Jones Oxidation: Orange Chromium Turns Green

The Mechanism: An Ester of Chromic Acid, Then One Broken C–H

Chromium trioxide is a solid anhydride. Dropped into water it hydrates to chromic acid, H₂CrO₄, which at the acidity and concentration of Jones reagent condenses further to dichromate and higher polychromates — the source of the deep orange. The reactive species is a small equilibrium population of protonated chromic acid, in which chromium is a hard, strongly electrophilic d⁰ centre.

Step 1 — esterification. The lone pair on the alcohol oxygen attacks chromium, a proton shifts, and water leaves. The product is a chromate ester, R₂CH–O–CrO₃H. This is fast, acid-catalysed and fully reversible — a pre-equilibrium, not the rate-determining step. The oxygen now carries a very good leaving group, and the carbinol hydrogen is still untouched.

Step 2 — the slow step. A water molecule acts as the base and plucks that carbinol C–H. The electron pair it leaves behind rolls into the C–O bond to make the carbonyl π bond, and the O–Cr bond breaks heterolytically, carrying the pair away on chromium. It is an E2-like collapse: the hydrogen has to sit roughly anti-periplanar to the chromium across the C–O axis. Out comes the ketone; chromium is left two electrons richer, as Cr(IV).

Step 3 — cleaning up the chromium. Cr(IV) is unstable and not the end of the story. It comproportionates with unreacted Cr(VI) to give two Cr(V), and Cr(V) is itself a competent two-electron oxidant that converts more alcohol to carbonyl while dropping to Cr(III). (A rival path, established by Jan Roček's group, has Cr(IV) abstract a hydrogen atom to give a carbon radical that Cr(VI) then intercepts; both end in the same place.) Every chromium atom finishes at Cr(III), which in water is kinetically inert and drops out of the chemistry as green sludge.

Three Alcohols Per Two Chromiums, and the Colour That Counts Them

Add the steps up and the balanced equation falls out: 3 R₂CHOH + 2 H₂CrO₄ + 6 H⁺ → 3 R₂C=O + 2 Cr³⁺ + 8 H₂O. The ratio is 3:2 for one reason only — each alcohol surrenders two electrons, and Cr(VI)→Cr(III) is a three-electron drop. Two chromiums accept six electrons; three alcohols supply them. The recipe encodes the same arithmetic: 26.7 g CrO₃ plus 23 mL of concentrated H₂SO₄ diluted to 100 mL is 2.67 M in chromium, and 2.67 × 3 = 8 N in oxidising equivalents.

Worked example. Oxidise 1.00 g of cyclohexanol (M = 100.16 g/mol, 9.98 mmol) to cyclohexanone. A secondary alcohol is a two-equivalent substrate, so you need 2 × 9.98 = 20.0 meq, and at 8 meq/mL that is 2.5 mL of Jones reagent — check it the other way: 9.98 × (2/3) = 6.65 mmol of Cr(VI), and 6.65/2.67 = 2.5 mL. At 85% yield you recover 0.83 g of cyclohexanone, and you generate 6.65 mmol × 52.0 g/mol = 0.35 g of chromium as toxic waste for every gram of substrate. A primary alcohol taken to the acid is a four-equivalent substrate and needs double that.

The colour is that electron ledger in the visible spectrum. Cr(VI) is d⁰, so it has no d electrons to excite; its absorption near 440 nm is a ligand-to-metal charge transfer band, oxygen 2p to chromium 3d, fully allowed and therefore intense — of order 10³ M⁻¹ cm⁻¹ at the near-UV band maximum and still of order 10² on the visible tail at 440 nm. Reduce it and you get d³ Cr(III), whose only visible transitions are the Laporte-forbidden d–d bands of [Cr(H₂O)₆]³⁺ near 408 nm and 575 nm (⁴A₂g → ⁴T₁g and ⁴A₂g → ⁴T₂g), with ε around 13–16 M⁻¹ cm⁻¹ — one to two orders of magnitude weaker. The 575 nm band is the octahedral splitting: 17,400 cm⁻¹, about 208 kJ/mol. The hexaaqua ion transmits in both the blue-green and the red and looks violet; in Jones' sulfuric liquor sulfate enters the coordination sphere, the red window closes, and green survives.

Westheimer's Deuterium: Proving Which Bond Breaks

Which step is the slow one was settled by isotopic substitution. In 1949 Frank Westheimer and N. Nicolaides oxidised isopropanol-2-d — ordinary isopropanol with the single carbinol hydrogen replaced by deuterium — and found it reacted about 6.7 times slower. A primary kinetic isotope effect that large can only mean the C–H bond is being broken in the transition state of the slow step.

Worked example. Where does 6.7 come from? The difference is zero-point energy. A C–H stretch sits near 2900 cm⁻¹ and a C–D stretch near 2100 cm⁻¹, because swapping H for D nearly doubles the reduced mass of the oscillator and ν ∝ 1/√μ. The zero-point energies are half of each, so C–D starts (2900 − 2100)/2 = 400 cm⁻¹ deeper in its well — and at the transition state, where the bond is half-broken, that advantage is largely gone. Convert: 400 cm⁻¹ × 11.96 J mol⁻¹ per cm⁻¹ = 4.79 kJ/mol. Then k_H/k_D = exp(ΔZPE/RT) = exp(4790 / (8.314 × 298)) = exp(1.93) = 6.9. Measurement and back-of-envelope agree to within 3%, which is why this became the textbook demonstration of the method.

The kinetics corroborate it. The rate is first order in alcohol, first order in Cr(VI) and acid-catalysed, steepening beyond first order in [H⁺] at high acidity where the Hammett function h₀ takes over — exactly the signature of a fast protonation-and-esterification pre-equilibrium feeding a slow C–H cleavage. The geometry requirement shows up too: in conformationally locked rings, an axial alcohol such as cis-4-tert-butylcyclohexanol is oxidised roughly 3 times faster than its equatorial epimer, because the bulky axial chromate ester is strained in the ground state and that strain is relieved as the carbon flattens to sp².

Making It, Running It, and Knowing When to Stop

The reagent is deliberately simple: dissolve 26.7 g of CrO₃ in water, add 23 mL of concentrated sulfuric acid with cooling, and dilute to 100 mL. It keeps for months in a glass bottle. The substrate goes into acetone — miscible with the aqueous reagent, a solvent for most organics, and already oxidised at its carbonyl carbon, so it does not compete. The reagent is added dropwise into the substrate at 0–25 °C, never the reverse, and never onto solid CrO₃, because chromic acid and acetone together can run away exothermically.

The reaction is its own indicator. Each drop of orange flashes green as it is consumed. When the alcohol is gone, the next drop stays orange — a persisting orange tinge is the endpoint, a titration you read with your eyes. Excess is quenched with a splash of isopropanol, which turns the last of it green, and the product is decanted or extracted away from the chromium sludge. Confirmation is routine: in the IR the O–H stretch near 3300 cm⁻¹ vanishes and a ketone C=O appears near 1715 cm⁻¹ (an acid shows a broad 2500–3300 cm⁻¹ O–H instead); in the ¹H NMR the carbinol proton at 3.5–4 ppm disappears.

Everything around the reaction is now governed by the reagent's toxicity, not its chemistry. Cr(VI) compounds are an IARC Group 1 carcinogen — established, not suspected, in humans. In the United States, OSHA's 2006 standard sets a permissible exposure limit of 5 µg/m³ of Cr(VI) as an 8-hour average, with an action level of 2.5 µg/m³. Waste streams are assayed by the diphenylcarbazide colorimetric method, in which Cr(VI) oxidises the reagent to diphenylcarbazone and the magenta Cr(III) complex that results is read at 540 nm, or by ion chromatography with post-column derivatisation at trace levels.

1946, the Breathalyzer, and the Sunset Date

The reagent comes from an acetylene paper. Bowden, Heilbron, Jones and Weedon, J. Chem. Soc. 1946, 39 were trying to oxidise acetylenic carbinols to acetylenic ketones — substrates whose triple bonds are wrecked by most oxidants and whose products are prone to rearrangement. Their trick was speed and segregated addition: substrate in acetone, chromium in its own flask of water and acid, meeting only a drop at a time, so only a trace of oxidant is ever present and the fragile product sees it for seconds. Ewart Jones's name stuck to the reagent.

The same redox couple then went out into the world in a very different box. In 1954 Robert Borkenstein, an Indiana State Police captain, built the first Breathalyzer, sold for decades as the Model 900. It trapped a fixed 52.5 mL sample of deep-lung breath and bubbled it through acidified potassium dichromate with a silver catalyst. Ethanol — a primary alcohol — was oxidised to acetaldehyde and on to acetic acid, bleaching the orange in proportion to the dose, and a photocell measured the loss of absorbance at 440 nm: the Cr(VI) charge-transfer band again. Blood alcohol was inferred through the legally fixed 2100:1 blood-to-breath ratio, a statutory average for the partition of ethanol between blood and alveolar air, not a constant of nature. Modern instruments read infrared absorption or use a fuel cell; the dichromate machines are museum pieces.

Regulation has since gone after the reagent itself. Chromium trioxide was placed on Annex XIV of the EU's REACH regulation and reached its authorisation sunset date on 21 September 2017: after that, using it in Europe requires a specific, time-limited authorisation. The same ion has a broader public reputation: the groundwater contamination at Hinkley, California, settled by PG&E for $333 million in 1996, was Cr(VI) used as a cooling-tower corrosion inhibitor.

How It Fails, and the Reagents It Is Mistaken For

Tertiary alcohols do not react. They form the chromate ester perfectly well, but there is no hydrogen on the carbinol carbon for water to remove, so the slow step has nothing to do and the ester simply hydrolyses back. Any apparent oxidation of a tertiary alcohol under Jones conditions is really acid-catalysed dehydration to an alkene, followed by oxidation of something else. That inertness is itself a diagnostic for a tertiary alcohol.

Primary alcohols cannot be stopped at the aldehyde, and this is the single most common misunderstanding. The reason is the water, not the chromium. An aldehyde in aqueous acid sits in equilibrium with its gem-diol hydrate, RCH(OH)₂, which has exactly what the mechanism wants — a hydroxyl to esterify and a C–H on the same carbon — so it is oxidised straight through to the carboxylic acid. Remove the water, as PCC (Corey and Suggs, 1975) and the Collins reagent do by working in dry dichloromethane, and the hydrate cannot form: the reaction halts at the aldehyde. Same metal, same oxidation state, different stopping point, and the difference is a solvent.

The other failures are consequences of the acid. Jones reagent is roughly 4 M in sulfuric acid, which is hostile to acetals, ketals, THP and trityl ethers, Boc carbamates and tertiary-alcohol-containing scaffolds; α-stereocentres next to the new carbonyl can epimerise; alkenes may isomerise into conjugation, and electron-rich ones are attacked outright. Vicinal diols are cleaved to two carbonyls. If the substrate cannot survive that, the answer is a milder oxidant — Swern at −78 °C or Dess–Martin periodinane at room temperature, both chromium-free. That, with the carcinogenicity and the 0.35 g of chromium waste per gram of substrate, is why one of the fastest and highest-yielding transformations in organic chemistry now appears mostly in teaching labs and in process routes where nothing else is cheap enough.

Jones against the chromium and non-chromium oxidants it is most often confused with

Oxidant	Medium	A primary alcohol gives	Why you would or would not use it
Jones (CrO₃, H₂SO₄, water, acetone)	Aqueous, strongly acidic, 0–25 °C	Carboxylic acid — the aldehyde hydrates and is oxidised again	Fast, cheap, 80–95%; destroys acetals, THP ethers, Boc groups and acid-sensitive stereocentres
PCC, pyridinium chlorochromate (Corey–Suggs, 1975)	Anhydrous CH₂Cl₂, weakly acidic	Aldehyde	No water means no hydrate, so it stops one rung short; leaves a tarry chromium residue
PDC, pyridinium dichromate (Corey–Schmidt, 1979)	Anhydrous CH₂Cl₂, or DMF	Aldehyde in CH₂Cl₂; carboxylic acid in DMF	Near-neutral; the solvent, not the reagent, chooses the stopping point
Collins / Sarett (CrO₃·2 pyridine)	Anhydrous CH₂Cl₂, or neat pyridine	Aldehyde	Basic and anhydrous, but needs a 6–12-fold excess and the reagent ignites if CrO₃ and pyridine are combined in the wrong order
Swern (DMSO, oxalyl chloride, Et₃N)	Anhydrous, −78 °C, basic	Aldehyde	Chromium-free and very mild; the by-product is dimethyl sulfide, and warming above −60 °C wrecks it
Dess–Martin periodinane	Anhydrous CH₂Cl₂, 25 °C, near-neutral	Aldehyde	Iodine(V), not chromium; the mildest of the set and the most expensive