The Allyl System

The allyl system's C1/C3 equivalence is not a museum piece confined to textbooks — it governs real synthetic and biological chemistry constantly. In synthesis, allylic substitution reactions (both classical $S_N1'$ ionization–recombination and modern transition-metal-catalyzed variants such as Tsuji–Trost palladium-catalyzed allylic alkylation) exploit exactly the symmetric charge distribution derived here: a $\pi$-allyl palladium complex is, structurally and electronically, almost a direct organometallic embodiment of the $\psi_1/\psi_2$ picture, with nucleophiles able to attack either terminus depending on steric and electronic biases layered on top of the intrinsically symmetric parent framework. In natural product and terpene biosynthesis, allylic cations generated from pyrophosphate ionization (for instance, from geranyl or farnesyl pyrophosphate) are central to essentially all terpenoid cyclization cascades; enzymes (terpene synthases) control which terminus of the resulting allylic cation is attacked by a subsequent nucleophile or alkene, using the enzyme active site to break the intrinsic symmetry that the isolated allyl cation would otherwise show. Radical allylic bromination (Wohl–Ziegler reaction with NBS) relies on the extra stability of the allylic C–H bond conferred by exactly the delocalization energy computed above, making allylic positions selectively reactive toward radical halogenation over unactivated alkyl positions elsewhere in a molecule. In materials and spectroscopy, the allyl radical's EPR hyperfine pattern (large equal terminal splittings, small central splitting) remains a standard diagnostic for identifying delocalized versus localized radical intermediates in mechanistic studies, and the general odd-alternant-hydrocarbon framework (allyl being the smallest member) extends directly to benzyl radicals/cations/anions, to the reactive intermediates of pericyclic reactions, and to conducting-polymer physics, where soliton-like defects in polyacetylene chains are themselves, at the level of a single repeat unit, allyl-like three-orbital problems with a symmetric nonbonding state.

Ask the tutor: How do the allyl coefficients change once a substituent breaks the mirror symmetry, and how does that predict which terminus reacts in an SN1' substitution?

Consider the allyl fragment C1–C2–C3, planar, each carbon $sp^2$-hybridized, each contributing one unhybridized $2p_z$ orbital perpendicular to the plane. This is structurally identical to how butadiene's four $p_z$ orbitals were set up, except now there are three centers instead of four, and — crucially — an odd number of $\pi$ electrons is possible (three, in the radical), which cannot happen in a system built from an even number of orbitals filled in pairs without leaving something unpaired. The three $2p_z$ orbitals, $\phi_1, \phi_2, \phi_3$ (on C1, C2, C3 respectively), combine linearly:

Because the framework has a mirror plane through C2 (swapping C1 ↔ C3), any eigenfunction of the Hamiltonian must be an eigenfunction of the reflection operator too — degenerate eigenvalues aside, which do not occur here. This means every solution has either $c_1 = c_3$ (symmetric) or $c_1 = -c_3$ (antisymmetric). That single geometric fact, before any arithmetic, already predicts equal $|c_1|=|c_3|$ in every orbital — the entire physical content of "allylic delocalization" is contained in it. The three resulting molecular orbitals stack up as bonding ($\psi_1$, no nodes between carbons), nonbonding ($\psi_2$, one node — right through C2), and antibonding ($\psi_3$, two nodes). Filling these with 2, 3, or 4 $\pi$ electrons gives the cation, radical, or anion respectively, and because $\psi_2$ is the frontier orbital in all three cases (either empty, half-filled, or filled), and $\psi_2$ has zero amplitude at C2, the chemistry of allylic species is dominated by an orbital that structurally excludes the central carbon.

Now overlay the electron count. The allyl cation has two $\pi$ electrons (from the original allylic C=C, since removing H⁻ from the sp³ carbon leaves an empty $p$ orbital that joins the $\pi$ system): both go into $\psi_1$, and $\psi_2,\psi_3$ are empty. The allyl radical has three $\pi$ electrons: $\psi_1$ full, $\psi_2$ singly occupied. The allyl anion has four: $\psi_1$ and $\psi_2$ both full. In every case the shape of the frontier density comes from $\psi_2$ (empty, half-full, or full) sitting on top of the always-full $\psi_1$, and since $\psi_2$ vanishes at C2, essentially all of the "interesting" charge or spin redistribution happens at C1 and C3.

The Hückel secular equations for a three-orbital, mirror-symmetric $\pi$ system reduce to a $3\times3$ determinant in the parameters $\alpha$ (Coulomb integral) and $\beta$ (resonance integral, connecting adjacent atoms only):

where: $\alpha$ is the energy of an electron in an isolated $2p_z$ orbital (same for every carbon by symmetry), $\beta$ is the stabilization from adjacent-orbital overlap (negative in sign, $\beta<0$), and off-diagonal zero (C1–C3) reflects the Hückel approximation that non-adjacent orbitals do not directly interact. Solving this cubic in $E$ gives three roots:

where: $E_1$ is the bonding level (lowest energy, since $\beta<0$ makes $\sqrt2\beta$ negative), $E_2$ is exactly the nonbonding, isolated-atom energy, and $E_3$ is the antibonding level (highest energy). The normalized coefficients that go with these energies are:

where: each $\psi_i$ is normalized ($\sum c^2=1$) and mutually orthogonal to the others, and the sign pattern of coefficients (all same sign, one sign flip, two sign flips) tracks directly with node count (0, 1, 2). The $\pi$-electron density at each carbon $r$ (summed over occupied MOs, each electron weighted by orbital occupation number $n_i$) is:

where: $n_i \in \{0,1,2\}$ is the occupation of orbital $i$, and $c_{ir}$ is the coefficient of atomic orbital $\phi_r$ in molecular orbital $\psi_i$. This single formula, applied with different $n_2$ (0, 1, or 2), generates the cation/radical/anion charge distributions directly. The total $\pi$-bond energy relative to three isolated $2p$ electrons is:

where: summing over occupied orbitals gives the total stabilization, and comparing $E_\pi$ to $n\alpha$ (the energy of $n$ non-interacting electrons) yields the delocalization (resonance) energy.

Example 1 (foundational) — Allyl cation from an SN1 ionization. 1-Chloro-2-butene, $\text{CH}_3\text{–CH=CH–CH}_2\text{–Cl}$, ionizes in a polar protic solvent to lose chloride and generate a substituted allyl cation. Set up the three $2p_z$ orbitals on C1 (=CH–CH₃ terminus, secondary-like), C2 (=CH–), and C3 (the original CH₂Cl carbon, now CH₂⁺-like). Because the leaving group departs from an $sp^3$ carbon adjacent to a $\pi$ bond, the empty $p$ orbital left behind aligns with the existing $\pi$ system, generating a genuine 3-orbital, 2-electron allylic cation — not a localized primary carbocation. By the coefficients derived above, the HOMO ($\psi_1$) places charge density $+\tfrac12$ at C1 and $+\tfrac12$ at C3 (using the parent, unsubstituted coefficients as a first approximation), with zero at C2. Nucleophilic attack (e.g., by water or the departed chloride recombining) can occur at either terminus: attack at C3 regenerates the starting connectivity, while attack at C1 gives the "crossed" product 3-chloro-1-butene — the defining signature of an Allylic Substitution and SN1' Reactions pathway. The key qualitative point, checked against the exact math: it is $\psi_1$'s coefficients (both termini equal in magnitude) that dictate this 50:50-type reactivity in the unsubstituted case, and substituent effects (here, a methyl at C1) perturb which terminus is favored without changing the fundamental delocalization pattern.

Example 2 (applied) — Bond lengths in the allyl anion versus a localized diene. Allyl anion (from deprotonating propene with a strong base such as $n$-BuLi, or observed directly as the lithium salt) shows two C–C bonds of essentially equal length, roughly 1.39–1.40 Å, intermediate between a typical C–C single bond (1.54 Å) and C=C double bond (1.34 Å). This is a direct structural consequence of the $\pi$-bond-order formula built from the same coefficients:

For the anion ($n_1=2,n_2=2$): $c_{1,1}c_{1,2} = \tfrac12\cdot\tfrac{1}{\sqrt2} = \tfrac{1}{2\sqrt2}$, contributing $2\times\tfrac{1}{2\sqrt2}=\tfrac{1}{\sqrt2}$; and $c_{2,1}c_{2,2} = \tfrac{1}{\sqrt2}\cdot 0 = 0$, contributing nothing from $\psi_2$. So:

By the identical calculation, $p_{23}=0.707$ as well (using $c_{1,2}c_{1,3}=\tfrac{1}{\sqrt2}\cdot\tfrac12$, doubled, plus the $\psi_2$ contribution $0\cdot(-\tfrac{1}{\sqrt2})=0$). Both allylic C–C bonds thus carry identical bond order 0.707 — exactly halfway between a pure single bond (order 0, in this normalized scale relative to $\pi$-only contribution) and a pure double bond (order 1), consistent with the crystallographically observed near-equal, intermediate bond lengths. This is the mechanistic reason X-ray and electron-diffraction structures of allyl anion salts show two indistinguishable "1.5 bonds," not one short and one long bond as a single Lewis structure would suggest.

Example 3 (exam/competition level) — EPR hyperfine pattern of the allyl radical. The allyl radical, generated for instance by photolysis of a peroxide in the presence of propene (H-atom abstraction) or by direct photolysis of 1,5-hexadiene, gives an EPR spectrum with hyperfine coupling to four equivalent terminal protons (two on each of C1 and C3) and a much smaller coupling to the two protons on C2. Predict the qualitative spin-density pattern and rationalize the near-zero splitting from the central protons. Reasoning. The unpaired electron in the allyl radical occupies $\psi_2$ (the SOMO), with coefficients $(c_1,c_2,c_3) = (\tfrac{1}{\sqrt2},0,-\tfrac{1}{\sqrt2})$. Spin density at each carbon is $\rho_r = c_{2,r}^2$ (only the singly-occupied orbital contributes to net spin, since $\psi_1$ is doubly filled and contributes zero net spin): $\rho_1 = (\tfrac{1}{\sqrt2})^2 = 0.5$, $\rho_2 = 0$, $\rho_3 = (\tfrac{1}{\sqrt2})^2 = 0.5$. These sum to $1.0$, correctly accounting for exactly one unpaired electron. The McConnell relation connects spin density on a carbon to the hyperfine coupling constant of an attached hydrogen:

where: $Q \approx -22.5\text{ to }-27\text{ G}$ is an empirically calibrated proportionality constant (from fitting to systems of known spin density), and $\rho_C$ is the $\pi$-spin density at the carbon bearing that hydrogen. Using $Q\approx-24$ G and $\rho_1=\rho_3=0.5$:

in good agreement with the experimentally observed allyl radical terminal-proton coupling of roughly $13.9$ G (the small discrepancy reflects the approximate nature of simple Hückel coefficients and of $Q$ itself). For the central proton, since $\rho_2=0$ from the nodal structure of $\psi_2$, the direct McConnell contribution vanishes; the small residual coupling actually observed for the C2 proton (~4 G) arises from higher-order effects (spin polarization of the C–H $\sigma$ bond by the adjacent $\pi$ spin density, not captured at the simple Hückel level) rather than any density predicted by $\psi_2$ itself. This is exactly the qualitative fingerprint used to identify allylic (versus localized, non-delocalized) radicals from EPR spectra: two large, equal terminal splittings and one small central splitting is the signature of $\rho_2=0$ forced by mirror symmetry.

The Hückel molecular orbital method used throughout this article was developed by Erich Hückel in the early 1930s, originally to explain the special stability of benzene and, more broadly, to give a tractable quantum-mechanical account of conjugated and aromatic $\pi$ systems using only the "top" $\pi$ electrons while treating the $\sigma$ framework as a fixed scaffold. His approach — reduce a many-electron problem to a simple, solvable one-electron eigenvalue problem for each conjugated system — became the template for essentially all qualitative organic MO reasoning that followed, including the allyl and butadiene treatments given here, and it was later systematized and popularized for organic chemists largely through the work of Roald Hoffmann and others in the extended Hückel and frontier-orbital traditions.