Double- versus Single-Barbed (Fishhook) Arrows

The distinction is load-bearing across nearly all of modern chemistry and biology. In polymer manufacturing, free-radical polymerization of ethylene, styrene, and acrylates (producing polyethylene, polystyrene, and acrylic plastics at industrial scale) is drawn entirely with fishhook arrows, from peroxide-initiator homolysis through propagation to radical termination — while ionic (living anionic or cationic) polymerizations used for precision block copolymers are drawn entirely with double-barbed arrows, and the choice of mechanism directly determines the polymer's molecular weight distribution and architecture. In atmospheric and combustion chemistry, ozone depletion by chlorofluorocarbons proceeds through a Cl• radical chain (fishhooks: Cl• + O₃ → ClO• + O₂, then ClO• + O → Cl• + O₂), a mechanism whose correct radical description was central to the Nobel-recognized work of Rowland, Molina, and Crutzen. In medicinal chemistry and toxicology, cytochrome P450 oxidations proceed through high-valent iron-oxo radical rebound mechanisms (fishhook C–H abstraction followed by radical recombination) that determine drug metabolite structures and sites of toxic bioactivation, in contrast to purely ionic phase-II conjugation chemistry (glucuronidation, sulfation) drawn with double-barbed arrows. In materials and battery chemistry, single-electron-transfer (SET) steps in redox flow batteries and in dissolving-metal reductions (Birch reduction, SmI₂-mediated radical cyclizations widely used in total synthesis) are fishhook chemistry, mechanistically distinct from the double-barbed hydride/proton-transfer steps that complete the same overall transformations. And in antioxidant biochemistry, vitamin E and other radical-trapping antioxidants function by donating a hydrogen atom (fishhook H-atom transfer, quenching a lipid peroxyl radical) rather than by the double-barbed proton-transfer chemistry of ordinary acid–base reactions — a mechanistic distinction with direct consequences for how antioxidant efficacy is measured and designed.

Ask the tutor: Why does the same C–Br bond in the same molecule sometimes break heterolytically and sometimes homolytically, and what specific factor decides it?

Return to the underlying object every arrow represents: a molecular orbital, occupied by either two electrons (a bonding pair or lone pair) or, transiently in a radical, by one electron (a singly occupied molecular orbital, SOMO). The curved arrow is not a cartoon of "motion" in real space; it is a bookkeeping device that says this orbital's electron population is being transferred into that orbital. The whole content of The Curved-Arrow Formalism is the rule that arrow tails start on electrons (a bond or lone pair) and heads point to where those electrons end up (a new bond, a new lone pair, or an atom that becomes electron-poor/rich). The double-barbed arrow moves a pair. Its tail sits on a filled orbital holding two electrons — a lone pair on a Lewis base, or a σ or π bonding pair. Its head lands on an atom (forming a new lone pair there), on another atom in a new bond, or between two atoms forming a new bond. Because two electrons move as a unit, the atom that loses them becomes formally more positive by one unit of charge (if it had electrons removed from a lone pair or one end of a bond it "owned") and the destination becomes more negative by one unit — this is exactly how formal charges update in an ionic mechanism. The species produced are still closed-shell: every electron remains paired, just redistributed. This is heterolytic cleavage or heterolytic bond formation, and it is the entire grammar of acid–base chemistry, SN1/SN2, E1/E2, carbonyl addition, and aromatic substitution. The single-barbed fishhook arrow moves one electron. Its tail sits on a singly occupied orbital (a radical) or on one lobe of a bond that is being split symmetrically; its head lands on a single new location, contributing one electron toward a new bond or a new SOMO. Crucially, because a covalent σ or π bond is a two-electron object, homolysis of a bond requires drawing two fishhooks, one for each electron, each hook sending its electron to a different fragment. The bookkeeping consequence: no formal charges are created (each fragment keeps exactly the number of electrons "its half" of the bond contributed), but each product becomes a radical — an atom or molecule with an odd, unpaired electron in a previously bonding or nonbonding orbital. The orbital-symmetry root of the difference is worth stating explicitly. Heterolysis is favored when the bond's electron pair can be stabilized asymmetrically — one fragment's LUMO/empty orbital accepts nothing, the other's HOMO keeps both electrons, typically because of polarity built into the bond (C–Br is already polarized Cδ+–Brδ−) or because an external nucleophile/base assists by donating into the σ* orbital as the bond breaks (concerted heterolysis, as in SN2). Homolysis is favored when the bond is nonpolar or nearly so (C–C, O–O peroxide bonds, X–X halogens under photolysis) so that splitting the pair symmetrically, one electron to each fragment, costs less energy than forcing an asymmetric ionic split in a medium that cannot stabilize the resulting ions (gas phase, nonpolar solvent, or high temperature/photochemical activation).

Formal charge bookkeeping under a double-barbed arrow step obeys the same rule you already know from Lewis structures:

where: an atom "owns" all its nonbonding electrons plus one electron from each bond it participates in; a double-barbed arrow moving a lone pair or bond pair changes this ownership by exactly 2 electrons at the two ends of the arrow (tail atom loses 1 unit of ownership if the pair becomes a new bond, or 2 if it becomes fully external as in a leaving anion; head atom gains correspondingly), always yielding integer, even-electron formal charges. Radical (fishhook) bookkeeping instead conserves spin and electron count atom-by-atom, with no formal charge created:

where: each homolysis of one σ or π bond creates two new unpaired electrons (net +2 to the radical count), and each radical–radical combination step destroys two unpaired electrons (net −2); a well-formed radical chain mechanism must have this quantity conserved at every single step (each elementary step changes it by exactly 0, +2, or −2, never by an odd number, since electrons cannot appear or vanish singly). The diagnostic equation that tells you which arrow to reach for is the comparison of activation pathways available in the medium:

where: $\Delta G^{\ddagger}_{\text{heterolysis}}$ is lowered by polar/protic solvents that solvate the resulting ion pair (or by an internal/external nucleophile assisting departure) and by substituents that stabilize charge (resonance, hyperconjugation, adjacent lone pairs); $\Delta G^{\ddagger}_{\text{homolysis}}$ is lowered by weak, nonpolar bonds (O–O, N–Br, halogen–halogen), by initiators (light, peroxides, heat) that supply the bond dissociation energy directly, and is essentially insensitive to solvent polarity since no charge is created. Whichever pathway has the lower $\Delta G^{\ddagger}$ under the actual reaction conditions is the one nature (and your mechanism) takes — and that in turn dictates which arrow type is correct, not merely stylistic.

Example 1 (foundational): Drawing the correct arrow for deprotonation of acetic acid by hydroxide. Acetic acid, CH₃COOH, is deprotonated by hydroxide. First identify the source: the O–H bond's bonding pair is the electrons that will move (not a lone pair on oxygen, since it is the bond itself that breaks). Hydroxide's oxygen lone pair is the nucleophile attacking the acidic proton. Draw arrow 1, double-barbed, tail on the OH⁻ lone pair, head landing on the acidic H of the carboxylic acid — this forms the new O–H bond of water. Draw arrow 2, double-barbed, tail on the old C(=O)–O–H bond (the one being broken), head moving onto the carboxyl oxygen, generating a new lone pair there. Check charge: OH⁻ (charge −1) becomes neutral H₂O (its oxygen ownership drops correctly), and the carboxylic acid (neutral) becomes carboxylate (charge −1, oxygen now owns an extra lone pair). Total charge conserved: −1 + 0 = 0 + (−1). Both arrows are double-barbed because every species — acetic acid, hydroxide, water, acetate — is closed-shell; the acetate anion product is stabilized by resonance delocalization of that new lone pair into the neighboring C=O π system, which is only meaningful because it is a discrete electron pair being delocalized, a claim you can only make with double-barbed formalism. As a limiting-case check: if instead of OH⁻ we used a radical base, this mechanism would be entirely different (H-atom abstraction, fishhooks) — the arrow type is dictated by the actual electronic character of the reagent, hydroxide's electron pair versus a hypothetical radical's single electron.

Example 2 (applied): NBS allylic bromination — distinguishing the ionic bromination pathway from the correct radical pathway. N-Bromosuccinimide (NBS) with cyclohexene under radical initiation (AIBN or light) is the textbook case where students are tempted to draw an ionic-style (double-barbed) bromination mechanism by analogy with Br₂ addition, and get it wrong. The correct picture: initiation generates a small, steady-state concentration of Br• via homolysis (two fishhooks splitting a weak N–Br or Br–Br bond, exactly as in the Cl₂ derivation above — one fishhook to each fragment). Propagation step 1: Br• abstracts an allylic hydrogen from cyclohexene. Draw this with fishhooks: one fishhook from the allylic C–H bonding pair goes to pair up with Br•'s own single fishhook (converging into the new H–Br bond, a legitimate 2-electron bond built from two 1-electron fishhook contributions), and the second fishhook from that same C–H bond moves onto the allylic carbon, generating an allylic radical. This allylic radical is resonance-delocalized — but drawn with fishhooks, not double-barbed arrows, because the delocalizing species is a SOMO (one electron), moving between the two termini of the allyl system: one fishhook, tail on the C=C π bond, head moving to the adjacent carbon, shifting the radical position while the unpaired electron count stays at exactly one throughout. Propagation step 2: this allylic radical abstracts a Br atom from Br₂ (regenerated in low concentration from NBS + HBr equilibrium), again via two fishhooks converging into a new C–Br bond and one fishhook forming a new Br• that continues the chain. At no point in this entire sequence does an ion appear; drawing double-barbed arrows here (as if Br+ were adding electrophilically across the alkene, Markovnikov-style) describes the wrong mechanism — that ionic pathway is what happens with Br₂ alone in the dark in a polar solvent, a mechanistically and regiochemically distinct reaction (vicinal dibromide via bromonium ion, not allylic monobromide via radical chain).

Example 3 (exam/competition level): Distinguishing a single mechanism that contains both a heterolytic and a homolytic step — retro-ene-adjacent radical clock ring-opening used to probe SN1 vs SN2 character. Consider a mechanistic puzzle used to probe whether a solvolysis proceeds through a discrete cation (heterolytic, double-barbed throughout) or has any radical character: a cyclopropylcarbinyl-type leaving group is solvolyzed, and the "radical clock" cyclopropylcarbinyl-to-homoallyl rearrangement rate is compared against a control. Two candidate mechanisms must be drawn with strictly self-consistent arrows. Candidate A (pure heterolytic/cationic pathway). The C–OTs (tosylate) bond ionizes: one double-barbed arrow, tail on the C–O bond, head moving fully onto the tosylate oxygen (departing as the resonance-stabilized tosylate anion, its negative charge delocalized over three oxygens via further double-barbed arrows into the sulfonyl group). This generates a cyclopropylcarbinyl cation. Ring-opening to the homoallyl cation is then drawn with a second double-barbed arrow: tail on a cyclopropane C–C bonding pair (one of the two bonds flanking the cationic carbon, whose orbital is aligned for hyperconjugative overlap with the empty p orbital), head moving to form a new π bond between the migrating carbon and the original cationic center, while the ring carbon that lost the bond becomes the new cationic center. This is legitimate because the empty orbital receiving the pair is a true LUMO (p orbital of a carbocation, zero electrons), consistent with a two-electron, heterolytic-style bond reorganization even though no external nucleophile is involved — it is a cationic rearrangement, still entirely double-barbed, because the driving orbital interaction is a filled σ(C–C) donating into an empty p orbital, a two-electron process. Candidate B (radical pathway, if homolysis competed). If instead the C–LG bond broke homolytically (relevant when radical initiators or radical-generating conditions are present, or when testing whether trace radical character exists), the ring-opening step must use fishhooks: one fishhook from the strained cyclopropane C–C bond moves to pair with the existing radical's SOMO (extending conjugation/forming a new π-bond electron pair correctly requires care — in the genuine radical mechanism, one fishhook moves from the breaking C–C σ bond to combine with the adjacent carbon radical's unpaired electron, forming a new π bond, while the second fishhook from that same σ bond becomes the new, ring-opened radical center on the far carbon). The diagnostic power of the "radical clock" is precisely that cyclopropylcarbinyl radicals ring-open extremely fast (~10⁸ s⁻¹), so if products show the ring-opened (homoallyl) skeleton even under conditions expected to be purely ionic, or if the ratio of ring-opened to unopened product tracks known radical rate constants rather than cationic ones, the mechanism has radical character and must be redrawn with fishhooks, not double-barbed arrows, for at least that step. This is the real experimental use of the arrow-type distinction: chemists infer which arrow is correct from kinetics and product ratios, then must draw the mechanism to match, never the reverse.

The curved-arrow formalism, from which both the double-barbed and single-barbed conventions descend, was systematized in the late 1920s and 1930s by the English chemist Sir Robert Robinson and, independently and influentially in pedagogical form, by Christopher Ingold, who provided much of the rigorous electronic-theory language (nucleophile, electrophile, heterolysis, homolysis — Ingold coined several of these very terms) that makes the two arrow types meaningful rather than merely decorative. Ingold's systematic classification of reaction mechanisms (SN1, SN2, E1, E2) explicitly rested on distinguishing concerted, two-electron, heterolytic bond reorganizations from the free-radical chemistry being simultaneously developed by others; the fishhook notation for one-electron processes matured alongside the growing understanding of radical chain reactions (combustion, polymerization) through the 1930s–1950s, with major mechanistic clarity contributed by workers such as Frank Whitmore (carbocation rearrangements, firmly in the double-barbed/heterolytic camp) and Morris Kharasch (who established the radical, anti-Markovnikov mechanism of HBr addition to alkenes under peroxide conditions — precisely the historical experiment that forced organic chemists to formally separate homolytic from heterolytic addition and therefore to need two distinct arrow symbols in the first place).