Carboxylation of Grignard Reagents and Nitrile Hydrolysis

A flask of dry ether, a coil of magnesium turnings, and a stream of bubbling carbon dioxide gas. From this simple setup emerges one of the most reliable ways to build a carboxylic acid. It attaches onto a carbon skeleton that previously had none. Push solid CO2 ("dry ice") into a Grignard solution. After aqueous acidic workup, you collect a carboxylic acid with one more carbon than you started with. It looks almost too easy. Yet the mechanism is a clean, textbook example of a nucleophile attacking an electrophilic carbon. It is precisely because the mechanism is so simple that the reaction is so trustworthy. This matters because carboxylic acids are everywhere in target molecules. They appear in pharmaceuticals, fatty acids, polymer monomers, and agrochemicals. Yet very few simple starting materials already carry a $-\text{CO}_2\text{H}$ group in the position you need it. Two complementary strategies solve this "chain-lengthening" problem from opposite directions. The Grignard reagents route builds the acid carbon de novo. It converts an alkyl or aryl halide, via its organometallic, directly into an acid with one extra carbon. The nitrile route instead installs a cyanide carbon first. This typically happens via $S_N2$ displacement of a halide by $^{-}\text{CN}$ . Only later, in a separate hydrolysis step, is that nitrile unmasked as a carboxylic acid. Both routes achieve the same net transformation — halide to acid, +1 carbon. But they diverge sharply in what functional groups they tolerate. They also differ in what mechanistic machinery does the work. The big picture is this: a carboxylic acid carbon is at the highest oxidation state before you reach $\text{CO}_2$ itself. So building one always requires one of two approaches. Either (a) capture carbon already at that oxidation level, as $\text{CO}_2$ does. This means feeding a carbanion equivalent into an electrophilic carbon. Or (b) start from a carbon at a lower or comparable oxidation state. A nitrile carbon works, formally between an imine and an acid. Then oxidize or hydrate it up. The Grignard– $\text{CO}_2$ reaction is example (a). A strongly nucleophilic carbanion equivalent attacks the electrophilic carbon of $\text{CO}_2$ . Nitrile hydrolysis is example (b). Water, activated by acid or base, adds across the polarized $\text{C}\equiv\text{N}$ triple bond in two sequential additions. It adds first to an amide, then to the acid. This article works through both mechanisms arrow-by-arrow. It covers the tetrahedral-like addition of a carbanion to $\text{CO}_2$ . It also covers the two-stage addition–addition mechanism of nitrile hydrolysis under both acidic and basic conditions. It shows how chemists choose between them based on what other functional groups must survive. It also shows how far down the carbon chain the new acid group needs to sit.

Key takeaways

$\text{RMgX} + \text{CO}_2 \to \text{RCO}_2\text{MgX} \xrightarrow{\text{H}_3\text{O}^+} \text{RCO}_2\text{H}$ adds exactly one carbon at the site of the original halide.
The Grignard carbanion attacks the electrophilic carbon of $\text{CO}_2$ in a single concerted nucleophilic addition — no tetrahedral intermediate collapse needed, since carboxylate is the resting state.
Solid $\text{CO}_2$ (dry ice) is added to the Grignard, not the reverse, to keep $\text{CO}_2$ in large excess and suppress the ketone-forming side reaction (a second $\text{RMgX}$ adding to the carboxylate).
$\text{R–X} + \text{NaCN} \to \text{R–CN}$ (via $S_N2$ ) followed by hydrolysis, $\text{R–CN} \to \text{RCONH}_2 \to \text{RCO}_2\text{H}$ , also adds one carbon but tolerates groups (esters, alkenes) that Grignard reagents destroy.
Acidic hydrolysis of a nitrile proceeds by protonation of nitrogen, water addition to the resulting nitrilium ion, tautomerization to the amide, then a second acid-catalyzed hydrolysis of the amide to the acid.
Basic hydrolysis proceeds by direct hydroxide addition to the nitrile carbon, giving an amidate that protonates to the amide, followed by classical addition–elimination hydrolysis of the amide via a tetrahedral intermediate.
Key idea: both routes lengthen a chain by exactly one carbon, but they attack the problem from opposite ends of the carbon's oxidation state — Grignard captures $\text{CO}_2$ directly as the acid carbon, nitrile hydrolysis first installs a lower-oxidation-state $\text{C}\equiv\text{N}$ carbon and hydrates it up in two discrete nucleophilic-addition steps.

The core insight

A carboxylic acid carbon can be built two ways: capture it already fully oxidized as electrophilic $\text{CO}_2$ using a carbanion nucleophile (Grignard route), or install it as a nitrile carbon via $S_N2$ and then hydrate the $\text{C}\equiv\text{N}$ across two sequential water additions, first to an amide and then to the acid (nitrile route). Both add one carbon; the choice between them is dictated by what other functionality in the molecule must survive strongly basic organometallic conditions versus aqueous acid or base at reflux.

Where this lives today

The Grignard carboxylation is still a first-choice laboratory method whenever a target molecule needs a carboxylic acid installed at a specific, otherwise inaccessible position on a carbon skeleton, particularly in small-scale and academic synthesis where its operational simplicity (dry ice is cheap, and the reaction requires no exotic catalyst) outweighs its intolerance of other functional groups. Industrially, the related but distinct Kolbe–Schmitt carboxylation (aromatic Grignard-like chemistry adapted to phenoxide nucleophiles and $\text{CO}_2$ under pressure) is the actual manufacturing route to salicylic acid, the precursor of aspirin — a direct industrial descendant of the same fundamental idea that a sufficiently nucleophilic carbanion (or carbanion-like aryloxide) will capture electrophilic $\text{CO}_2$ to build a new carboxylic acid carbon. Nitrile hydrolysis, meanwhile, has found a second life in green and biocatalytic chemistry: nitrile hydratase and nitrilase enzymes, isolated from soil bacteria, catalyze exactly the same net transformation — nitrile to amide, or nitrile directly to acid — under extraordinarily mild aqueous conditions (room temperature, neutral pH) compared with the reflux, strong-acid/strong-base conditions of classical laboratory hydrolysis. This biocatalytic route is used industrially at multi-tonne scale to manufacture acrylamide (from acrylonitrile) and nicotinamide (from 3-cyanopyridine), avoiding the harsh conditions and unwanted polymerization side reactions that classical acid hydrolysis of acrylonitrile would otherwise cause. In pharmaceutical synthesis, the nitrile route is routinely favored over the Grignard route specifically because so many drug intermediates already carry esters, ketones, or other Grignard-incompatible functionality; a cyanide displacement, run under comparatively mild conditions, followed by a carefully controlled partial or full hydrolysis, threads that needle far more reliably. Both reactions also underpin the logic taught for planning Methods for Preparing Carboxylic Acids more broadly, and understanding the amide-intermediate mechanism here is the direct conceptual bridge into the broader study of Structure and Reactivity of Nitriles as an electrophilic functional group in its own right.

Ask the tutor: How does the choice between the Grignard–CO2 route and the nitrile hydrolysis route change if the target molecule also contains a stereocenter that must be preserved?

The physical picture

Carbon dioxide is a linear, doubly electrophilic molecule: $\text{O}=\text{C}=\text{O}$ , with the central carbon flanked by two oxygens pulling electron density away through both the $\sigma$ framework and the two orthogonal $\pi$ systems. The central carbon sits at an unusually high effective positive charge for a carbon with no formal charge, precisely because it is doubly $\pi$ -bonded to the most electronegative common element. A Grignard reagent, by contrast, is the archetypal carbon nucleophile: the $\text{C–Mg}$ bond is strongly polarized toward carbon (electronegativity of C ≈ 2.5, Mg ≈ 1.3), so the alkyl or aryl carbon carries substantial carbanion character. Bring the two together and the outcome is almost thermodynamically inevitable: the carbanion-like carbon attacks the electrophilic carbon of $\text{CO}_2$ , one of the two π bonds breaks heterolytically, and the electrons land on an oxygen, generating a magnesium carboxylate salt. No proton transfer, no leaving group, no tetrahedral intermediate that has to collapse — carboxylate is the resting state, stabilized by two equivalent resonance structures spreading the negative charge over both oxygens. Nitrile hydrolysis works on a very different electrophile. The nitrile carbon is $sp$ -hybridized, triple-bonded to nitrogen through one $\sigma$ and two orthogonal $\pi$ bonds, and — like a carbonyl but more so — polarized $\text{C}^{\delta+}\!\equiv\!\text{N}^{\delta-}$ . Nitrogen's lone pair sits in an $sp$ orbital pointing away from carbon, largely uninvolved in resonance donation back into the ring of electrophilicity; this makes the nitrile carbon distinctly attackable by nucleophiles, though less so than a ketone or ester carbonyl because the linear, only mildly polarized triple bond offers less charge separation than a $\pi$ bond to oxygen. Water — poor nucleophile, needing activation either way — adds to this carbon twice in sequence, each addition mechanistically resembling carbonyl-addition chemistry: first water (activated by acid, or as hydroxide in base) attacks $\text{C}\equiv\text{N}$ to give, after tautomerization, an amide; then the same nucleophile attacks the resulting amide carbonyl a second time, expelling ammonia (or its conjugate base) to give the acid. Two sequential additions, each borrowing directly from the machinery of carbonyl chemistry, are what convert a $\text{C}\equiv\text{N}$ carbon all the way to a $-\text{CO}_2\text{H}$ carbon.

Figure

Carbanion attack on the electrophilic carbon of CO₂: one π bond breaks, electrons localize on oxygen

Figure

Nitrile hydrolysis: two sequential water additions, nitrile → amide → acid

Figure

Why excess CO₂ matters: carboxylate is a weak enough electrophile that a second RMgX rarely adds — unless RMgX is added slowly to a limited CO₂ supply

The core ideas

The overall transformations are summarized by two net equations, each hiding a distinct mechanism underneath.

\text{R–MgX} + \text{O=C=O} \;\longrightarrow\; \text{R–CO}_2^-\,\text{MgX}^+ \;\xrightarrow{\;\text{H}_3\text{O}^+\;}\; \text{R–CO}_2\text{H} + \text{MgX(OH)}

where: R is alkyl, aryl, or vinyl (any group that forms a stable, non-eliminating Grignard reagent); the first arrow is a single nucleophilic addition step; the second is a simple Brønsted acid–base protonation of the carboxylate on workup.

\text{R–X} + \,^{-}\text{CN} \;\xrightarrow{S_N2}\; \text{R–CN} \;\xrightarrow[\text{H}_2\text{O}]{\text{H}_3\text{O}^+\text{ or OH}^-,\ \Delta}\; \text{R–CO}_2\text{H} + \text{NH}_3\ (\text{or NH}_4^+)

where: the first step is a backside-attack substitution requiring an unhindered primary or secondary alkyl halide (aryl and vinyl halides do not undergo this $S_N2$ under normal conditions); the second step is a two-stage hydration of the nitrile, passing through the isolable amide. A useful bookkeeping relationship tracks the carbon count explicitly:

n_{\text{acid}} = n_{\text{halide}} + 1

where: $n_{\text{halide}}$ is the number of carbons in the starting alkyl/aryl halide skeleton (including the carbon bearing the halogen), and the "+1" is the new carbon supplied either by $\text{CO}_2$ or by cyanide. The regiochemical outcome is fixed by where the new carbon attaches: directly onto the carbon that formerly bore the leaving group (or the metal), in both routes, since both the Grignard carbon and the cyanide carbon substitute at that same position.

Derivation AMechanism 1 — RMgX carboxylation with CO₂, from carbanion to acid

Goal. Show, arrow by arrow, how a Grignard reagent converts $\text{CO}_2$ into a magnesium carboxylate, and why aqueous acidic workup — not aqueous base — is used to liberate the free acid.

Step 1 — Generation of the nucleophile. The Grignard reagent $\text{R–MgX}$ (formed separately from $\text{R–X}$ and $\text{Mg}^0$ in dry ether or THF, via a single-electron-transfer radical mechanism at the metal surface) exists in solution as an aggregate, but for mechanistic purposes it behaves as a source of $\text{R}^{\delta-}$ tightly ion-paired to $\text{Mg}^{2+}\text{X}^-$ . The C–Mg bond is highly polarized: carbon bears substantial negative charge because magnesium (electronegativity ≈ 1.3) donates electron density into the C–Mg $\sigma$ bond far more than carbon does.

Step 2 — Orbital alignment with CO2. Carbon dioxide's central carbon is $sp$ -hybridized with two orthogonal, empty-in-character $\pi^*$ orbitals (formally: two filled $\pi$ bonding orbitals to the two oxygens, but the LUMO is a $\pi^*$ combination heavily weighted on the central carbon due to the electronegativity difference). The Grignard carbanion's filled orbital — essentially an $sp^3$ lone pair-like orbital on R — aligns with one of these $\pi^*$ orbitals of $\text{CO}_2$ , approaching along the axis perpendicular to that particular C=O π system.

Step 3 — The bond-forming step. A single curved arrow moves the electron pair from the C–Mg bond (or, viewed as the carbanion lone pair) onto the central carbon of $\text{CO}_2$ , forming the new C–C bond. Simultaneously, one full C=O π bond of $\text{CO}_2$ breaks heterolytically, and that electron pair moves entirely onto the corresponding oxygen. This step is concerted: bond-making and bond-breaking happen in the same electronic event, exactly analogous to a Grignard's addition to a ketone carbonyl, except here the electrophile is $\text{CO}_2$ rather than $\text{R}_2\text{C}=\text{O}$ .

Step 4 — Formation of the carboxylate. The immediate product is a magnesium carboxylate, $\text{R–CO}_2^{-}\ \text{MgX}^+$ , in which the new carbon bears one C–R bond, one remaining C=O π bond, and one C–O⁻ bond — but because both oxygens are equivalent by resonance (the negative charge delocalizes over both C–O bonds, each with roughly 1.5 bond order), this is best drawn as a resonance hybrid, not a static single structure. This resonance delocalization is why the carboxylate, unlike a simple alkoxide, is a comparatively poor electrophile: the negative charge is smeared over two oxygens instead of concentrated on one, making a second nucleophilic attack by excess $\text{RMgX}$ much slower than the first addition to $\text{CO}_2$ itself.

Step 5 — Aqueous acidic workup. Treating the reaction mixture with dilute aqueous acid (commonly $\text{H}_3\text{O}^+$ , e.g. dilute HCl or $\text{H}_2\text{SO}_4$ ) protonates the carboxylate oxygen. This is a simple, fast, thermodynamically favorable proton transfer — the carboxylate ( $\text{p}K_a$ of the conjugate acid ≈ 4–5) is a far stronger base than water, so protonation is essentially complete. The magnesium salt of the halide, $\text{MgX(OH)}$ or $\text{Mg(OH)X}$ , dissolves into the aqueous layer, and the free carboxylic acid is isolated, typically by extraction, from the organic layer (or by direct crystallization if sufficiently insoluble in water).

Why acid, not base, for the workup. If aqueous base were used instead, the product carboxylic acid would simply be deprotonated back to the water-soluble carboxylate salt, which cannot then be separated from the aqueous magnesium salts by extraction. Acidic workup ensures the acid is protonated (neutral, extractable into an organic solvent) while the magnesium and halide ions remain in the aqueous phase as simple inorganic salts.

Consistency check — atom economy and carbon count. Starting from $\text{R–X}$ with $n$ carbons, forming $\text{R–MgX}$ preserves $n$ carbons; adding $\text{CO}_2$ supplies exactly one new carbon, giving $\text{R–CO}_2\text{H}$ with $n+1$ carbons — consistent with the bookkeeping relationship stated above.

Go deeper · Derivation B · 4 minMechanism 2 — Nitrile hydrolysis under acid and base, amide as the obligatory intermediate

Derivation BMechanism 2 — Nitrile hydrolysis under acid and base, amide as the obligatory intermediate

Goal. Derive, mechanistically, why nitrile hydrolysis to a carboxylic acid must pass through an isolable amide intermediate, under both acidic and basic aqueous conditions, and show the electron flow at each stage.

Step 1 (acidic pathway) — Protonation of nitrogen. In strong aqueous acid (e.g. concentrated HCl or $\text{H}_2\text{SO}_4$ , heated), the nitrile nitrogen — which bears a lone pair in an $sp$ orbital, not involved in the $\pi$ system — is protonated first. This generates a nitrilium ion, $\text{R–C}\equiv\text{N}^+\text{–H}$ , which is dramatically more electrophilic at carbon than the neutral nitrile, because the positive charge withdraws electron density directly through the $\pi$ system onto the very carbon that must be attacked next.

Step 2 (acidic) — Water addition to the nitrilium carbon. A water molecule, acting as a weak nucleophile despite protonation elsewhere in the medium, attacks the electrophilic carbon of the nitrilium ion. One curved arrow moves a lone pair from water's oxygen onto that carbon; simultaneously, one of the two π bonds of the (formally triple, now partially cationic) C≡N⁺ bond breaks, its electrons localizing onto nitrogen. The result, after loss of a proton from the newly formed oxocarbenium-like oxygen, is the imidic acid tautomer $\text{R–C(OH)=NH}$ — a species exactly analogous to an enol, but with nitrogen instead of one of the carbons.

Step 3 (acidic) — Tautomerization to the amide. The imidic acid, like an enol, is the minor tautomer; proton transfer (formally: protonation of nitrogen and deprotonation of oxygen, or a concerted acid/base-catalyzed shift) converts $\text{R–C(OH)=NH}$ into the far more stable primary amide $\text{R–C(=O)–NH}_2$ . This tautomerization is strongly downhill: the amide's C=O π bond is stabilized by resonance donation from the adjacent nitrogen lone pair (amide resonance), an energetic stabilization (~15–20 kcal/mol of resonance stabilization) that has no counterpart in the imidic acid tautomer.

Step 4 (acidic) — The amide is itself hydrolyzed, by the standard addition–elimination mechanism. Because the reaction medium is strongly acidic and hot, the amide does not survive; it is hydrolyzed further by the same four-step logic used for ester or amide hydrolysis generally: (i) protonation of the amide carbonyl oxygen, making the carbonyl carbon strongly electrophilic; (ii) nucleophilic attack by a second water molecule at that carbon, generating a tetrahedral intermediate bearing OH, OH2+, NH2, and R substituents; (iii) proton transfers within the tetrahedral intermediate to place a good leaving group (protonated nitrogen, as $-\text{NH}_3^+$ ) properly positioned anti to the departing electrons; (iv) collapse of the tetrahedral intermediate, expelling ammonia (which is immediately protonated to the ammonium ion under the strongly acidic conditions, driving the equilibrium irreversibly toward products) and regenerating the carbonyl, now as the neutral carboxylic acid after a final deprotonation.

Step 5 (basic pathway) — Direct hydroxide addition. Under basic conditions (aqueous NaOH, heat), no protonation of nitrogen occurs; instead, hydroxide itself — a much better nucleophile than neutral water — attacks the nitrile carbon directly. One curved arrow moves a hydroxide lone pair onto the nitrile carbon; one C≡N π bond breaks, its electrons localizing fully onto nitrogen as a negatively charged imidate anion, $\text{R–C(O}^-\text{)=NH}$ .

Step 6 (basic) — Protonation and tautomerization to the amide anion / amide. The imidate anion is protonated (by water, regenerating hydroxide catalytically) to give the same imidic acid tautomer seen in the acidic pathway, which tautomerizes to the primary amide exactly as before. Alternatively, the amidate can be viewed as directly tautomerizing to the resonance-stabilized amide anion before final protonation; either description converges on the same amide intermediate.

Step 7 (basic) — Base-catalyzed hydrolysis of the amide to the carboxylate. Hydroxide attacks the amide carbonyl carbon directly (no protonation needed, since hydroxide is nucleophilic enough on its own), forming a tetrahedral alkoxide intermediate bearing O⁻, OH, NH2, and R. This intermediate collapses by kicking out amide nitrogen as the strongly basic $^{-}\text{NH}_2$ leaving group — a poor leaving group in principle, but the reaction is driven forward because the initially formed carboxylic acid is instantly deprotonated by the departing $^{-}\text{NH}_2$ (itself becoming $\text{NH}_3$ ) in a fast, essentially irreversible proton transfer, converting the neutral acid to the resonance-stabilized, unreactive carboxylate and pulling the entire multi-step equilibrium to completion.

Step 8 (basic) — Final acidification. Because the basic hydrolysis stops at the carboxylate stage (not the free acid — no acid can survive in excess hydroxide), a separate acidic workup step (dilute aqueous HCl) is required afterward to protonate the carboxylate and liberate the neutral carboxylic acid, exactly as in the Grignard route's final step.

Consistency check — why the amide cannot be skipped. Both mechanisms funnel through the same electronic logic: a nitrile carbon, once attacked once by water/hydroxide, becomes an imidic acid that overwhelmingly prefers its amide tautomer due to amide resonance; only a second, independent nucleophilic attack — this time on the amide's carbonyl carbon, following the standard tetrahedral-intermediate addition–elimination pathway — completes the journey to the carboxylic acid. This is why controlled nitrile hydrolysis (mild conditions, low temperature, or careful stoichiometry) can be halted cleanly at the amide stage, a synthetically useful stopping point in its own right.

Worked examples

Example 1 (foundational) — Propanoic acid from ethyl bromide, two independent routes compared. Target: $\text{CH}_3\text{CH}_2\text{CO}_2\text{H}$ (propanoic acid, 3 carbons) from $\text{CH}_3\text{CH}_2\text{Br}$ (ethyl bromide, 2 carbons). Grignard route. Ethyl bromide reacts with magnesium turnings in dry diethyl ether under nitrogen (rigorously anhydrous — Grignard reagents are destroyed instantly by any trace of water or acidic protons, including those on terminal alkynes or on OH/NH groups) to form ethylmagnesium bromide, $\text{CH}_3\text{CH}_2\text{MgBr}$ . This solution is then poured onto crushed dry ice (solid $\text{CO}_2$ , large local excess to suppress double addition). The ethyl carbanion attacks the central carbon of $\text{CO}_2$ , breaking one π bond and forming magnesium propanoate, $\text{CH}_3\text{CH}_2\text{CO}_2\text{MgBr}$ . Acidic aqueous workup (dilute HCl) protonates this salt to give propanoic acid.

\text{CH}_3\text{CH}_2\text{Br} \xrightarrow{\text{Mg, Et}_2\text{O}} \text{CH}_3\text{CH}_2\text{MgBr} \xrightarrow{\text{(i) CO}_2\text{(s)}\text{, (ii) H}_3\text{O}^+} \text{CH}_3\text{CH}_2\text{CO}_2\text{H}

Nitrile route. Ethyl bromide is instead treated with sodium cyanide in a polar aprotic solvent (e.g. DMSO or DMF, which solvate $\text{Na}^+$ well but leave $^{-}\text{CN}$ poorly solvated and highly nucleophilic), displacing bromide via backside $S_N2$ attack to give propionitrile, $\text{CH}_3\text{CH}_2\text{CN}$ . Reflux with aqueous $\text{H}_2\text{SO}_4$ (or aqueous NaOH followed by acidic workup) hydrolyzes the nitrile through the amide, $\text{CH}_3\text{CH}_2\text{CONH}_2$ , to the acid.

\text{CH}_3\text{CH}_2\text{Br} \xrightarrow{\text{NaCN, DMSO}} \text{CH}_3\text{CH}_2\text{CN} \xrightarrow{\text{H}_3\text{O}^+, \Delta} \text{CH}_3\text{CH}_2\text{CO}_2\text{H}

Both routes give the same three-carbon acid. Carbon-count check: 2 (halide) + 1 (CO2 or CN carbon) = 3, matching propanoic acid. This symmetric-looking example is chosen deliberately simple to expose the two mechanisms cleanly before applying them to a case where the two routes are not interchangeable.

Example 2 (applied) — Chain-lengthening a hindered secondary halide, where only one route works. Suppose the target is 2-methylbutanoic acid, $\text{CH}_3\text{CH}_2\text{CH(CH}_3)\text{CO}_2\text{H}$ , starting from 2-bromobutane, $\text{CH}_3\text{CH}_2\text{CHBrCH}_3$ (a secondary alkyl bromide). Attempted nitrile route. Treating 2-bromobutane with $^{-}\text{CN}$ under $S_N2$ conditions is problematic: cyanide is a small, strongly basic, and only moderately polarizable nucleophile, and secondary alkyl halides bearing β-hydrogens are notoriously prone to E2 elimination competing with $S_N2$ substitution when a nucleophile is even mildly basic. In practice, secondary substrates give substantial alkene byproduct (2-butene, mixture of E/Z) alongside the desired 2-methylbutanenitrile, lowering the yield of usable nitrile and complicating purification. The reaction is not mechanistically forbidden, but it is synthetically unreliable for secondary substrates carrying accessible β-hydrogens. Grignard route. 2-Bromobutane forms the secondary Grignard reagent, $\text{CH}_3\text{CH}_2\text{CH(MgBr)CH}_3$ , cleanly with magnesium metal in ether — Grignard formation at the metal surface does not proceed through the same free-carbanion/base-mediated elimination pathway that plagues $S_N2$ with cyanide, so competing elimination is far less of a problem here (though it is not entirely absent for very hindered or highly branched systems). Carboxylation with solid $\text{CO}_2$ followed by acidic workup delivers 2-methylbutanoic acid directly and in good yield.

\text{CH}_3\text{CH}_2\text{CHBrCH}_3 \xrightarrow{\text{Mg, Et}_2\text{O}} \text{CH}_3\text{CH}_2\text{CH(MgBr)CH}_3 \xrightarrow{\text{(i) CO}_2\text{, (ii) H}_3\text{O}^+} \text{CH}_3\text{CH}_2\text{CH(CH}_3)\text{CO}_2\text{H}

This example demonstrates the practical decision rule: for secondary (and especially branched or hindered) alkyl halides, the Grignard– $\text{CO}_2$ route is generally more reliable than cyanide displacement, precisely because Grignard formation at a metal surface sidesteps the $S_N2$ -versus-E2 competition that a discrete anionic nucleophile like $^{-}\text{CN}$ cannot avoid.

Example 3 (exam/competition-level) — Multi-step synthesis with a protecting-group-style constraint, choosing the nitrile route because Grignard reagents are incompatible with an ester. Suppose the target is the diacid derivative $\text{HO}_2\text{C–CH}_2\text{CH}_2\text{CH}_2\text{–CO}_2\text{H}$ (glutaric acid, a 5-carbon diacid), and the only reasonably available starting material is methyl 4-bromobutanoate, $\text{BrCH}_2\text{CH}_2\text{CH}_2\text{CO}_2\text{CH}_3$ — a primary alkyl bromide that already carries a methyl ester at the other end of the chain. Why Grignard fails here. Forming a Grignard reagent from this substrate is mechanistically doomed from the start: as soon as any $\text{RMgBr}$ begins to form at the metal surface, the nucleophilic carbanion end of one molecule will attack the electrophilic ester carbonyl of a second, unreacted molecule of starting material (or of product Grignard still bearing its ester), well before enough $\text{CO}_2$ can ever be introduced. Grignard reagents and esters are mutually incompatible — a Grignard adds twice to an ester (through a tetrahedral alkoxide intermediate that expels methoxide, giving a ketone, which is itself immediately attacked by a second equivalent of Grignard) to give, after workup, a tertiary alcohol. This uncontrollable self-condensation/polymerization makes the direct Grignard route unusable while an ester is present elsewhere in the same molecule. Nitrile route succeeds. Cyanide, by contrast, is not nucleophilic enough at typical $S_N2$ reaction temperatures to attack the ester carbonyl competitively — esters are reasonably resistant to attack by weak, poorly basic nucleophiles like $^{-}\text{CN}$ under the mild conditions ( $\text{DMSO}$ or $\text{DMF}$ , room temperature to mild heating) used for halide displacement, because $S_N2$ at the ester's saturated (nonexistent) electrophilic carbon isn't the operative pathway anyway — cyanide's only viable reaction is the $S_N2$ backside attack at the primary alkyl bromide carbon, which proceeds cleanly:

\text{BrCH}_2\text{CH}_2\text{CH}_2\text{CO}_2\text{CH}_3 \xrightarrow{\text{NaCN}} \text{N}\!\equiv\!\text{C–CH}_2\text{CH}_2\text{CH}_2\text{CO}_2\text{CH}_3

Subsequent hydrolysis under aqueous acid (hot, dilute $\text{H}_2\text{SO}_4$ , reflux, excess water) then hydrolyzes both the nitrile (through its amide intermediate, as derived above) and the methyl ester (by the standard acid-catalyzed ester hydrolysis mechanism — protonation of the carbonyl, addition of water, proton transfer, and loss of methanol) simultaneously, since both functional groups are susceptible to the same forcing aqueous-acid conditions:

\text{N}\!\equiv\!\text{C–CH}_2\text{CH}_2\text{CH}_2\text{CO}_2\text{CH}_3 \xrightarrow[\Delta]{\text{H}_3\text{O}^+,\ \text{excess H}_2\text{O}} \text{HO}_2\text{C–CH}_2\text{CH}_2\text{CH}_2\text{–CO}_2\text{H}

Carbon count: the starting bromoester has 5 carbons (4 in the chain plus the ester's methyl carbon, which is lost as methanol during hydrolysis, so effectively 4 carbons retained in the chain); adding the nitrile carbon gives 5 carbons in the product diacid skeleton, matching glutaric acid, $\text{HOOC(CH}_2)_3\text{COOH}$ . This example is the clearest illustration of the central design rule of this article: choose the nitrile route whenever the molecule already carries an electrophilic functional group (ester, ketone, aldehyde, epoxide, or acidic N–H/O–H) that a strongly basic, strongly nucleophilic Grignard reagent would attack uncontrollably; choose the Grignard–CO2 route when no such vulnerable group is present, or when the substrate is secondary/branched and prone to E2 elimination under S_N2-forcing cyanide conditions.

Limiting cases, scope & common traps

Scenario	What happens	Why
Excess $\text{CO}_2$ (correct protocol: RMgX added to solid $\text{CO}_2$ )	Clean carboxylate, no ketone byproduct	Carboxylate's delocalized negative charge makes it a poor electrophile relative to CO2 itself; a second RMgX addition is slow when CO2 is never depleted locally
CO2 gas bubbled slowly into excess RMgX	Ketone, then tertiary alcohol, as major byproducts	Local excess of RMgX allows a second nucleophilic addition onto the intermediate carboxylate/derived ketone before workup
Primary unhindered alkyl halide + $^{-}\text{CN}$	Clean $S_N2$ , nitrile forms in high yield	No steric hindrance, no competing β-hydrogen elimination pathway favored
Secondary or branched alkyl halide + $^{-}\text{CN}$	Significant E2 elimination competes, lowering nitrile yield	Cyanide is a moderately strong, small, hard base; steric hindrance disfavors backside $S_N2$ attack, favoring proton abstraction instead
Aryl or vinyl halide + $^{-}\text{CN}$	No reaction (without Pd/Cu catalysis)	$S_N2$ requires backside attack through the halide's C–X σ*; this orbital is inaccessible/geometrically wrong at an $sp^2$ carbon in the standard uncatalyzed pathway
Aryl or vinyl halide + Mg, then CO2	Works well — aryl/vinyl Grignards form readily and carboxylate cleanly	Grignard formation proceeds by a radical mechanism at the metal surface, not $S_N2$ ; C(sp2)–Mg bonds are perfectly competent nucleophiles toward CO2
Substrate bearing an ester, ketone, aldehyde, or acidic O–H/N–H, taken through Grignard route	Grignard reagent self-destructs or attacks the wrong carbonyl uncontrollably	RMgX is both a strong base and a strong carbon nucleophile; incompatible with any electrophilic or acidic group elsewhere in the same molecule
Same substrate, taken through nitrile route	Ester or other group typically survives $S_N2$ /cyanide step; may or may not survive the harsh hydrolysis step later, but that step is controllable/separable in time	Cyanide is weakly nucleophilic toward carbonyls under mild $S_N2$ conditions; hydrolysis conditions can sometimes be tuned (partial hydrolysis stopping at amide)
Nitrile hydrolysis under mild/controlled conditions	Stops cleanly at the primary amide	Amide resonance stabilizes the amide intermediate strongly; a second, separate nucleophilic attack (harsher conditions) is required to push on to the acid
As temperature → very low or catalyst absent	Nitrile hydrolysis essentially does not proceed at a useful rate	Water is a poor nucleophile and the uncatalyzed addition to a nitrile has a substantial activation barrier; acid/base catalysis (or nitrile hydratase enzymes) is required

Common trapMisconception: "CO2 always adds to give a ketone or over-addition product"

Students sometimes assume that because Grignard reagents add twice to esters (giving tertiary alcohols after a ketone intermediate), the same double-addition must happen with $\text{CO}_2$ . It does not, provided the standard protocol is followed (RMgX poured onto excess solid CO2). The key structural difference is that the first addition to $\text{CO}_2$ produces a resonance-delocalized carboxylate, not a localized alkoxide — the negative charge is smeared over two equivalent oxygens, which markedly lowers the electrophilicity of that carbon toward a second RMgX. Ketone/tertiary-alcohol byproducts only dominate when the addition order is reversed (limited CO2 bubbled into excess Grignard solution), because then local RMgX excess persists even after the first addition, so a second equivalent can attack before the resonance-stabilized carboxylate has a chance to be the only species present.

The chemist behind it

Victor Grignard discovered the organomagnesium halide reagents that bear his name in 1900, while working in the laboratory of Philippe Barbier at the University of Lyon on related organozinc chemistry. Barbier had been struggling with an unreliable organozinc addition to an unsaturated ketone and suggested his student try magnesium instead — magnesium turnings reacting with alkyl halides in dry ether gave a far more robust, generally applicable organometallic reagent. Grignard's systematic exploration of the scope of these new reagents, including their addition to carbon dioxide to make carboxylic acids, formed the basis of his 1901 doctoral thesis and, remarkably quickly, became one of the most widely adopted reactions in all of organic synthesis. He shared the 1912 Nobel Prize in Chemistry with Paul Sabatier (for catalytic hydrogenation), a recognition unusually swift for a reaction discovered barely a decade earlier — testament to how immediately useful and mechanistically transparent the chemistry proved to be. Grignard went on to a long academic career at Nancy and Lyon, continuing to develop organometallic methodology and later contributing to wartime chemical detection methods during the First World War.

Origins · 1 minVictor Grignard and how this was discovered

Victor Grignard

1871–1935 · French organic chemist, University of Lyon and Nancy; discovered organomagnesium reagents in Barbier's laboratory in 1900; shared the 1912 Nobel Prize in Chemistry

History

- 1900 — Grignard, working under Barbier at Lyon, reports that alkyl halides react with magnesium in dry ether to give a new, general organometallic reagent.

- 1901 — Grignard's doctoral thesis systematically explores the scope of the reagent, including addition to carbonyl compounds and to carbon dioxide to give carboxylic acids.

- 1912 — Grignard and Paul Sabatier jointly awarded the Nobel Prize in Chemistry.

- Early-to-mid 20th century — nitrile hydrolysis (already known from Pinner-type and classical hydrolysis studies of the 19th century) is systematically combined with the Kolbe nitrile synthesis (alkyl halide + cyanide) into the standard two-step chain-lengthening sequence taught today.

- Mid-to-late 20th century — enzymatic nitrile hydrolysis (nitrile hydratases and nitrilases) is characterized and later exploited industrially, offering a mild biocatalytic alternative to forcing acid/base hydrolysis conditions.

Why does the Grignard reagent add to CO2 in a single step, without forming a stable tetrahedral intermediate first?

It does form a short-lived tetrahedral-like arrangement transiently at the moment of bond formation, but that arrangement is never a discrete, observable intermediate the way it is in ester or amide hydrolysis. The reason is that the carbon of CO2 only has two substituents (the two oxygens) before attack, so after the carbanion adds, the resulting carboxylate has exactly the substitution pattern (R, O⁻, =O) that is already the thermodynamically preferred resting state — there is no extra atom to expel afterward, unlike in ester substitution where a leaving group (alkoxide) must depart from a genuinely tetrahedral sp3 carbon. The immediate product of addition to CO2 is the final product (as its magnesium salt), so there is nothing further for a "tetrahedral intermediate" to collapse into.

Can Grignard reagents be carboxylated with anything other than solid CO2?

Yes — CO2 gas can be used if bubbled carefully in excess (with good stirring and cooling to keep it dissolved rather than escaping), and in industrial or specialized settings supercritical CO2 has also been used as both solvent and reagent. Solid CO2 (dry ice) remains the standard undergraduate laboratory choice mainly for convenience and safety: it is easy to weigh out a large local excess, it keeps the reaction mixture cold (suppressing Grignard reagent decomposition and side reactions), and it avoids the need for specialized gas-handling equipment. The underlying mechanism — nucleophilic attack on the electrophilic central carbon of CO2 — is identical regardless of the physical form in which CO2 is delivered.

Does the nitrile hydrolysis mechanism change qualitatively between acid and base conditions, or is it fundamentally the same reaction?

The two mechanisms share the same overall logic (two sequential nucleophilic additions to increasingly electrophilic carbons, nitrile then amide) but differ in exactly which species is the attacking nucleophile and which intermediate is protonated versus anionic. Under acid, the nitrile is first activated by protonation at nitrogen (making the carbon a much better electrophile) before neutral water attacks; under base, hydroxide — already a strong nucleophile — attacks the unactivated nitrile directly, with no need for prior protonation. The amide hydrolysis stage likewise differs: acid conditions require protonating the amide carbonyl before water attacks, while base conditions let hydroxide attack the neutral amide directly, and the reaction is driven to completion by irreversible deprotonation of the product acid to unreactive carboxylate.

Why can't aryl halides be converted to aryl nitriles by simply treating them with NaCN, the way alkyl halides can?

The SN2 mechanism absolutely requires backside attack of the nucleophile along the axis directly opposite the C–X bond, through the σ orbital of that bond. At an sp2, aryl carbon, that backside trajectory is blocked by the ring itself — the carbon is planar and part of a rigid aromatic system, geometrically incompatible with the required inversion pathway, and there is no accessible σ orbital oriented correctly relative to an approaching nucleophile in the same simple sense as at a tetrahedral sp3 carbon. Industrially, aryl nitriles are instead made by Pd- or Cu-catalyzed cyanation (Rosenmund–von Braun reaction, or modern palladium-catalyzed cyanation), which operates through an entirely different oxidative-addition/reductive-elimination catalytic cycle rather than a direct SN2 displacement.

If I only need an amide, not a carboxylic acid, is nitrile partial hydrolysis a reliable way to make one?

Yes, and it is a genuinely useful synthetic strategy, not merely an unwanted side-stop. Because the amide is considerably more resistant to a second nucleophilic attack than the nitrile is to its first, careful control of stoichiometry, temperature, and reaction time (or use of milder catalytic conditions, such as basic hydrogen peroxide in the Radziszewski reaction, which selectively converts nitriles to amides) allows the reaction to be halted cleanly at the amide stage in good yield. This is a standard method for preparing primary amides that would otherwise be inconvenient to access directly from a carboxylic acid (which would require activating the acid, e.g. as an acid chloride, before amide formation).

Why does the Grignard route add exactly one carbon, never more, even though CO2 itself contains one carbon flanked by two oxygens?

Only one new carbon–carbon bond forms in the entire sequence: the carbanion attacks the single carbon atom of CO2 exactly once. The two oxygens of CO2 do not contribute any additional carbons to the product — they simply end up as the two oxygens of the resulting carboxylic acid group (one as the C=O, one as the C–OH after protonation). So the "+1" in the carbon-count bookkeeping refers strictly to that one central carbon of CO2 becoming the new carboxylic acid carbon, regardless of how many heteroatoms happen to be attached to it.

What happens if trace water is present in the ether during Grignard formation or before CO2 addition?

Any trace of water (or any acidic proton source — alcohols, amines, terminal alkynes) reacts with the Grignard reagent essentially instantly and irreversibly, protonating the carbanion to regenerate the simple alkane (R–H) and destroying the nucleophile before it ever reaches CO2. This is why Grignard reactions are run under rigorously anhydrous conditions (oven-dried glassware, dry solvent freshly distilled from a drying agent, inert atmosphere) and why even atmospheric moisture is excluded with a nitrogen or argon blanket; a Grignard reagent is simultaneously a very strong base and a very strong nucleophile, and water — being both a proton source and a competing nucleophile — defeats it on both counts simultaneously.

Carboxylation of Grignard Reagents and Nitrile Hydrolysis

Where this lives today

The physical picture

The core ideas

Worked examples

Limiting cases, scope & common traps

The chemist behind it

Before this

Where this leads