Lewis Acids and Bases: The Electron-Pair Definition

The Lewis acid-base framework is not a historical curiosity — it is the operating vocabulary of enormous swaths of modern chemistry and technology. Industrial Friedel-Crafts alkylation and acylation, still used to manufacture pharmaceuticals, fragrances, and polymer precursors, run on AlCl₃ and related Lewis acids activating alkyl or acyl halides by coordinating to the halide's lone pair and generating a reactive electrophile. Ziegler-Natta and metallocene catalysts for polyethylene and polypropylene production depend on Lewis-acidic aluminum or zirconium centers activating alkene monomers. Zeolite catalysts in petroleum refining present both Brønsted-acidic bridging hydroxyls and Lewis-acidic framework aluminum sites, and controlling the balance between the two is a major lever in catalyst design. In biological and medicinal chemistry, Lewis acid-base interactions govern metalloenzyme active sites: zinc in carbonic anhydrase and carboxypeptidase acts as a Lewis acid, polarizing a bound water molecule to make it a better nucleophile, while iron in hemoglobin and the cytochromes coordinates oxygen and other ligands through exactly the dative-bond mechanism described here. Chelation therapy for heavy-metal poisoning is designed entirely around competitive Lewis base ligands (EDTA, DMSA) that out-compete biological Lewis bases for toxic metal cations. In materials science, Lewis acid-base chemistry underlies semiconductor doping and surface passivation, gas-phase adduct chemistry used in chemical vapor deposition (trimethylaluminum and ammonia adducts feature in GaN growth chemistry), and battery electrolyte design, where Lewis acidic/basic additives modulate lithium-ion solvation. Perhaps most strikingly, the twenty-first-century discovery of frustrated Lewis pairs — bulky phosphine/borane combinations sterically prevented from forming a classical adduct, yet still cooperatively activating small molecules like H₂ and CO₂ — showed that the geometric, steric side of the Lewis picture (not just the electronic side) can be engineered to do genuinely new chemistry, including metal-free catalytic hydrogenation. That story, Adduct Formation and Frustrated Lewis Pairs, depends entirely on the frontier-orbital and steric reasoning built here, and the classification of how strongly a given acid prefers a given base leads directly into The HSAB Principle: Hard and Soft Classification and the systematic trends surveyed in Lewis Acidity Trends in Main-Group Halides.

Ask the tutor: Why does a good Lewis base like pyridine sometimes fail to displace a weaker base already bound to a Lewis acid, even when thermodynamics favors the exchange?

Begin with the Lewis structures you already trust. A molecule's chemistry at a given site is controlled by what orbitals are available there: filled, nonbonding lone pairs are potential electron-pair donors, and empty, energetically accessible orbitals are potential electron-pair acceptors. Gilbert Lewis's insight, formalized in his 1923 book Valence and the Structure of Atoms and Molecules, was to define the acid/base relationship purely in these terms, stripped of any reference to hydrogen or to a particular solvent. A Lewis base is a species with a filled, non-bonding, or "loosely held" orbital — usually a lone pair — that is energetically and geometrically positioned to overlap with an empty orbital on a partner. A Lewis acid is a species with an empty (or very reactive, low-lying) valence orbital — a $p$, $d$, or hybrid orbital — capable of receiving that pair. When they meet, the base's HOMO and the acid's LUMO combine to give a new bonding molecular orbital, filled with the donated electron pair, and (usually) a new empty antibonding orbital higher in energy. This is exactly the same orbital mathematics used to build any two-electron covalent bond — the only asymmetry is that both electrons in the new bond came from one partner, which is precisely what "coordinate" or "dative" means. The classic textbook case is boron trifluoride and ammonia. BF₃ is trigonal planar ($sp^2$ boron), because boron has only three valence electrons pairs and an empty $2p_z$ orbital perpendicular to the molecular plane. That empty $2p_z$ is boron's LUMO — boron is electron deficient, six electrons around it rather than eight. Ammonia's nitrogen lone pair, sitting in an $sp^3$-like hybrid orbital, is its HOMO. When the two approach, the nitrogen lone pair donates into boron's empty $p_z$, boron rehybridizes from $sp^2$ to $sp^3$, the BF₃ fragment puckers out of planarity (F–B–F angles contract from 120° toward ~109.5°), and a new B–N dative bond forms. The product, H₃N·BF�3 (or F₃B·NH₃), has a tetrahedral boron and a pyramidal, nearly unchanged nitrogen.

Compare this to what happens with a proton. When NH₃ meets H⁺, exactly the same event occurs: the nitrogen lone pair (HOMO) donates into the empty 1s orbital of the bare proton (LUMO). The only special feature of this case is that the acceptor orbital belongs to a proton — no nucleus-shielding electrons, so the interaction is unusually strong and the product, NH₄⁺, is the ordinary Brønsted acid-base adduct we call "protonated ammonia." Brønsted acidity is thus a proton-specific special case nested inside the broader Lewis framework: every Brønsted acid is a Lewis acid (because the transferable H⁺ has an empty acceptor orbital), and every Brønsted base is a Lewis base (it must have a lone pair to bind that proton). The converse fails: BF₃, Al³⁺, Ag⁺, and SO₃ are all potent Lewis acids that have no proton to offer at all.

The picture generalizes further. Any species with an accessible empty valence orbital of the right symmetry and energy is a Lewis acid: main-group halides with empty $p$ orbitals (BF₃, AlCl₃, SiF₄ after ligand expansion via $d$-orbital or $\sigma^*$ participation), transition-metal cations with empty $d$ orbitals (Ag⁺, Cu²⁺, Fe³⁺), and even neutral molecules that can polarize into an electrophilic center, such as CO₂ (electrophilic carbon flanked by two oxygens) or SO₃. On the base side, any lone pair — on N, O, S, P, halides, or π-systems (alkenes acting as bases toward Pt(II), for instance) — is a candidate donor.

The central relationship is not a numeric formula so much as an energetic bookkeeping statement for the adduct-formation reaction

where:

• A is the Lewis acid, contributing an empty acceptor orbital (LUMO),
• :B is the Lewis base, contributing a filled donor orbital (HOMO) with an electron pair,
• A←B is the adduct, containing one new two-electron coordinate covalent bond.

The strength of that new bond — and hence the equilibrium constant for adduct formation — is governed qualitatively by second-order perturbation theory for two interacting orbitals:

where:

• $H_{\text{AB}}$ is the resonance (overlap-dependent) interaction integral between the acid's LUMO and the base's HOMO,
• $E_{\text{LUMO(A)}}$ and $E_{\text{HOMO(B)}}$ are the orbital energies of acceptor and donor, respectively,
• the denominator is the energy gap between donor and acceptor orbitals — smaller gaps give larger stabilization for a given $H_{\text{AB}}$.

Reading: this is the same frontier-molecular-orbital expression used throughout organic and organometallic reactivity (it underlies the Klopman-Salman treatment that later motivates hard/soft classification). Two levers control adduct strength: how large the interaction matrix element $H_{\text{AB}}$ is (which scales with orbital overlap, itself controlled by orbital size and directionality — sterics matter here), and how small the energy gap between donor HOMO and acceptor LUMO is. A good Lewis acid has a low-lying LUMO; a good Lewis base has a high-lying HOMO. When the gap is small and overlap is good, $\Delta E$ is large and negative (stabilizing), and the adduct is strong. A second essential relationship is the thermodynamic cycle connecting Lewis basicity to measurable adduct formation constants:

where:

• $K_{\text{f}}$ is the adduct formation (stability) constant,
• $\Delta H^\circ$ is dominated by the new bond's strength (orbital-overlap driven) offset by any distortion (reorganization) energy needed to pyramidalize the acid or compress bond angles,
• $\Delta S^\circ$ is usually unfavorable (negative) because two particles combine into one, so entropy opposes adduct formation and must be overcome by a sufficiently exothermic $\Delta H^\circ$.

Reading: even a very good orbital match can fail to produce a stable, isolable adduct if the entropic penalty of combining two free molecules into one outweighs the enthalpic gain — this is one reason gas-phase or low-temperature adduct formation constants differ so much from solution values, and why solvent competition (the solvent itself may be a Lewis base) matters enormously in practice.

Example 1 — Foundational: identifying acid and base sites and drawing the adduct for BCl₃ + (CH₃)₃N. Trimethylamine, N(CH₃)₃, has nitrogen bonded to three methyl groups plus one lone pair — an $sp^3$-hybridized, pyramidal nitrogen whose lone pair is the HOMO and is even more accessible (more basic) than in NH₃, because the electron-donating methyl groups raise the lone-pair energy (inductive effect) relative to N–H. BCl₃ is trigonal-planar boron with an empty $2p_z$ orbital (LUMO), by the same VSEPR/hybridization argument as BF₃ — three bonding domains, no lone pair, boron uses $sp^2$ hybrids, leaving $2p_z$ empty. Bringing them together: the nitrogen lone pair donates into boron's empty $2p_z$, forming a new N–B σ-bond. Boron rehybridizes $sp^2 \to sp^3$ and its three Cl substituents pucker from 120° toward ~109°; nitrogen, already pyramidal, changes little. The product is written

with formal charges $+1$ on N and $-1$ on B in the fully ionic-arrow convention (though the actual bond is strongly covalent and these formal charges are bookkeeping, not real charge separation). Because BCl₃ is a stronger Lewis acid than BF�3 toward amines (weaker Cl→B π-donation to lose, as established in the derivation above) and because trimethylamine's lone pair is more basic than ammonia's, this adduct's formation constant is larger than that of F₃B·NH₃ — consistent with measured gas-phase and calorimetric Lewis basicity/acidity scales.

Example 2 — Applied: ranking Lewis acidity of the boron trihalides using the reorganization-energy argument, and checking against real thermodynamic data. The naive expectation from electronegativity is that BF₃, with the most electronegative, most electron-withdrawing halogen, should be the strongest Lewis acid (most electron-poor boron, lowest LUMO). Experimentally, however, gas-phase and solution enthalpies of adduct formation with a common reference base (pyridine, or NH₃) show the opposite trend for many bases:

in Lewis acidity (with occasional inversions depending on the exact base and measurement conditions, but this ordering is the standard qualitative result). Apply the reorganization-energy framework:

Step 1. In each BX₃, boron's empty $2p_z$ accepts π-donation from filled halogen $p$ orbitals in the free acid. This donation is strongest when the halogen $p$ orbital is closest in energy and size to boron's $2p$ orbital.

Step 2. Fluorine's $2p$ orbital is very close in size to boron's $2p$ orbital (both period 2), giving excellent F→B π-overlap in free BF₃ — so free BF₃ is unusually stabilized by π-donation, meaning its "true" electrophilicity at boron is partly quenched already.

Step 3. Chlorine and bromine use $3p$ and $4p$ orbitals respectively, which are much larger and more diffuse than boron's compact $2p$ orbital: the size mismatch gives poor π-overlap, so BCl₃ and BBr₃ retain more of boron's intrinsic electron deficiency and thus have a hungrier, less-quenched LUMO.

Step 4. When each BX₃ pyramidalizes to accept the base's lone pair, that same π-donation channel is destroyed (Step 3 of the earlier derivation). BF₃ therefore pays the largest reorganization penalty (it loses the most stabilizing π-donation) while gaining the same-order new σ-bond; BBr₃ pays the smallest penalty because it had little π-stabilization to lose. Net result: BBr₃ is the strongest Lewis acid of the series, BF₃ the weakest, even though a naive inductive-effect argument predicts the reverse. Checking a limiting case: if there were no halogen p→B π-donation available at all (imagine hypothetically replacing halogens with purely σ-donating, non-π-active substituents), the ordering would collapse back to the naive electronegativity prediction, since the reorganization penalty would vanish for all three and only inductive electron withdrawal (favoring F) would remain. The fact that real BX₃ acidity is not ordered this way is itself evidence for the importance of the π-donation/reorganization mechanism.

Example 3 — Exam/competition level: quantitative estimate of adduct-formation spontaneity using a simplified thermodynamic cycle. Suppose calorimetric measurements give the following data (representative literature-style values, not universal constants) for the reaction

with $\Delta H^\circ = -184\ \text{kJ mol}^{-1}$ and $\Delta S^\circ = -285\ \text{J mol}^{-1}\text{K}^{-1}$ (both plausible orders of magnitude for a bimolecular gas-to-solid adduct-forming reaction with a large loss of translational/rotational entropy). Find $\Delta G^\circ$ at $T = 298\ \text{K}$ and at $T = 450\ \text{K}$, and determine the temperature at which the adduct becomes thermodynamically unfavorable to form (the "ceiling temperature" for this adduct).

Step 1. At 298 K:

Strongly favorable — the adduct is stable and forms spontaneously at room temperature, consistent with F₃B·NH₃ being an isolable, bottleable solid.

Step 2. At 450 K:

Still favorable, but noticeably less so — the entropic penalty is eating into the enthalpic stabilization as temperature rises, exactly as expected since adduct formation reduces particle count and translational/rotational freedom.

Step 3 — ceiling temperature. Set $\Delta G^\circ = 0$ to find where the reaction stops being spontaneous:

Above roughly 646 K, this particular adduct would be thermodynamically disfavored to form (or would dissociate if already formed) — a direct, checkable numeric prediction from the enthalpy/entropy data, and a useful sanity limit: as $T \to \infty$, the $-T\Delta S^\circ$ term dominates and any bimolecular adduct-forming reaction with negative $\Delta S^\circ$ eventually becomes non-spontaneous, no matter how strong the new dative bond is. This is the general reason why many Lewis adducts that are stable solids at room temperature dissociate back into free acid and base in the gas phase at high temperature.

Gilbert Newton Lewis (1875–1946) is one of the most influential yet Nobel-less figures in twentieth-century chemistry. Trained at Harvard, he spent the bulk of his career building the chemistry department at the University of California, Berkeley, into a research powerhouse. He is best known for the shared-electron-pair covalent bond and the Lewis structure notation every chemistry student learns in the first weeks of general chemistry, but his 1923 book Valence and the Structure of Atoms and Molecules went much further: it proposed that acid-base behavior itself should be redefined around electron pairs rather than protons, deliberately generalizing the Brønsted-Lowry proton-transfer definition (published independently the same year) into a framework that could describe non-aqueous, non-protic, and metal-ion chemistry on equal footing. Lewis was notoriously never awarded the Nobel Prize despite 41 nominations — one of the most cited omissions in the Prize's history — a fact that itself became a case study in the sociology and occasional arbitrariness of scientific recognition. His electron-pair theories of both bonding and acidity, however, became permanent, load-bearing infrastructure for the rest of chemistry.