached by any reagent in this five-step sequence.
Worked examples
Example 1 — Foundational: (±)-Mandelic acid from benzaldehyde. Benzaldehyde is treated with NaCN/H₂SO₄ (generating HCN in situ, buffered to control the equilibrium) to give the cyanohydrin, mandelonitrile:
The addition is reversible and unselective without a catalyst, giving racemic mandelonitrile — a mixture of (R) and (S) at the new stereocenter, each 50%. Refluxing mandelonitrile in 6 M HCl for several hours drives Mechanism I to completion. Track the atoms explicitly. Protonation of the nitrile nitrogen gives the nitrilium ion; water adds to carbon; tautomerization gives the amide, Ph–CH(OH)–C(=O)–NH₂ (mandelamide); protonation of the amide oxygen, a second water addition, and loss of ammonium ion gives mandelic acid:
Because the cyanohydrin was racemic, the product is (±)-mandelic acid. This is exactly the reaction run industrially (with resolution or enzymatic asymmetric cyanation upstream) to manufacture mandelic acid, an intermediate for cephalosporin antibiotics and a urinary antiseptic in its own right. Limiting-case check: if hydrolysis were stopped at the amide stage (milder conditions, shorter time — a common trap discussed below), the isolated product would be mandelamide, Ph–CH(OH)–CONH₂, not the acid; heating in acid for a shorter time or under basic conditions (which hydrolyze amides far more slowly than nitriles hydrolyze to amides) reliably isolates the amide as the kinetic product, confirming the two-step nature of Mechanism I.
Example 2 — Applied: an amino-alcohol synthesis toward a phenylpropanolamine-type skeleton. Suppose the target is 2-amino-1-phenylpropan-1-ol from propiophenone-type logic, illustrating chain-extension-plus-amination strategy used in adrenaline/ephedrine-class syntheses. Start from acetophenone-derived cyanohydrin chemistry adapted to an aldehyde partner; here we use phenylacetaldehyde as the concrete, asymmetric (non-round) example so every stereocenter and every hydride count is explicit. Phenylacetaldehyde, Ph–CH₂–CHO, is treated with HCN/base catalyst (mild, e.g. catalytic KCN buffered near neutral pH, favoring the reversible addition without epimerizing any existing stereocenters) to give the cyanohydrin:
This creates one new stereocenter at the former carbonyl carbon (now CH(OH)CN). If an (S)-selective hydroxynitrile lyase is used instead of achiral base catalysis, the product is obtained in high ee, say 96% ee favoring the (S)-cyanohydrin — a number entirely set by the chiral catalyst, unrelated to any subsequent step. Full reduction with excess LiAlH₄ (four electrons, two hydride equivalents, no attempt at partial reduction here) reduces the nitrile all the way through the imine intermediate (Mechanism II, Step 1–2 logic) to the primary amine without isolating the imine, because LiAlH₄'s "naked", highly reactive hydride reduces the intermediate imine before it can be isolated or hydrolyzed:
The product, 3-phenyl-1-amino-2-propanol (a 1,3-relationship counting the amine and the original phenyl ring, and a genuine 1,2-amino alcohol counting the new C–OH and C–NH₂ carbons themselves) retains the stereocenter set at cyanide addition — if the cyanohydrin was 96% ee (S), the amino alcohol is 96% ee at the corresponding carbon, because LiAlH₄ reduction of a nitrile several bonds away from that stereocenter cannot invert, racemize, or otherwise touch a C–O or C–H bond it never approaches. Limiting-case check: if racemic base-catalyzed cyanohydrin formation were used instead of the enzyme (0% ee, a racemic 50:50 mixture), the final amino alcohol would likewise be racemic (0% ee) — confirming that the reduction step contributes exactly zero net enantioselectivity of its own; all stereochemical information is set solely at the cyanide-addition step, and the ee value is transmitted, never generated or degraded, downstream.
Example 3 — Exam/competition level: chain-shortening trap and the decarboxylative pathway. A classic exam question: cyclohexanone cyanohydrin, 1-hydroxycyclohexanecarbonitrile, is hydrolyzed exhaustively in hot aqueous H₂SO₄, and the α-hydroxy acid product is then heated strongly with catalytic acid. What is the final isolated product, and why? First, apply Mechanism I unchanged: cyclohexanone + HCN gives 1-hydroxycyclohexane-1-carbonitrile, and hot aqueous acid hydrolyzes the nitrile through the amide to 1-hydroxycyclohexanecarboxylic acid:
This α-hydroxy acid is a tertiary alcohol directly adjacent to a carboxylic acid — a substitution pattern primed for acid-catalyzed dehydration-decarboxylation. Under continued strong heating with acid catalyst, the tertiary C–OH is protonated, water leaves to generate a tertiary carbocation stabilized by the adjacent carboxylic acid's inductive/resonance donation (through the developing conjugation, this is closely related to a retro-ene/decarboxylation pathway for β-hydroxy and α-hydroxy acids); the net classical outcome for a tertiary α-hydroxy acid under vigorous acid and heat is loss of the elements of formic acid or, more commonly taught at this level, simple acid-catalyzed dehydration of the tertiary alcohol to give an α,β-unsaturated carboxylic acid, cyclohex-1-ene-1-carboxylic acid, via standard E1 loss of water from the protonated tertiary alcohol (carbocation formation, then loss of a β-hydrogen from the ring to re-form a π bond conjugated with the carboxyl group):
The trap in this question is assuming that "heating an α-hydroxy acid with acid" automatically means decarboxylation (loss of CO₂); that pathway (retro-ene, or via a cyclic transition state) is well documented for β-keto acids and certain α-hydroxy acids under specific forcing thermal conditions, but the dominant, kinetically far more accessible pathway for a tertiary α-hydroxy acid under standard mineral-acid dehydration conditions is simple E1 dehydration to the conjugated acrylic-acid-type product, exactly as it would be for any tertiary alcohol. Recognizing which pathway dominates requires comparing the ordinary, low-barrier E1 dehydration of a tertiary carbinol against the higher-barrier, six-membered-transition-state decarboxylation pathway — under standard aqueous mineral acid and reflux, dehydration wins.
Limiting cases, scope & common traps
| Situation | What actually happens | Why |
|---|---|---|
| Hydrolysis stopped early (mild base, short time, or dilute acid, room temperature) | Isolate the amide, not the acid | Nitrile → amide is markedly faster than amide → acid; the second hydrolysis needs more forcing acid/base and heat |
| DIBAL-H used at room temperature instead of −78 °C | Over-reduction to the amine, or a messy mixture | The aluminum-imine intermediate is only kinetically stable in the cold; warming allows ligand exchange and a second hydride delivery |
| LiAlH₄ used when an aldehyde is wanted | Get the amine, not the aldehyde | LiAlH₄'s reactive hydride reduces the intermediate imine before workup; there is no stable intermediate to intercept |
| Racemic cyanide addition (no chiral catalyst) upstream | Final acid/amine is racemic, regardless of how carefully downstream steps are run | Stereochemistry is set only at the C–CN bond-forming step; nothing downstream can create ee that wasn't already there |
| α,α-disubstituted (fully substituted, no α-H) cyanohydrin heated strongly in acid | Simple dehydration pathways to alkene-acids are blocked if no β-H is anti-periplanar/available; may instead survive or decarboxylate via alternate pathway | Structure-dependent — always check for available β-hydrogens before assuming E1 dehydration |
| Attempting cyanohydrin formation on a very hindered or highly electron-rich carbonyl | Equilibrium lies toward free carbonyl + HCN; poor conversion | The addition equilibrium constant is sensitive to steric bulk and to conjugative stabilization of the carbonyl (as covered under Cyanohydrin Formation) |
The chemist behind it
The modern strategic use of cyanohydrins as chiral, chain-extending building blocks owes an enormous debt to the discovery and application of hydroxynitrile lyases (oxynitrilases), enzymes first studied in the context of cyanogenesis in plants — many plants (almond, cassava, sorghum) store cyanogenic glycosides and release HCN defensively via exactly the reverse of cyanohydrin formation, catalyzed by these same enzyme families run in the biosynthetic direction. Chemists later realized that running these enzymes in the synthetic direction — enzyme plus HCN plus aldehyde, under carefully controlled pH — gives access to cyanohydrins in very high enantiomeric excess, a discovery that transformed cyanohydrin chemistry from a racemic curiosity into an asymmetric synthesis workhorse.
Origins · 1 minFranz Effenberger and how this was discovered
Effenberger's group carried out extensive systematic studies through the 1980s–2000s on hydroxynitrile lyases from almond (Prunus amygdalus) and other plant sources, mapping which aldehydes and ketones each enzyme would accept and with what enantioselectivity, and demonstrating that the resulting chiral cyanohydrins could be carried forward — by exactly the hydrolysis and reduction mechanisms detailed above — into enantiopure α-hydroxy acids and amino alcohols on preparative and even industrial scale. This body of work turned enzymatic cyanohydrin synthesis into a standard entry in the asymmetric-synthesis toolkit, sitting alongside Sharpless epoxidation and enzymatic kinetic resolution as a reliable way to install a single, well-defined stereocenter early in a synthesis and then propagate it, untouched, through several further steps.
Why does cyanide, specifically, make such a good one-carbon nucleophile for this whole strategy?
Cyanide is small, a strong nucleophile despite its resonance stabilization, and — crucially — the nitrile group it installs is a compact, chemically robust placeholder that survives many reaction conditions unreacted until you deliberately activate it. Few other one-carbon nucleophiles offer this combination: a Grignard-derived cyanide surrogate would be far too reactive and non-selective, while a stabilized carbanion equivalent (say, a malonate anion) adds a full two-carbon unit rather than one. The nitrile's linear, electronically simple C≡N unit is also uniquely suited to controlled, stepwise activation — first as an electrophile for hydrolysis or reduction — which is exactly why it works so well as the "extra carbon" in chain-extension strategies.
Can you form a cyanohydrin from a carboxylic acid derivative instead of an aldehyde or ketone?
No — cyanohydrin formation specifically requires nucleophilic addition of cyanide to a genuine carbonyl π* system where addition is thermodynamically reasonably accessible (aldehydes and ketones). Esters, amides, and carboxylic acids are far less electrophilic at carbon (their carbonyl is stabilized by resonance donation from an adjacent oxygen or nitrogen lone pair) and, even where cyanide could in principle add, the tetrahedral intermediate would simply expel the original leaving group (alkoxide, amide anion) rather than persist as a stable cyanohydrin analog. This is exactly the same electronic reasoning that makes esters and amides resistant to direct Grignard mono-addition compared to aldehydes and ketones.
Does the choice of acid or base for hydrolysis change the final product, or just the rate?
Both acid- and base-catalyzed nitrile hydrolysis ultimately deliver the carboxylic acid (or its conjugate base, a carboxylate, under basic conditions, requiring a final acidic workup to protonate it) via the same overall bond changes, but the detailed mechanism differs: acid-catalyzed hydrolysis proceeds through nitrilium and imidic acid/amide intermediates as shown in Mechanism I, while base-catalyzed hydrolysis proceeds through direct hydroxide addition to the nitrile carbon (forming an amidate anion directly, without a protonation-activation step) and then a second hydroxide addition to the resulting amide. The base route is often preferred industrially because it avoids strongly acidic waste streams, but both routes pass through an isolable amide intermediate if conditions are insufficiently forcing, and both leave the adjacent stereocenter completely untouched.
Why is the stereocenter in a cyanohydrin created at all — what determines whether you get R or S?
The stereochemical outcome of cyanide addition to a prochiral carbonyl is governed entirely by which face of the planar sp² carbonyl carbon the nucleophile approaches, following Bürgi–Dunitz trajectory considerations for the uncatalyzed reaction (giving racemic product from an achiral carbonyl, since both faces are equally accessible to free cyanide in solution) or by the shape of a chiral catalyst's active site (in enzymatic or chiral-catalyst-mediated additions), which selectively shields one face and forces nucleophilic attack from the other. This entire question of facial selectivity belongs to the mechanism and stereochemistry of the addition step itself, covered under Cyanohydrin Formation; this article's contribution is simply to show that whatever selectivity is achieved there survives, completely unaltered, through every downstream transformation.
What happens if you try to reduce the cyanohydrin's nitrile with catalytic H₂ over a metal catalyst instead of a hydride reagent?
Catalytic hydrogenation (H₂, Pd/C, Pt, Raney nickel, or similar) can also reduce nitriles to primary amines, delivering two equivalents of H₂ across the triple bond stepwise on the catalyst surface via imine-like adsorbed intermediates, arriving at the same primary amine product as LiAlH₄ reduction. The main practical complication is that primary amine products can undergo secondary condensation with any remaining imine intermediate to give a secondary amine byproduct (a classic side reaction in nitrile hydrogenations), usually suppressed by running the hydrogenation in the presence of ammonia or acid to protonate the primary amine as it forms, preventing it from acting as a nucleophile toward the imine. Mechanistically, though, the destination — full four-electron reduction to the 1,2-amino alcohol — is identical to the LiAlH₄ pathway described in Mechanism II's logic, just carried out heterogeneously.
Is there a limiting case where the cyanohydrin chemistry described here simply fails or gives no useful product?
Yes — highly hindered or strongly deactivated carbonyls (very electron-rich aryl ketones, or extremely bulky dialkyl ketones) may never form a useful concentration of cyanohydrin in the first place, because the addition equilibrium (governed by the thermodynamics discussed under Cyanohydrin Formation) lies too far toward starting materials; in that limiting case, none of the downstream hydrolysis or reduction chemistry in this article has a substrate to act on at all, and an alternative one-carbon homologation strategy (such as a Wittig or an Arndt–Eistert-type approach) would be chosen instead.