Reactive Intermediates: Overview and Classification

Carbocation chemistry underlies petroleum refining at industrial scale: acid-catalyzed alkylation and isomerization processes that upgrade gasoline octane rating proceed through exactly the hyperconjugatively stabilized tertiary carbocations described here, generated on solid acid catalysts (zeolites, sulfated zirconia) that mimic Olah's superacid environments on an industrial surface. The same electron-deficient carbon chemistry governs Friedel–Crafts alkylations and acylations used to build pharmaceutical intermediates and fine chemicals. Carbanion chemistry is the backbone of nearly all carbon–carbon bond-forming methodology in synthetic organic chemistry: enolate alkylations, aldol reactions, Claisen condensations, and organolithium/Grignard additions all depend on precisely the resonance and inductive stabilization arguments developed here to predict which proton is removed and which carbon attacks an electrophile — directly setting up quaternary stereocenters in total synthesis of natural products and drug candidates. Radical intermediates now sit at the center of modern photoredox catalysis, used throughout the pharmaceutical industry to form C–C and C–N bonds under exceptionally mild conditions, and radical polymerization (free-radical and controlled/living radical methods like ATRP and RAFT) manufactures the majority of commodity and specialty polymers used industrially. Radical stability arguments here connect forward directly to Bond Dissociation Energies and Radical Stability, the quantitative backbone of that entire field, and to combustion and atmospheric chemistry, where radical chain propagation governs everything from engine knock to stratospheric ozone depletion. Carbenes and their metal-stabilized carbenoid cousins are indispensable in modern catalysis: N-heterocyclic carbenes (NHCs), persistent singlet carbenes stabilized by adjacent nitrogen lone pairs, are now among the most widely used ligands in transition-metal catalysis (Grubbs-generation olefin metathesis catalysts, Buchwald–Hartwig-type cross-couplings) and in organocatalysis (benzoin and Stetter reactions). Simmons–Smith cyclopropanation, exploiting a carbenoid's singlet-like stereospecific behavior, remains a standard method for stereocontrolled cyclopropane synthesis in total synthesis, and diazo-carbene C–H insertion chemistry (Rh- and Ru-catalyzed) is now a mainstream late-stage functionalization tool in medicinal chemistry.

Ask the tutor: How does the singlet/triplet carbene distinction connect quantitatively to bond dissociation energies when we start ranking radical stability numerically?

Start from a generic carbon with four substituents in a tetrahedral $sp^3$ arrangement — methane's geometry. A reactive intermediate is generated whenever one bond to that carbon is broken without immediately being replaced by a new bond of the same type. What differs between the four classes is what is left behind in the nonbonding space once that bond is gone.

Carbocation. Remove a group along with both bonding electrons (heterolysis, electrons leave with the leaving group). Carbon is left with only three bonds and zero nonbonding electrons at that center. To minimize energy, the carbon rehybridizes from $sp^3$ to $sp^2$: the three remaining substituents spread into a trigonal plane (120° angles), each $sp^2$ orbital making a strong $\sigma$ bond, and the leftover, unhybridized $2p_z$ orbital is left perpendicular to that plane — empty. This empty $p$ orbital is the entire chemical personality of a carbocation: it is a potent, geometrically exposed electrophile, and anything that can donate electron density sideways into it (a lone pair, a $\pi$ bond, or a well-aligned $\sigma_{C-H}$ bond) stabilizes the cation enormously.

Carbanion. Remove a group but let both bonding electrons stay on carbon (heterolysis, electrons stay). Now carbon has three bonds and a full lone pair — four electron domains, just like an amine nitrogen. VSEPR says: stay (or become) $sp^3$, pyramidal, lone pair occupying one vertex of a tetrahedron, rapidly inverting through a planar $sp^2$ transition state (much like ammonia) unless geometrically locked. Because the lone pair is what defines this species, carbanion stability improves whenever that pair can be pulled away from carbon — by electronegative neighbors (induction) or by delocalization into an adjacent $\pi$ system (resonance, as in an enolate).

Radical. Break the bond symmetrically (homolysis): each fragment keeps one electron. The carbon left behind has three bonds and one unpaired electron. Geometrically it sits between the cation and the anion — pyramidalization is shallow because a half-filled orbital gives much less angle strain preference than a full lone pair, so many simple alkyl radicals are close to planar or only slightly pyramidal, best described as $sp^2$ with the unpaired electron in a $p$ orbital (with rapid inversion if any pyramidal character exists). This half-filled $p$ orbital can act as a weak acceptor and a weak donor simultaneously, which is why radical stabilization (again hyperconjugation and resonance) works in the same direction as cation stabilization — electron-donating alkyl groups stabilize both — but far more weakly, because a half-filled orbital's appetite for extra density is much smaller than a truly empty one.

Carbene. Here a whole substituent's worth of both bonding electrons doesn't leave — instead, we remove two different substituents (formally, take a $CH_2$ fragment out of a bond and detach both its remaining bonds), leaving one carbon with only two $\sigma$ bonds and two nonbonding electrons. Two electrons in two available nonbonding orbitals (one roughly $sp^2$-like in-plane orbital, one unhybridized $p$ orbital perpendicular to the plane) can arrange two ways: pair up in the lower-energy in-plane orbital (singlet carbene, bent, ~100–105°, empty $p$ orbital, full $sp^2$-like lone pair — simultaneously electrophilic and nucleophilic) or spread out with parallel spins, one electron in each orbital (triplet carbene, more linear, ~130–150°, behaves like a 1,1-diradical). Which state is the ground state depends delicately on substituents, and this single choice controls the stereochemical outcome of the carbene's most famous reaction, cyclopropanation.

Now overlay stability trends onto this geometry. For carbocations, stabilization is a race to fill that empty $p$ orbital — alkyl groups help through hyperconjugative donation of $\sigma_{C-H}$ (or $\sigma_{C-C}$) electron density, and full stabilization comes from adjacent lone pairs (oxygen, nitrogen) or $\pi$ systems (allyl, benzyl) that can conjugate directly. For carbanions, stabilization is the opposite race — pulling the lone pair's density away from carbon, done by electronegative atoms (induction) or adjacent $\pi$-acceptors (resonance into a carbonyl, nitro group, or aromatic ring). For radicals, the same donors that help cations help radicals, but weakly, because a half-filled orbital's electron demand is intermediate. For carbenes, the two rules coexist on the same atom: substituents with lone pairs (like the oxygens on alkoxycarbonyl-, or the halogens on dihalocarbenes) donate into the empty $p$ orbital while simultaneously withdrawing inductively from the filled orbital — a push-pull effect that strongly favors the singlet state.

The unifying quantity behind every stability ranking here is not a single equation but a consistent bookkeeping of where electron density sits relative to an electron-deficient or electron-rich center, formalized through resonance energy and hyperconjugative stabilization. Three relationships do essentially all the work.

Cation/radical stabilization by hyperconjugation (donation from a filled $\sigma_{C-H}$ or $\sigma_{C-C}$ into the empty or half-filled $p$ orbital) is maximized when the donor orbital and the acceptor $p$ orbital are parallel:

where: the numerator is the overlap-weighted interaction (larger for more, better-aligned C–H/C–C bonds — hence 3° > 2° > 1° > methyl), and $\Delta \varepsilon$ is the energy gap between donor $\sigma$ and acceptor $p$ (smaller gap → stronger mixing). Reading it: more adjacent C–H/C–C bonds, properly aligned, means more stabilization — this is the entire orbital reason tertiary cations beat secondary beat primary beat methyl.

Carbanion/radical stabilization by resonance delocalization into a $\pi$ acceptor is captured by the standard resonance stabilization energy,

where: the delocalized structure mixes multiple valid Lewis contributors (e.g., enolate O⁻ ↔ C=C–O structures), and the energy difference is always negative (stabilizing) because mixing two nearly degenerate structures always lowers the ground state below either alone. Reading it: whenever you can draw a second legitimate resonance structure that moves the charge or radical onto a more electronegative atom or into a larger $\pi$ system, the intermediate is more stable than the single-structure picture suggests.

Singlet–triplet gap of a carbene, which decides its entire reactivity profile, is set by the competition between exchange stabilization of parallel electrons and the energy cost of populating a higher orbital:

where: $\Delta\varepsilon_{\sigma-p}$ is the orbital energy gap between the in-plane lone-pair-like orbital and the perpendicular $p$ orbital (widened by $\pi$-donor substituents like halogens or alkoxy groups, and by geometric bending), and $K_{\text{exch}}$ is the exchange stabilization gained by having two electrons with parallel spin in different orbitals (favors the triplet, roughly constant, largest when the two orbitals are close in energy and diffuse, as for simple alkyl carbenes). Reading it: a large orbital gap (pushed apart by π-donors and by bending) favors the singlet; a small gap lets the exchange term win and gives a triplet ground state.

Example 1 (foundational) — Ranking three carbocations by hyperconjugative donor count. Rank the stability of the following cations, all derived by loss of Cl⁻ from the corresponding chloride: (a) $(CH_3)_3C^+$ (tert-butyl), (b) $CH_3CH_2^+$ (ethyl), (c) $(CH_3)_2CHCH_2^+$... but let's instead use a cleaner asymmetric triad to force real counting rather than memorized labels: 2-chloro-2-methylbutane → cation A, 2-chlorobutane (at C2) → cation B, 1-chlorobutane → cation C. Step 1. Draw each resulting cation and identify the number of adjacent C–H (and C–C) bonds available for hyperconjugation into the empty $p$ orbital. Cation A, $(CH_3)(CH_3CH_2)(CH_3)C^+$ — a tertiary cation — has three separate alkyl substituents, contributing $3\times3 = 9$ adjacent hyperconjugating C–H bonds (each methyl offers 3, treating the ethyl's $CH_2$ methyls similarly gives comparably many donors; the key structural fact is three separate carbon substituents, each supplying multiple aligned C–H bonds). Step 2. Cation B, $CH_3$–$CH^+$–$CH_2CH_3$, is secondary: only two carbon substituents flank the empty $p$ orbital, giving roughly six hyperconjugating C–H bonds (three from each flanking methyl/methylene), fewer donor bonds than A. Step 3. Cation C, $CH_3CH_2CH_2CH_2^+$, is primary: only one carbon substituent (the propyl chain) flanks the cationic carbon, and the empty $p$ orbital is otherwise flanked by two hydrogens (which cannot hyperconjugate as effectively as C–H bonds on a neighboring carbon in the same donor-count sense) — giving the fewest strong donor interactions. Step 4. Stability order follows directly from donor count: $A \text{ (3°)} > B \text{ (2°)} > C \text{ (1°)}$. This is not a memorized "tertiary beats secondary" rule — it is the direct consequence of counting aligned $\sigma_{C-H}/\sigma_{C-C}$ donors feeding the empty $p$ orbital (Mechanism 1, Step 4). Quantitatively, gas-phase hydride affinities track this exactly: tert-butyl cation is roughly 15–20 kcal/mol more stable than isopropyl, which is itself roughly 15 kcal/mol more stable than ethyl.

Example 2 (applied) — Predicting carbanion stability and reaction outcome for an asymmetric 1,3-dicarbonyl deprotonation. Consider ethyl 3-oxopentanoate (ethyl propionylacetate), $CH_3CH_2$–$C(=O)$–$CH_2$–$C(=O)$–$OCH_2CH_3$, treated with one equivalent of NaOEt. Which proton is removed, and how stable is the resulting carbanion relative to a simple alkyl carbanion? Step 1. Identify the acidic site: the central $CH_2$ flanked by a ketone carbonyl and an ester carbonyl. Removing one of those protons with ethoxide (curved arrow from the C–H bond to the departing H, picked up by ethoxide's lone pair) leaves a lone pair on that central carbon. Step 2. That lone pair is not localized — it delocalizes into the adjacent ketone $\pi^*$ system (draw the resonance arrow: lone pair moves in to form a new C=C bond, C=O π bond breaks, electrons move onto the ketone oxygen) giving an enolate resonance structure with negative charge on oxygen. It simultaneously delocalizes into the ester carbonyl the same way, giving a second enolate-type resonance structure with negative charge on the ester's carbonyl oxygen. Step 3. Compare this to deprotonating an isolated alkyl $CH_2$ with no adjacent carbonyl (e.g., a proton on the ethyl group's terminal $CH_3$, $pK_a \approx 50$). The dicarbonyl-flanked proton has $pK_a \approx 11$, roughly $39$ $pK_a$ units — about $53\ \text{kcal/mol}$ in $\Delta G$ terms at 298 K via $\Delta G = 2.303\,RT\,\Delta pK_a$ — more acidic, entirely because the resulting carbanion is stabilized by two independent resonance delocalizations rather than zero. Step 4. This is exactly the resonance-stabilization relationship from the core ideas: $\Delta E_{\text{resonance}}$ is large and negative because two legitimate low-energy delocalized structures (into each carbonyl) both lie close in energy to the localized picture, and mixing multiple nearly-degenerate resonance structures compounds the stabilization — which is why 1,3-dicarbonyl compounds (not simple ketones alone, $pK_a\approx20$) are the classic substrates for enolate alkylation chemistry.

Example 3 (exam/competition level) — Predicting stereochemical outcome from carbene spin state. A synthetic chemist treats cis-2-butene with two different carbene sources: (i) $:CCl_2$ generated from $CHCl_3$/strong base (phase-transfer conditions), and (ii) $:CH_2$ generated by photolysis of diazomethane in the gas phase at low pressure (conditions known to give the free, unrelaxed triplet due to collisional deactivation being too slow to intervene). Predict the stereochemical outcome of each cyclopropanation and justify mechanistically. Step 1 — Establish the spin state of each carbene from substituent/generation logic. $:CCl_2$ carries two chlorine π-donor substituents, which (Mechanism 2) widen the $\sigma$–$p$ gap via lone-pair donation into the empty $p$ orbital, giving a ground-state singlet. Gas-phase $:CH_2$ generated with insufficient collisional relaxation is famously obtained as the higher-energy triplet (the ground state of methylene itself, as derived above), because there is no third body available quickly enough to remove the excess energy and let it relax, and because $:CH_2$'s intrinsic ground state is already triplet. Step 2 — Apply the singlet mechanism (concerted, one-step, both C–C bonds form simultaneously). A singlet carbene has one filled orbital ($\sigma$, lone pair) and one empty orbital ($p$). It reacts with an alkene $\pi$ bond in a single concerted step: the alkene's filled $\pi$ donates into the carbene's empty $p$ orbital while the carbene's filled $\sigma$ lone pair donates back into the alkene's $\pi^*$ — both new C–C σ bonds form essentially simultaneously, in a single transition state, with no intervening species that can rotate. Because cis-2-butene's two methyl groups never have a chance to rotate relative to each other during this one-step closure, the product cyclopropane retains the alkene's original relative stereochemistry: cis-2-butene + $:CCl_2 \rightarrow$ the cis-dichlorocyclopropane (methyls remain cis to each other) — stereospecific, retention. Step 3 — Apply the triplet mechanism (stepwise, radical-like, via a diradical intermediate). A triplet carbene has two singly-occupied orbitals with parallel spins; it cannot form both new σ bonds at once because spin conservation forbids forming two new paired bonds from two electrons that both started out unpaired-with-carbene and unpaired-with-alkene-π in a single concerted step — the reaction must instead proceed by forming one new C–C bond first (one carbene electron pairs with one alkene π electron), generating an open-shell 1,3-diradical intermediate with a genuine, freely rotating C–C single bond at the former alkene carbons. Step 4 — Track loss of stereochemistry through bond rotation, then ring closure. This diradical intermediate lives long enough (relative to the sub-picosecond rotational barrier around a single bond) for rotation to scramble the relative orientation of the two methyl groups before the second C–C bond closes (which additionally requires an intersystem crossing, spin-flip, before the second bond can pair up, adding a further time delay that only worsens the stereochemical scrambling). The result: cis-2-butene + triplet $:CH_2$ gives a mixture of cis- and trans-1,2-dimethylcyclopropane — non-stereospecific, loss of alkene geometry — the classic diagnostic test (Skell's rule) used experimentally to assign carbene spin states.

The clean electron-counting language used for reactive intermediates converges on the work of several chemists, but the physical-organic architecture of "structure of the transition state/intermediate determines reactivity" is most associated with George Willard Wheland and, for carbocation structure specifically, Saul Winstein and George Olah. Olah's achievement was to move carbocations from a purely mechanistic inference — a species drawn mid-arrow in someone else's mechanism — into a directly observable chemical entity.