Norrish Type I and II Photoreactions

How the excited carbonyl behaves

Both Norrish reactions begin the same way. Absorbing a UV photon promotes an electron from a non-bonding oxygen lone pair into the antibonding π* orbital of the carbonyl — the classic n→π* transition that gives simple ketones a weak absorption around 280–320 nm. Intersystem crossing usually delivers the molecule to the lowest triplet state (T₁), whose excited oxygen is electron-poor and behaves much like an alkoxy radical.

This diradical-like oxygen is the reactive species. Whether it fragments the neighbouring C–C bond (Type I) or abstracts a distant hydrogen (Type II) depends on the substrate: the availability of an abstractable γ-hydrogen, the stability of the radicals that would form on α-cleavage, temperature, and phase. In the gas phase and in solution the two pathways genuinely compete, and many ketones show both.

Norrish Type I: α-cleavage

In the Type I (or Norrish I) reaction, the bond between the carbonyl carbon and the adjacent α-carbon breaks homolytically. One electron goes to each fragment, producing an acyl radical (R–C•=O) and an alkyl radical (•R′). This is favoured when the resulting radical is stabilised — for example when the α-carbon is tertiary or benzylic.

The radical pair then has several fates:

• Decarbonylation — the acyl radical ejects carbon monoxide (CO) to give a second alkyl radical, which combines with the first to form a new C–C bond. Dibenzyl ketone, for instance, loses CO on irradiation to give bibenzyl.
• Recombination back to starting material, or to an isomeric ketone.
• Disproportionation to an alkene plus an aldehyde.
• Ring cleavage in cyclic ketones. Cyclohexanone, on α-cleavage and intramolecular hydrogen transfer, opens to an unsaturated aldehyde (5-hexenal) or, after decarbonylation, to 1-pentene and CO.

Because it generates clean radical pairs, Norrish I is a workhorse for studying radical cage effects and is exploited in modern photoinitiators such as benzoin ethers and acylphosphine oxides, whose α-cleavage produces the radicals that start acrylate polymerisation in UV-cured coatings, inks, and dental resins.

Norrish Type II: γ-hydrogen abstraction and the 1,4-biradical

The Type II reaction needs a hydrogen on the carbon three positions from the carbonyl (the γ-carbon). The excited oxygen abstracts this γ-H through a six-membered cyclic transition state — the same six-membered geometry that makes 1,5-hydrogen transfers so common in radical chemistry. The result is a 1,4-biradical: one radical centre on oxygen (now a hydroxyl-bearing carbon) and one on the former γ-carbon.

This biradical then partitions between two outcomes:

• Fragmentation (Norrish II cleavage) — the central Cα–Cβ bond breaks, splitting the molecule into an alkene and an enol, which tautomerises to a smaller ketone or aldehyde. 2-Hexanone, for example, cleaves to give propene and the enol of acetone.
• Cyclization (the Yang reaction) — the two radical ends couple to form a four-membered ring, giving a cyclobutanol. Discovered by Nien-Chu Yang in 1958, this Norrish–Yang cyclization is a rare direct route to strained cyclobutanols and is valued in complex-molecule synthesis.

The ratio of fragmentation to cyclization depends on how the biradical is oriented and on how quickly the singlet biradical closes versus the triplet biradical returns to starting material. Solvents that hydrogen-bond to the incipient hydroxyl (alcohols, water) lengthen the biradical lifetime and typically raise the fragmentation yield, sometimes above 90%, by disrupting the intramolecular hydrogen bond that otherwise favours reversion.

Selectivity, stereochemistry, and controlling the outcome

Several factors decide which reaction wins and how selective it is:

• Available γ-H: No γ-hydrogen means no Type II, so the molecule defaults to Type I α-cleavage. Conversely, a weak, well-positioned γ-C–H bond (tertiary or benzylic) accelerates Type II.
• Excited-state multiplicity: Triplet ketones (n→π* triplets such as those of aryl alkyl ketones) abstract hydrogen efficiently and are the usual Type II reactants. Singlet pathways are shorter-lived and show more reversion.
• Confinement: Carrying out the reaction inside zeolites, micelles, or cyclodextrin cavities restricts biradical geometry and can dramatically boost cyclization over fragmentation, and even induce diastereoselectivity in the cyclobutanol products.
• Ring size in Type I: Small strained ketones open readily; strain relief drives ring cleavage.

Because the intermediate is a diradical, absolute stereocontrol is hard, but the cyclobutanols from Yang cyclization form with defined relative configuration, and chiral confining media or auxiliaries have been used to make the products enantioenriched.

Applications: polymers, atmosphere, and synthesis

Polymer photodegradation. Carbonyl groups introduced into polyolefins — deliberately in some biodegradable plastics, or accidentally by oxidation — make the material susceptible to sunlight. Norrish I and II reactions cleave the backbone: Type II chain scission is a major reason polyethylene and polypropylene become brittle and crack on prolonged UV exposure. Poly(methyl vinyl ketone) and related ketone-bearing polymers are model systems for studying this photo-scission.

Atmospheric chemistry. Carbonyl compounds such as acetone and larger aldehydes released into the troposphere absorb near-UV sunlight and undergo Norrish I photolysis, generating radicals (including CO and acyl radicals) that feed into the free-radical cycles governing ozone and smog formation. The photolysis quantum yields and rates of these carbonyls are input data for atmospheric models.

Photoinitiators and curing. As noted, Norrish I α-cleavage of designed ketones is the chemical heart of UV-cured coatings, adhesives, 3D-printing resins, and dental composites, where sub-second irradiation triggers hardening.

Total synthesis. Norrish–Yang cyclization builds cyclobutane and cyclobutanol rings that are otherwise awkward to access, and controlled Type II fragmentation has been used as a mild, non-basic way to cleave carbon chains under neutral conditions using only light.

A short history

Ronald Norrish began systematic gas-phase photolysis of ketones and aldehydes at Cambridge in the late 1920s and 1930s, distinguishing the two competing modes now named after him. After the Second World War he and his student George Porter developed flash photolysis — using an intense light pulse to create excited molecules and radicals in high concentration, then probing them spectroscopically on microsecond timescales. This let chemists directly observe the transient acyl radicals and biradicals that the Norrish mechanisms predicted.

For pioneering the study of extremely fast chemical reactions by these methods, Norrish and Porter shared the 1967 Nobel Prize in Chemistry with Manfred Eigen. Nien-Chu Yang reported the cyclobutanol-forming variant of the Type II reaction in 1958, and the combined process is often called the Norrish–Yang reaction. Flash photolysis itself grew into the femtosecond spectroscopy that later earned Ahmed Zewail the 1999 Nobel Prize, making the Norrish work a foundation stone of modern reaction dynamics.

Norrish Type I vs Type II at a glance

Feature	Norrish Type I	Norrish Type II
Key step	Homolysis of the α C–C bond next to C=O	Intramolecular abstraction of a γ-hydrogen
Intermediate	Acyl + alkyl radical pair	1,4-biradical
Geometric need	None special	Six-membered cyclic transition state (γ-H)
Typical products	Decarbonylation, ring-opening, recombination	Alkene + enol/ketone (fragmentation) or cyclobutanol (Yang)
Favoured by	Strained or stabilised α-radicals	Molecules with abstractable γ C–H bonds