Methods for Preparing Carboxylic Acids

Oxidative routes to carboxylic acids remain industrially central. Terephthalic acid, the monomer feeding polyethylene terephthalate (PET) production for textiles and plastic bottles, is manufactured on a colossal scale by the Amoco (Mid-Century) process: catalytic air oxidation of p-xylene using a Co/Mn/Br catalyst system, which is mechanistically a radical-chain benzylic oxidation closely related in spirit (though catalytically distinct in detail) to the KMnO₄ benzylic-cleavage chemistry taught as the classroom archetype — both exploit the same weakened benzylic C–H bond as the point of attack. Benzoic acid itself, produced industrially by air oxidation of toluene, is used as a food preservative (as sodium benzoate) and as a precursor to plasticizers and pharmaceutical intermediates. In pharmaceutical synthesis, controlled oxidation of primary alcohols to carboxylic acids (or stopping cleanly at the aldehyde with PCC- or Dess–Martin-type reagents, and finishing with a milder Pinnick oxidation, NaClO₂, when acid-sensitive groups are present) is a routine step in building carboxyl-bearing drug fragments, since many active pharmaceutical ingredients — ibuprofen, naproxen, and countless others — carry a carboxylic acid pharmacophore that must be installed with full retention of nearby stereocenters, exactly as illustrated in the worked stereochemistry example above. The carbon-extending routes previewed here — carboxylation of Grignard reagents and hydrolysis of nitriles — dominate laboratory-scale synthesis whenever the target acid needs a specific isotopic label at the carboxyl carbon (using ¹³CO₂ or ¹⁴CO₂ to make isotopically labeled acetic or benzoic acid for mechanistic or metabolic tracer studies) or whenever the required carbon skeleton is most conveniently built by first making an alkyl or aryl halide, converting it to a Grignard reagent or nitrile, and only then installing the carboxyl carbon. These routes are developed fully, with complete curved-arrow mechanisms for nucleophilic addition to CO₂ and for the tetrahedral-intermediate cascade of nitrile hydrolysis, under Carboxylation of Grignard Reagents and Nitrile Hydrolysis.

Ask the tutor: How does the mechanism change if I use ozone or hydrogen peroxide instead of chromium/manganese oxidants to cleave an alkylbenzene side chain?

The organizing idea is oxidation state. Assign oxidation numbers to carbon by treating C–H bonds as contributing –1 to carbon (H is less electronegative) and C–O, C–N, C–halogen bonds as contributing +1 to carbon (O, N, halogen are more electronegative), with C–C bonds contributing 0. A primary alcohol carbon, R–CH₂–OH, has two C–H bonds and one C–O bond: oxidation number –1. An aldehyde carbon, R–CHO, has one C–H and a C=O (counted as two C–O contributions since a double bond counts twice): oxidation number +1. A carboxylic acid carbon, R–COOH, has no C–H at all and three oxygen-type contributions (one C=O double bond plus one C–OH): oxidation number +3. Going from alcohol to acid is therefore a four-electron oxidation carried out in two two-electron steps, with the aldehyde as the obligatory, isolable-in-principle intermediate. The mechanistic reason the reaction does not simply stop at the aldehyde in aqueous oxidizing conditions is that the aldehyde is not itself the species oxidized in the second step. Aldehydes in water exist in equilibrium with their hydrates, R–CH(OH)₂, formed by nucleophilic addition of water across the carbonyl π orbital. The hydrate carbon has two C–OH groups and one C–H: oxidation number +1, same as the aldehyde (hydration doesn't change oxidation state — it's an addition, not a redox step), but crucially the hydrate has an O–H bond positioned exactly like an alcohol O–H, so it can undergo the same alkoxide-elimination oxidation mechanism a second time, now producing the carboxylic acid (oxidation number +3). Anhydrous reagents such as PCC deny the aldehyde any water to hydrate with, so this second oxidation event simply cannot occur, and the reaction halts cleanly at the aldehyde. For alkylbenzenes, the ring itself is inert to these oxidants (aromatic π systems resist the electrophilic Cr(VI)/Mn(VII)-based oxidations used here), but any C–H bond benzylic* to the ring — i.e., on the carbon directly attached to the aromatic ring — is weakened by hyperconjugative/resonance stabilization of the radical or cationic intermediate formed upon its removal. Hot permanganate or dichromate abstracts these benzylic hydrogens repeatedly, walking the benzylic carbon up the same +3 oxidation ladder, and in the process cleaves away every carbon beyond the benzylic position as CO₂ and small fragments. The net transformation is brutal but reliable: any n-alkylbenzene with a benzylic C–H becomes benzoic acid, regardless of how long the original chain was.

The oxidation-state accounting that governs every route in this article can be written compactly. Assign each carbon an oxidation number $N$ by

where: bonds to carbon contribute zero; a double bond to a heteroatom counts as two bonds to that heteroatom. Reading this equation: it is a bookkeeping device, not a physical charge — it simply tracks how many two-electron oxidation events separate one functional group from another, and it predicts how many discrete oxidation steps a given transformation requires. For the alcohol-to-acid sequence the overall electron count is

where: $\Delta N$ is the total number of electrons formally removed from the carbon (a 4-electron, i.e. two-step, oxidation), matching the stoichiometric requirement for two sequential 2-electron oxidant equivalents (e.g., two Cr(VI)→Cr(IV) half-steps, or reduction of two separate Cr(VI) centers each undergoing a 3-electron change per Cr as part of the overall balanced redox couple). For the classic Jones oxidation the balanced half-reactions are

where: each Cr goes from +6 to +3 (a 3-electron change per Cr, 6 electrons per Cr₂O₇²⁻), and each alcohol carbon undergoes the 4-electron change computed above; the coefficients (3 alcohol : 2 dichromate) balance total electrons transferred, $3 \times 4 = 12$ electrons donated $= 2\times 6$ electrons accepted. For benzylic oxidation with hot permanganate, the overall equation for an unbranched n-alkylbenzene ArCH₂R′ (any chain, as long as a benzylic C–H exists) is summarized by

where: every carbon past the benzylic position is destroyed (oxidized to CO₂/carbonate or small acids) and only the benzylic carbon survives, now bonded to the ring as the carboxyl carbon.

Example 1 (foundational): oxidation-state accounting for 1-propanol → propanoic acid. Start from 1-propanol, CH₃CH₂CH₂OH. Assign the oxidation number of C1 (the carbinol carbon): it bears two C–H bonds and one C–O bond (plus one C–C bond, contributing 0), giving

In the product, propanoic acid, CH₃CH₂COOH, C1 bears one C=O (counted as +2) and one C–OH (+1), zero C–H bonds, and one C–C bond:

The net change is

confirming a genuine 4-electron (two-step) oxidation, exactly as the mechanism above requires (one 2-electron step to the aldehyde, a second 2-electron step from the hydrate). Using Jones reagent (CrO₃/H₂SO₄/acetone–water), the reaction proceeds cleanly through propanal and its hydrate to propanoic acid in a single pot, because water is present throughout. If instead PCC in CH₂Cl₂ is used, the reaction stops at propanal, isolable in good yield, because Step 3 of the mechanism (hydration) cannot occur without water. As a limiting-case check: if one deliberately runs the Jones oxidation in strictly anhydrous conditions (not the normal protocol, but instructive), the reaction is observed to stall at the aldehyde stage as well — directly confirming that it is the presence of water, not some special property of chromium(VI), that drives the reaction past the aldehyde.

Example 2 (applied): converting 4-tert-butyltoluene to a single pure carboxylic acid using hot KMnO₄. 4-tert-butyltoluene has two alkyl substituents on the ring: a methyl group (–CH₃) at one position and a tert-butyl group (–C(CH₃)₃) at the para position. Only the methyl group has benzylic C–H bonds (three of them); the tert-butyl group's ipso carbon is fully substituted with three methyl groups and bears no C–H at all at the benzylic position. Applying the mechanism above: hot excess KMnO₄ initiates benzylic C–H abstraction only at the methyl group (Step 1 has no entry point at the tert-butyl carbon, per Step 5 of the second derivation), oxidizing –CH₃ stepwise through –CH₂OH, –CHO/hydrate, to –COOH. The tert-butyl group is untouched because none of its carbons are benzylic C–H bearing carbons subject to the initiating abstraction. The isolated product is 4-tert-butylbenzoic acid, HOOC–C₆H₄–C(CH�3)₃, in which the methyl arm has become a carboxyl group while the bulky tert-butyl arm survives completely intact. This is a standard industrial-adjacent strategy: using a bulky, oxidation-resistant substituent to mark a ring position while a separate methyl arm is deliberately sacrificed to install a –COOH exactly where wanted — a design logic used in preparing substituted benzoic acid building blocks for polymers and pharmaceuticals.

Example 3 (exam/competition level): distinguishing four isomeric oxidation routes to the same acid, and predicting which reagent set is required. Suppose a synthesis target is (S)-2-methylbutanoic acid, and four possible precursors are proposed: (a) (S)-2-methyl-1-butanol (a primary alcohol) oxidized with Jones reagent; (b) 2-methylbutanal (S)-configured at C2, oxidized with Tollens' reagent or mild Ag₂O; (c) sec-butylbenzene-type analog — irrelevant, discard, since there's no aromatic ring in the target; (d) 1-sec-butylbenzene-type chain — again discard for the same reason. Route (a): the stereocenter at C2 is not the alcohol carbon (C1) undergoing oxidation — C1 is CH₂OH, not a stereocenter, and C2 (bearing H, CH₃, CH₂OH-chain, and CH₂CH₃) is the actual stereocenter and is never touched by the oxidation mechanism (Steps 1–4 of the alcohol-oxidation derivation act exclusively on C1). Since no bond to the C2 stereocenter is broken at any point in the chromate-ester/elimination mechanism, configuration at C2 is fully retained, and Jones oxidation of (S)-2-methyl-1-butanol delivers (S)-2-methylbutanoic acid with complete stereochemical fidelity — no racemization, because the stereocenter is a mere spectator two bonds away from all the redox chemistry. Route (b): mild oxidation of the corresponding aldehyde, 2-methylbutanal, with Tollens' reagent (Ag(NH₃)₂⁺) proceeds by nucleophilic addition of hydroxide to the aldehyde carbonyl (forming a tetrahedral alkoxide, exactly analogous to Step 3's hydration but with hydroxide instead of water as nucleophile), followed by hydride transfer from the resulting alkoxide's C–H to Ag(I), reducing it to Ag(0) and oxidizing the carbon to the carboxylate. Again, C2 is a spectator; stereochemistry is retained. This route is milder than Jones oxidation and useful when other oxidation-sensitive groups are present, but functionally gives the same stereochemical outcome by a related mechanistic logic (both mechanisms funnel through a tetrahedral, sp³ oxyanion intermediate at the carbonyl carbon, never touching C2). The competition-level insight: because oxidation state changes occur only at the carbon directly bearing the leaving oxidant fragment (chromate ester oxygen, or the hydroxide/hydride acceptor), any pre-existing stereocenter not itself being oxidized survives both routes with retention — a general and testable prediction of the mechanism, not a memorized fact.

The oxidative degradation of alkylbenzene side chains to benzoic acids is inseparable from the history of structural organic chemistry in the nineteenth century, when chemists had no spectroscopic tools and relied on exhaustive chemical degradation to determine connectivity. August Wilhelm von Hofmann and, more directly relevant, Adolf von Baeyer and contemporaries used vigorous oxidation (originally with nitric acid, later with permanganate once Condy's crystals — potassium permanganate — became a standard laboratory oxidant in the mid-1800s) as a structure-proof tool: if an unknown alkylbenzene, upon exhaustive oxidation, gave benzoic acid, that proved a monosubstituted benzene ring with the alkyl chain attached at a single ring carbon; if it gave a dicarboxylic acid (phthalic, isophthalic, or terephthalic acid), that fixed the substitution pattern (ortho, meta, or para) of the original disubstituted ring. This "exhaustive oxidation" method — destroy the aliphatic parts, read off the surviving aromatic skeleton as a pattern of carboxyl groups — was for decades the primary chemical tool for assigning substitution patterns on benzene rings, decades before X-ray crystallography or NMR existed. The chromic acid oxidation of alcohols, meanwhile, traces to the general nineteenth-century development of chromium(VI) as a laboratory oxidant (chromic acid, "cleaning solution" made from K₂Cr₂O₇/H₂SO₄, was a standard glassware cleaner precisely because of its oxidizing power), refined into synthetically controllable forms much later: Jones reagent was introduced by Sir Ewart Jones in the 1950s as a fast, mild, in-situ chromic acid oxidation compatible with sensitive substrates, and Elias James Corey introduced pyridinium chlorochromate (PCC) in 1975 specifically to solve the "stop cleanly at the aldehyde" problem that plain aqueous Cr(VI) reagents could not solve.