The Scholl Reaction

What the Scholl reaction does

Almost every carbon–carbon bond-forming reaction you learn first — Grignard, aldol, Suzuki, Wittig — needs a handle: a halide, a carbonyl, a boronic acid, a metal. The Scholl reaction needs none of that. It takes two aromatic rings sitting near each other, each with an ordinary, unactivated C–H bond, and it welds them together directly:

    Ar–H  +  H–Ar′   ──[oxidant]──→   Ar–Ar′   +   2 H⁺  +  2 e⁻

Two C–H bonds vanish, one new C–C bond appears, and two electrons are stripped away by an oxidant. Because the two rings are usually already tethered together in the same molecule (say, two phenyl groups hanging off the same biphenyl core), the "coupling" is really an intramolecular ring closure — a cyclodehydrogenation. Do it once and a bay region snaps shut into a new six-membered ring. Do it many times on a big pre-built scaffold and a floppy tangle of benzene rings collapses into a single rigid, perfectly flat sheet of fused aromatic carbon — a piece of molecular graphene.

That is the whole appeal. One oxidant, applied to the right precursor, can form six new C–C bonds in a single flask and turn hexaphenylbenzene into a disc of thirteen fused rings. No other reaction does bottom-up graphene chemistry this cheaply.

The mechanism — arrows, electrons, and a radical cation

The Scholl reaction has been argued about for a century, and the honest answer is that it runs by two closely related pathways depending on the substrate and the oxidant. For the electron-rich polycyclic precursors used in materials chemistry — the ones that actually matter — the dominant route is the radical-cation (electron-transfer) mechanism:

1. First one-electron oxidation. The oxidant (say FeCl₃) removes a single electron from the arene's π system, giving an arene radical cation Ar•⁺. This species is both a radical and a cation — it is far more electrophilic than the neutral arene, and its spin density is highest at the ring-fusion carbon.
2. Intramolecular C–C bond formation. The electron-rich neighboring ring attacks that electrophilic carbon. Its π electrons swing in to form the new σ bond, producing a σ-complex — a distonic species that is a protonated dihydro-arene radical cation. This is the arenium-like intermediate, with an sp³ carbon bearing the "extra" H at each newly joined position.
3. First deprotonation. A base (chloride, or the conjugate base of the acid) pulls off one of the two sp³ protons, rearomatizing that carbon and leaving a neutral radical delocalized over the new ring system.
4. Second one-electron oxidation. The oxidant removes a second electron, regenerating a cation at the remaining sp³ center.
5. Second deprotonation. Loss of the last proton restores full aromaticity across both fused rings. The biaryl bond is now locked in, and the ring system is flat.

    Ar–H⋯H–Ar′        (neutral, two rings held close)
        │  – e⁻   (oxidant: FeCl₃ → FeCl₂ + Cl⁻)
        ▼
    [Ar–H⋯H–Ar′]•⁺     radical cation — electrophilic at the fusion C
        │  C–C bond forms (ring attacks radical-cation carbon)
        ▼
    σ-complex•⁺         two sp³ C–H's at the new junction, delocalized radical cation
        │  – H⁺
        ▼
    dihydro radical
        │  – e⁻  ,  – H⁺
        ▼
    Ar–Ar′             fully aromatic, planar, +1 new ring

The competing arenium (proton-first) mechanism starts instead with protonation of the arene by a strong acid to give a Wheland-type arenium ion, which is then attacked by the neighboring ring; two sequential losses of H⁺ (assisted by the oxidant re-aromatizing) give the same product. This route dominates under super-acidic, less-oxidizing conditions and for more electron-poor arenes. The practical fingerprint that tells the two apart: the radical-cation route shows a substantial kinetic isotope effect and correlates with substrate oxidation potential, whereas the arenium route shows methyl/methoxy migration (1,2-shifts of the cationic intermediate) that scrambles substituents. Both converge on the same fused product — which is why the Scholl reaction is so robust in practice even when the mechanism is murky.

Reagents, oxidants, and real conditions

Every Scholl reaction is a balance between an oxidant strong enough to reach the arene's oxidation potential and mild enough not to chlorinate, over-oxidize, or rearrange the substrate. The common choices:

• FeCl₃ (anhydrous). The workhorse for nanographenes. Typically 2–3 equivalents of FeCl₃ per new C–C bond, in dry dichloromethane, often with a few drops of nitromethane to solubilize the iron and boost the oxidizing power. Runs at 0–25 °C over minutes to hours. Cheap, scalable, but liberates HCl and can chlorinate very electron-rich rings. A single hexa-peri-hexabenzocoronene closure (six bonds) can call for ~12–18 equivalents of FeCl₃.
• DDQ + strong acid. 2,3-Dichloro-5,6-dicyano-1,4-benzoquinone with triflic acid (TfOH), methanesulfonic acid, or BF₃·OEt₂. The DDQ/acid pair (the Rathore conditions) is milder, gives cleaner mass balance, and largely avoids the chlorination and migration problems of FeCl₃. Preferred for delicate or valuable substrates.
• MoCl₅ or Mo(V)/Mo(VI) systems. Strong one-electron oxidants, effective for higher-oxidation-potential arenes that FeCl₃ cannot touch.
• Hypervalent iodine. PIFA [PhI(OCOCF₃)₂] or PhI(OTf)₂ with BF₃·OEt₂ — metal-free, useful when trace iron contamination is unacceptable (e.g. for electronic-grade materials).
• Classic Scholl conditions. The original: molten AlCl₃/NaCl eutectic at 120–180 °C, or AlCl₃ in an inert solvent. Harsh, low-yielding by modern standards, and now mostly of historical interest.

A universal caution: the reaction generates protons (2 H⁺ per bond), so an acid-tolerant setup and a proton acceptor (or simply excess oxidant) keep it turning. Water and oxygen must usually be excluded because the radical-cation intermediates are quenched by nucleophiles.

Scope, selectivity, and planarity

The Scholl reaction is not a general "couple any two arenes" tool. Three requirements govern whether it works:

• Geometry. The two rings must be able to reach a bay-region contact where a six-membered ring can close. Intramolecular closures with the two reacting carbons already held ~2.5–3.5 Å apart are strongly favored; genuinely intermolecular Scholl couplings are rare and low-yielding.
• Electron density. Electron-rich arenes (alkyl-, alkoxy-substituted, or extended polycyclics with low oxidation potential) react readily. Electron-poor rings bearing nitro, cyano, or carbonyl groups resist oxidation and typically fail — the same deactivation logic that governs electrophilic aromatic substitution.
• Thermodynamic payoff. Every closure gains aromatic stabilization by extending the conjugated π system, and it relieves the steric strain of crowded ortho-phenyl groups. That downhill drive is what lets a single reaction snap shut many bonds cooperatively.

There is no stereochemistry to speak of in the classic sense — the products are flat, planar aromatics. The relevant "selectivity" is regioselectivity: which of several possible bay regions closes. For symmetric precursors like hexaphenylbenzene this is a non-issue (all six closures are equivalent). For less symmetric substrates, substituent migration and competing closure modes can give isomer mixtures, which is the practical Achilles' heel of the method.

Scholl vs the metal-catalyzed biaryl couplings

	Scholl reaction	Suzuki / Ullmann coupling
Bond made	Ar–Ar (biaryl)	Ar–Ar (biaryl)
Handles needed	None — plain C–H on both rings	Halide + boronic acid (Suzuki); two halides (Ullmann)
Driving force	Oxidant removes 2 e⁻ + 2 H⁺	Pd⁰/Cu redox cycle
Catalyst	None (stoichiometric oxidant)	Pd or Cu catalyst
Typical reagent	FeCl₃, DDQ/TfOH, MoCl₅	Pd(PPh₃)₄ + base; Cu + ligand
Regiocontrol	Set by substrate geometry only	Set precisely by where you put the halide
Bonds per operation	Many at once (up to 6+ in one flask)	One at a time
Best for	Fusing crowded rings into flat nanographenes	Joining two distinct fragments with a defined connectivity
Main weakness	Substituent migration, isomer scrambling	Cost, halide pre-installation, one bond per step

The two are complementary, and modern nanographene syntheses use both: Suzuki or Diels–Alder polymerization first lays down the connectivity of a polyphenylene precursor, then a single Scholl reaction fuses all the rings that geometry has pre-aligned.

Worked example: hexaphenylbenzene → hexabenzocoronene

The textbook demonstration of the Scholl reaction's power is the one-step conversion of hexaphenylbenzene into hexa-peri-hexabenzocoronene (HBC) — a disc of thirteen fused benzene rings and 42 carbon atoms, the smallest "nanographene." Six new C–C bonds form in a single operation:

    hexaphenylbenzene (C₄₂H₃₀)   ──FeCl₃ (excess), CH₂Cl₂ / MeNO₂, RT──→
                                    HBC (C₄₂H₁₈)   +   6 [2 H⁺ + 2 e⁻]

    net:  C₄₂H₃₀  →  C₄₂H₁₈  +  12 H⁺ + 12 e⁻   (six cyclodehydrogenations)

• Precursor. Hexaphenylbenzene, itself made by cobalt- or Ni-catalyzed [2+2+2] cyclotrimerization of a diaryl-acetylene, or by Diels–Alder of a tetraphenylcyclopentadienone with an alkyne. The six pendant phenyls are held in a crowded "paddle-wheel" that badly wants to flatten.
• Oxidant. Anhydrous FeCl₃, roughly 12–18 equivalents (2–3 per bond), in CH₂Cl₂ with a little nitromethane. DDQ/TfOH works too and is cleaner.
• Conditions. 0–25 °C, minutes to a couple of hours, argon atmosphere, dry.
• Workup. Quench into methanol/water; the insoluble HBC disc precipitates. Reported yields under optimized FeCl₃ conditions exceed 80%.

Because all six bay regions are equivalent, there is no regiochemistry problem — the crowded propeller simply snaps flat. HBC and its alkyl-substituted derivatives self-assemble into columnar liquid-crystalline stacks and are model discotic semiconductors, which is why the Müllen group built a whole family of them this way.

Real application: bottom-up graphene nanoribbons

The headline modern use is the bottom-up synthesis of atomically precise graphene nanoribbons (GNRs). Lithographically cut graphene has ragged edges that scatter electrons; chemistry can do better. The recipe:

1. Build a polyphenylene precursor. An A₂B₂-type Diels–Alder polymerization (or Suzuki/Yamamoto polycondensation) makes a long, non-planar polymer whose backbone carries dozens of pendant benzene rings in a defined pattern — but not yet fused.
2. Zip it shut with one Scholl reaction. Treat the polymer with a large excess of FeCl₃. In a single operation, every geometrically pre-aligned pair of rings undergoes cyclodehydrogenation. The floppy chain collapses into a rigid, flat, armchair-edge graphene nanoribbon — with the width and edge structure dictated exactly by the monomer design.

Müllen, Fasel, and co-workers used this solution-phase route (and an on-surface variant, where the Scholl-type fusion is driven thermally on a Au(111) surface) to make ribbons up to hundreds of nanometers long with defined band gaps. The on-surface version, reported in Nature in 2010, produces defect-free armchair GNRs one atom-row at a time and is now a standard technique in nanoelectronics labs. In both cases the ring-fusing chemistry is the Scholl cyclodehydrogenation — it is the reaction that turns "a chain of benzene rings" into "a strip of graphene."

Limitations & side reactions

• Substituent migration. With methoxy or other strong donors at the reacting positions, the cationic intermediate undergoes 1,2-shifts, scrambling substituents and giving unexpected regioisomers. Switching to DDQ/acid, using alkyl instead of alkoxy solubilizers, or adding a tert-butyl blocking group are the standard fixes.
• Chlorination. FeCl₃ can act as a chlorinating agent on very electron-rich rings, installing unwanted Cl atoms. Nitromethane co-solvent and controlled stoichiometry suppress it; DDQ/acid avoids it entirely.
• Over-oxidation and ring-opening. Excess strong oxidant can push past the desired product to quinones or fragmented material.
• Failure on electron-poor arenes. Nitro, cyano, and carbonyl groups raise the oxidation potential above what FeCl₃ can reach; those substrates simply don't react.
• Incomplete closure on large precursors. In big GNR syntheses, a few bay regions can be left un-fused ("defects"), which is why on-surface, thermally driven cyclodehydrogenation is preferred for the highest-purity ribbons.
• Solubility collapse. As the product flattens, its π–π stacking makes it dramatically less soluble — which is great for isolation by precipitation but can trap unreacted intermediates inside aggregates.

Historical discovery — who and when

The reaction is named for Roland Heinrich Scholl (1865–1945), a Swiss-German organic chemist. In 1910, working with aluminium chloride at high temperature, Scholl and co-workers observed that heating two aromatic rings with AlCl₃ formed a direct bond between them with loss of hydrogen — most famously converting perylene-related and naphthalene systems into fused polycyclics. The transformation was formalized in his papers through the 1910s–1920s (Scholl, Seer, Weitzenböck), and the "loss of two aryl hydrogens to form a biaryl bond under Lewis-acid/oxidative conditions" became known as the Scholl reaction.

For most of the twentieth century it was a niche, capricious reaction with poor yields. Its renaissance came in the 1990s and 2000s when Klaus Müllen's group at the Max Planck Institute for Polymer Research recognized that the modern oxidant FeCl₃ (and later DDQ/acid) turned Scholl chemistry into a reliable, scalable engine for making nanographenes, hexabenzocoronenes, and graphene nanoribbons. That reframing — from historical curiosity to the central C–C bond of bottom-up graphene synthesis — is why the reaction is taught and used today.

Materials & safety notes

• Corrosive by-products. FeCl₃ conditions release HCl gas; run in a fume hood with a scrubber, and keep glassware and reagents rigorously dry — FeCl₃ hydrolyzes and loses activity on contact with moisture.
• DDQ toxicity. DDQ is a strong oxidant and an irritant; its reduced form (DDQH₂) and the acid co-reagents (TfOH, MsOH, BF₃·OEt₂) are corrosive. Handle with standard acid/oxidizer precautions.
• Metal-free grades. For electronic-grade nanographenes and OLED materials, trace iron from FeCl₃ is a device poison; hypervalent-iodine (PIFA) or DDQ conditions are chosen specifically to keep the product metal-free.
• Why it matters industrially. Scholl chemistry underpins research-scale production of discotic semiconductors, organic-electronics dyes, and the graphene nanoribbons being explored for next-generation transistors — the reaction's atom economy (many bonds per flask, no pre-installed halides) is a large part of why it scales.