Halogen Bonding

The sigma-hole: why a halogen turns electrophilic

A halogen covalently bonded to carbon looks, in a simple Lewis picture, like it should be all lone pairs and negative charge. But the real electron density is anisotropic. Along the extension of the C–X bond axis, at the “back” of the halogen opposite the bond, the electrostatic potential is actually positive. This region is called the σ-hole (a term coined by Timothy Clark and Peter Politzer around 2007).

The origin is the depopulation of the p_z orbital that points along the C–X bond: putting one electron of that orbital into the bonding pair leaves reduced density on the outer lobe. Perpendicular to the bond, the two occupied p orbitals form an electron-rich equatorial belt (negative potential). So a bound halogen is simultaneously a Lewis base at its sides and a Lewis acid at its tip. The σ-hole lines up a Lewis base collinearly with the C–X bond, which is why R–X···B angles cluster near 180°.

The σ-hole grows more positive as the halogen becomes larger and more polarizable (I > Br > Cl > F) and as electron-withdrawing groups on R drain density from X. Fluorine is so electronegative and hard that its σ-hole is negligible; C–F essentially does not form halogen bonds. This is the opposite of the periodic trend one might naively expect and is the single most important idea in the field.

What controls the strength

Halogen-bond strength is engineered by two knobs:

• Choice of halogen. Larger, softer halogens have bigger, more positive σ-holes. A typical ranking is I > Br > Cl, with iodine bonds commonly the strongest and most exploited.
• Backbone activation. Attaching X to an electron-poor scaffold amplifies the σ-hole. Perfluoroalkyl and perfluoroaryl iodides (e.g. 1-iodoperfluorooctane, or 1,4-diiodotetrafluorobenzene, C₆F₄I₂) are workhorse donors precisely because the fluorines withdraw density and deepen the hole.

On the acceptor side, strength grows with Lewis basicity: anions (halides, especially I⁻) and amine nitrogens give the strongest bonds, followed by pyridine-type N, carbonyl O, and π systems. The strongest neutral halogen bonds (e.g. to pyridines) reach roughly 30 kJ mol⁻¹; charge-assisted halogen bonds to anions can be substantially stronger. Because the interaction is dominated by electrostatics plus charge transfer (with a real dispersion component from the polarizable halogen), it is well captured by both electrostatic-potential analysis and natural-bond-orbital donation from the base lone pair into the C–X σ* antibonding orbital.

Geometry, and how it differs from a hydrogen bond

The defining signature is a short, linear contact. The halogen···base distance falls inside the sum of van der Waals radii, often by 0.2–0.5 Å, and the R–X···B angle sits within about 10–15° of 180°. In contrast, the base approaches the halogen at roughly 90° on the equatorial belt only when acting as a donor to something else — a reminder of the halogen's dual character.

Compared with a hydrogen bond, a halogen bond is:

• More directional. The σ-hole is a tight cap, so the geometry is tighter than a typical N–H···O bond.
• Hydrophobic on the donor side. The R–X donor is greasy, so halogen bonds can operate in lipophilic pockets where hydrogen bonds are disfavored.
• Tunable through the periodic table, letting a chemist dial strength by swapping Cl → Br → I without changing the rest of the molecule.

These distinctions are why halogen bonding is treated as a complementary, orthogonal design element rather than a mere copy of hydrogen bonding.

History: from Hassel's diffraction to a named interaction

The phenomenology predates the name by more than a century. In 1863 Frederick Guthrie prepared what we now recognize as an ammonia–iodine charge-transfer adduct. The structural breakthrough was Odd Hassel's 1954 single-crystal X-ray study of the 1:1 complex between molecular bromine and 1,4-dioxane, which showed a strikingly linear O···Br–Br arrangement — the geometric fingerprint of what we now call a halogen bond. Hassel's broader conformational work earned the 1969 Nobel Prize in Chemistry (shared with Derek Barton).

Robert Mulliken framed such contacts within charge-transfer theory in the 1950s. But the systematic, deliberate use of the interaction, and the crisp name “halogen bond,” emerged in the late 1990s and 2000s from the groups of Metrangolo and Resnati, who showed it could organize crystals as reliably as hydrogen bonding. IUPAC codified a formal definition in 2013, cementing halogen bonding as a recognized member of the noncovalent-interaction family alongside hydrogen bonds and van der Waals forces.

Applications: crystal engineering, drugs, and catalysis

Crystal engineering & materials. Because the interaction is strong and predictable, ditopic donors like 1,4-diiodotetrafluorobenzene co-crystallize with ditopic bases (bipyridines, dinitriles) to give designed 1-D chains and networks. Halogen bonding is used to build co-crystals, tune solubility and melting point of pharmaceuticals, template porous solids, orient nonlinear-optical chromophores, and drive anion recognition in synthetic receptors and sensors.

Medicinal chemistry. Roughly a quarter of marketed small-molecule drugs and a large fraction of leads carry at least one halogen, and many form halogen bonds to protein backbones and side chains. A classic example is the thyroid hormone system, where thyroxine's iodines make halogen bonds; the anticoagulant, kinase-inhibitor, and antiviral literature exploit Cl/Br/I contacts to backbone carbonyl oxygens to gain affinity and selectivity in hydrophobic pockets where a hydrogen bond would not fit.

Catalysis. A newer frontier is halogen-bond-donor organocatalysis, in which cationic or highly fluorinated iodine/bromine donors activate substrates by abstracting or polarizing a Lewis-basic leaving group — a Lewis-acid-like role analogous to hydrogen-bonding thiourea catalysts. Multidentate iodotriazolium and iodoperfluoroarene catalysts accelerate reactions such as halide abstraction from benzhydryl halides and certain Michael and Mannich-type additions, and anion-binding halogen-bond catalysts are an active research area.

How to identify one experimentally

Halogen bonds are diagnosed by a convergence of evidence rather than a single test:

• Crystallography. A short, near-linear R–X···B contact well inside the van der Waals sum is the gold standard; the Cambridge Structural Database contains tens of thousands of such contacts.
• Spectroscopy. Forming the bond red-shifts and weakens the C–X stretch in the IR/Raman (mirroring the O–H red shift of a classical hydrogen bond) and shifts NMR resonances of both partners; ¹²⁹Xe and halogen-nucleus NQR/NMR can probe the donor directly.
• Thermodynamics. Titrations (UV-vis, NMR, ITC) give association constants that follow the I > Br > Cl and electron-poor-backbone trends predicted by the σ-hole model.

A useful mental check: if a halogen sits closer to an electron donor than van der Waals radii allow, and does so along the C–X axis, you are almost certainly looking at a halogen bond rather than a random crystal-packing contact.

Halogen bond vs hydrogen bond

Property	Halogen bond (R–X···B)	Hydrogen bond (D–H···A)
Electrophilic atom	Halogen (Cl, Br, I)	Hydrogen
Origin of attraction	σ-hole; electrostatics + charge transfer	Electrostatics; positive H
Directionality	Very high (≈180°, linear)	High but more flexible
Strength trend	I > Br > Cl (F essentially none)	F–H > O–H > N–H
Tunability	Tune by halogen + backbone electron-withdrawal	Tune by donor/acceptor pKₐ
Hydrophobicity	Donor is hydrophobic	Donor is polar/hydrophilic