Bond Dissociation Energies and Radical Stability

Break a single C–H bond in isobutane homolytically and you pay 96.5 kcal/mol. Break one in methane and the bill is 105.0 kcal/mol. That's an 8.5 kcal/mol discount just for sitting on a tertiary carbon instead of a primary one. That gap is not an accident of tabulated numbers. It is the thermochemical fingerprint of how well a half-filled p orbital on carbon can be propped up by the bonds around it. Understanding why radicals get cheaper to make as substitution increases matters greatly. It is the single most useful piece of physical organic chemistry for predicting where a chain reaction will strike. Bond dissociation energy (BDE) is the enthalpy of homolysis of a bond in the gas phase. It generates two radicals with unpaired electrons on the fragments. Unlike bond energies measured by adding up average values from a table of "typical" bonds, a BDE is different. It is a real, measurable thermochemical quantity for one specific bond in one specific molecule. It is exquisitely sensitive to the local environment. This includes the identity of the carbon losing a hydrogen. It also includes the substituents around it, and any conjugation available to the resulting radical. This sensitivity is precisely what makes BDEs useful. They are a direct experimental probe of radical stability. The big picture connects three things that look separate at first. These are Classifying Carbons and Hydrogens: 1°, 2°, 3°, 4° (a purely structural classification), radical stability (a purely electronic/orbital concept), and BDEs (a purely thermochemical, measurable quantity). The order 3° > 2° > 1° > methyl in radical stability is read off the BDE table. The mechanistic reason for that order is hyperconjugation. This is donation from adjacent C–H (or C–C) σ bonds into the half-filled 2p orbital on the radical carbon. This single orbital-overlap argument explains the substitution trend. It does so when run through a Born–Haber-style thermochemical cycle. It also explains why allylic and benzylic radicals are dramatically more stable still. And it explains why this whole apparatus is the load-bearing foundation. That foundation is for predicting selectivity in radical chain reactions. This article builds the BDE/radical-stability toolkit from the ground up. It covers how a BDE is actually defined and measured. It presents the curved-arrow/orbital picture of homolysis. It gives a full mechanistic derivation of why hyperconjugation stabilizes radicals. This includes the SOMO–σ* overlap picture. It offers a worked Hess's-law cycle showing how BDE tables predict reaction enthalpies. Finally, it gives the numerical anchors you should be able to reproduce on demand: methyl ~105, primary ~101, secondary ~98.6, tertiary ~96.5 kcal/mol.

Key takeaways

A bond dissociation energy (BDE) is the ΔH° for homolytic cleavage of one specific bond, gas phase, 298 K, generating two radicals.
Radical stability order: 3° > 2° > 1° > methyl, read directly from decreasing C–H BDEs (105.0, 101.1, 98.6, 96.5 kcal/mol for methane, ethane(1°), propane(2°), isobutane(3°)).
The stabilization comes from hyperconjugation: overlap of filled, properly aligned C–H (or C–C) σ bonds with the half-filled 2p orbital (SOMO) on the radical carbon.
More alkyl substituents on the radical carbon means more σ(C–H) donors available to overlap with the SOMO, hence a lower-energy, more stable radical, hence a lower BDE for the bond that made it.
BDE tables let you compute ΔH°rxn for any radical step by Hess's law: ΔH°rxn = Σ BDE(bonds broken) − Σ BDE(bonds formed).
Allylic and benzylic C–H BDEs (~88 and ~90 kcal/mol) are far lower than any alkyl C–H because the radical is resonance- (not just hyperconjugatively-) stabilized — a preview of extended radical delocalization.
Key idea: BDE is a direct thermochemical measurement of radical stability, and radical stability tracks the number of adjacent σ-bonds (C–H/C–C) available to hyperconjugate into the singly-occupied p orbital.

Where this lives today

BDE reasoning underlies why radical chain halogenation of alkanes shows the selectivity it does — the mechanistic backbone of industrial processes producing chlorinated solvents and intermediates, where tertiary and secondary positions react preferentially over primary ones for exactly the hyperconjugative reasons developed here. In polymer chemistry, the choice of radical initiator (azo compounds, peroxides, persulfates) for controlled radical polymerization is a direct BDE-table exercise: pick the weakest bond that decomposes at the desired process temperature without being so weak that it decomposes uncontrollably during storage. In combustion and atmospheric chemistry, BDE tables for C–H bonds in fuels and volatile organic compounds predict which positions are attacked first by hydroxyl radicals, directly informing models of tropospheric ozone formation and fuel autoignition chemistry. In pharmaceutical metabolism, cytochrome P450 enzymes perform C–H hydroxylation preferentially at weaker (more substituted, benzylic, or allylic) C–H bonds of drug molecules — BDE and radical-rebound mechanisms are used to predict metabolically vulnerable sites and design metabolically stable analogs. In materials science, autoxidation and degradation of polymers, lubricants, and foodstuffs (rancidity) proceed through radical chains whose rate-limiting H-abstraction step is governed by exactly the same BDE ladder, with allylic and bis-allylic C–H bonds (very low BDE, ~75–80 kcal/mol in polyunsaturated systems) being the dominant sites of attack — the mechanistic seed that later grows into full treatments of allylic/benzylic resonance stabilization and ring-strain analysis via Heats of Combustion as a Measure of Ring Strain.

Ask the tutor: Why does bromination show much higher 3°:1° selectivity than chlorination, if both follow the same BDE-based radical stability order?

The physical picture

Homolysis of a C–H bond means the two bonding electrons split evenly, one going to carbon and one to hydrogen, rather than both going to one atom (heterolysis, which makes ions). Curved-arrow notation for a homolytic step uses single-barbed (fish-hook) arrows, each showing the movement of one electron, in contrast to the double-barbed arrows of heterolytic ionic mechanisms. Two fish-hooks point away from the middle of the bond being broken, one toward each fragment.

R{-}H \;\xrightarrow{\Delta H = \text{BDE}}\; R^{\bullet} + H^{\bullet}

The carbon radical R• is sp²-ish (nearly planar at the radical center, though the barrier to pyramidal inversion is tiny) with the unpaired electron sitting in a 2p orbital perpendicular to the plane of the three remaining substituents — this p orbital is the SOMO (singly occupied molecular orbital). Everything about substituent-dependent radical stability comes down to what can donate electron density sideways into that empty "half," i.e., overlap with the SOMO. A neighboring C–H σ bond, if its orbital axis is roughly parallel to the SOMO's axis, can donate a small amount of electron density from the filled σ(C–H) bonding orbital into the SOMO. This is a three-electron interaction (two from σ, one from the SOMO) that is net stabilizing because two electrons drop in energy (into the resulting lower combination) while only one rises (into the higher combination) — the same bookkeeping that makes ordinary two-orbital, two-electron bonding favorable, just with an odd electron thrown in. This is hyperconjugation, and each adjacent C–H or C–C σ bond that is well-aligned is one more stabilizing donor.

Figure

Homolysis of a C–H bond: fish-hook arrows and the resulting radical

Now compare the number of hyperconjugating C–H bonds available at each radical center. A primary radical (•CH₂–CH₃) has the SOMO on a carbon flanked by only the one adjacent CH₃ group, giving three C–H σ bonds on that neighboring carbon that can (partially, depending on rotamer averaging) align with the SOMO. A tertiary radical, e.g., •C(CH₃)₃, has three separate methyl groups flanking the radical carbon, each contributing its own set of C–H σ donors — roughly three times the hyperconjugative donor density of the primary case. More donors, more stabilization, lower energy radical, and — because the radical is a product of homolysis — a lower BDE for the bond whose cleavage produced it.

Figure

Hyperconjugation: SOMO on the radical carbon overlapping an adjacent C–H σ bond

The BDE order therefore reads directly off the substitution pattern of the carbon losing the H atom:

Figure

BDE ladder: methyl, 1°, 2°, 3° C–H bonds

The core ideas

The formal definition of the BDE for a bond A–B is the standard enthalpy of the homolysis reaction:

\text{BDE(A–B)} \equiv \Delta H^{\circ}_{298}\big[\,A{-}B(g) \rightarrow A^{\bullet}(g) + B^{\bullet}(g)\,\big]

where:

$A{-}B(g)$ is the neutral molecule in the gas phase (BDEs are gas-phase quantities; solvent effects on radicals are usually small but not zero)
$A^{\bullet}, B^{\bullet}$ are the two radical fragments, each retaining one electron from the broken bond
$\Delta H^{\circ}_{298}$ is measured or back-calculated from heats of formation: $\Delta H^{\circ} = \Delta H_f^{\circ}(A^\bullet) + \Delta H_f^{\circ}(B^\bullet) - \Delta H_f^{\circ}(A{-}B)$

Reading: the BDE is always positive for a real bond (breaking bonds costs energy), and a smaller BDE at the same B (commonly H•) means the radical A• produced is more stable relative to its parent molecule — stability and BDE move in opposite directions for a fixed leaving fragment. The Hess's-law consequence, used constantly in predicting whether a radical propagation step is exothermic:

\Delta H^{\circ}_{\text{rxn}} = \sum \text{BDE(bonds broken)} - \sum \text{BDE(bonds formed)}

where:

bonds broken are on the reactant side (endothermic contributions, positive)
bonds formed are on the product side (their BDEs are subtracted because forming a bond releases energy equal to its BDE)
this is an enthalpy estimate; it ignores entropy and assumes gas-phase, additive bond energetics — good for exothermicity trends, not for absolute rates

Derivation AWhy hyperconjugation lowers the SOMO energy — a two-orbital, three-electron argument

Goal. Show mechanistically, from orbital overlap, why a filled adjacent σ(C–H) bond lowers the energy of a radical relative to a case with no such neighbor, and why more such neighbors (going 1°→2°→3°) lowers it further.

Step 1 — Identify the two interacting orbitals. At a radical carbon, the SOMO is a 2p orbital, singly occupied, energy $\varepsilon_{SOMO}$ . On the adjacent (β, relative to the numbering convention where the radical carbon is α) carbon, a C–H bond has a filled bonding σ orbital, energy $\varepsilon_{\sigma}$ , lower in energy than the SOMO because σ(C–H) is a strong, well-overlapped two-center bond while the SOMO is a nonbonding-ish p orbital.

Step 2 — Symmetry/overlap requirement. Only a σ(C–H) bond whose C–H axis is roughly parallel to the p-orbital axis of the SOMO has non-zero overlap integral $S = \langle \sigma | p \rangle \neq 0$ . Free rotation about the Cα–Cβ bond means that, on a time-average (or in the lowest-energy conformer), some rotamer always places a C–H bond close to this alignment — this is why hyperconjugation is treated as an always-available stabilization for sp³ neighbors, not a rare special geometry.

Step 3 — Build the perturbation-theory mixing. Treating σ(C–H) and the SOMO p-orbital as a two-orbital interacting system (standard second-order perturbation / simple Hückel-type mixing), the two new combinations are a lower, bonding-like combination $\psi_{-}$ and a higher, antibonding-like combination $\psi_{+}$ :

\varepsilon_{\pm} = \frac{\varepsilon_{\sigma} + \varepsilon_{SOMO}}{2} \pm \sqrt{\left(\frac{\varepsilon_{\sigma}-\varepsilon_{SOMO}}{2}\right)^2 + \beta^2}

where $\beta$ is the (negative-convention) resonance/interaction integral, proportional to the overlap $S$ .

Step 4 — Electron count: this is the key three-electron bookkeeping. The σ(C–H) bond supplies 2 electrons; the SOMO supplies 1 electron. Total 3 electrons to place in $\psi_-$ and $\psi_+$ . Two electrons go into the lower orbital $\psi_-$ (energy drop relative to $\varepsilon_\sigma$ of roughly $|\beta|^2/(\varepsilon_{SOMO}-\varepsilon_\sigma)$ each, by second-order perturbation theory since $\varepsilon_\sigma < \varepsilon_{SOMO}$ ), and the third electron goes into the higher orbital $\psi_+$ (energy rise). Because two electrons drop while only one rises, and the drops and rise are of comparable magnitude per electron, the net effect is stabilizing — the total energy of the three electrons is lower than $2\varepsilon_\sigma + \varepsilon_{SOMO}$ (the unmixed reference).

Step 5 — Compare 1°, 2°, 3° radical centers. A primary radical carbon has one adjacent carbon supplying up to 3 aligned/averaged C–H σ donors. A secondary radical carbon has two adjacent carbons, each supplying its own set of C–H σ donors — roughly double the donor count of the primary case. A tertiary radical carbon has three adjacent carbons, each donating — roughly triple. Since each additional independent donor orbital contributes its own (weakly interacting, additive to good approximation) stabilization increment via Step 4, the total hyperconjugative stabilization increases monotonically 1° < 2° < 3°.

Step 6 — Connect stabilization to BDE. The parent alkane R–H has all bonding electrons in ordinary σ bonds (no partial promotion, no SOMO). Homolysis promotes one electron pair into the R• radical + H•, and it is precisely the radical's energy that is lowered by hyperconjugation (the starting alkane has no such stabilization to lose, since C–H σ bonds are already fully bonding, not "half-donating"). So:

\text{BDE} = \big[H_f^\circ(R^\bullet) + H_f^\circ(H^\bullet)\big] - H_f^\circ(RH)

A more hyperconjugatively stabilized R• has a lower $H_f^\circ(R^\bullet)$ , and therefore a lower BDE, for the same starting alkane framework and same H• byproduct. This is exactly the observed order:

\text{BDE(CH}_3\text{–H}) > \text{BDE(1°)} > \text{BDE(2°)} > \text{BDE(3°)}

105.0 > 101.1 > 98.6 > 96.5 \ \text{kcal/mol}

Go deeper · Derivation B · 2 minHess's-law cycle: predicting the enthalpy of an initiation step from a BDE table

Derivation BHess's-law cycle: predicting the enthalpy of an initiation step from a BDE table

Goal. Use tabulated BDEs to compute $\Delta H^{\circ}$ for a concrete radical step — the abstraction of a tertiary hydrogen from 2-methylpropane (isobutane) by a chlorine atom — and show the arithmetic explicitly, with non-round numbers carried through.

Step 1 — Write the elementary step with fish-hook arrows. Chlorine atom (from initiation, Cl₂ → 2 Cl•, itself with BDE(Cl–Cl) = 58.0 kcal/mol) abstracts the tertiary H of isobutane:

\text{Cl}^{\bullet} + (CH_3)_3CH \longrightarrow \text{HCl} + (CH_3)_3C^{\bullet}

One fish-hook goes from the C–H bonding pair to the incipient Cl–H bond (one electron migrates to pair with Cl's unpaired electron); the other fish-hook stays on carbon, becoming the new unpaired electron on the tertiary carbon.

Step 2 — Identify bonds broken and bonds formed. Bond broken: the tertiary C–H bond of isobutane, BDE = 96.5 kcal/mol. Bond formed: the H–Cl bond, BDE = 103.0 kcal/mol.

Step 3 — Apply the Hess's-law relation.

\Delta H^{\circ}_{\text{step}} = \text{BDE(3° C–H)} - \text{BDE(H–Cl)}

\Delta H^{\circ}_{\text{step}} = 96.5 \ \text{kcal/mol} - 103.0\ \text{kcal/mol}

\Delta H^{\circ}_{\text{step}} = -6.5\ \text{kcal/mol}

Step 4 — Sanity-check against the alternative (primary) abstraction. If Cl• instead abstracted one of the nine equivalent primary hydrogens of isobutane (BDE(1° C–H) = 101.1 kcal/mol for a comparably branched primary position, close to the ethane-type value used here for illustration):

\Delta H^{\circ}_{\text{step, 1°}} = 101.1 - 103.0 = -1.9\ \text{kcal/mol}

Both steps are exothermic (Cl• is reactive enough to abstract almost any C–H bond), but tertiary abstraction is exothermic by 4.6 kcal/mol more, precisely the enthalpic reflection of the extra hyperconjugative stabilization built into Step 5 of the previous derivation. This differential exothermicity is part (not all — polar and steric effects also matter for Cl• vs Br•) of why radical halogenation shows partial selectivity for more substituted C–H bonds, a theme picked up fully when discussing chain-mechanism selectivity.

Step 5 — Full-cycle check via heats of formation (limiting-case consistency). As an internal consistency check, using $H_f^\circ[(CH_3)_3CH] = -32.1$ kcal/mol, $H_f^\circ(\text{Cl}^\bullet) = 29.0$ kcal/mol, $H_f^\circ(\text{HCl}) = -22.1$ kcal/mol, and solving for $H_f^\circ[(CH_3)_3C^\bullet]$ from BDE = 96.5:

H_f^\circ[(CH_3)_3C^\bullet] = \text{BDE} + H_f^\circ[(CH_3)_3CH] - H_f^\circ(H^\bullet)

H_f^\circ[(CH_3)_3C^\bullet] = 96.5 + (-32.1) - 52.1 = 12.3\ \text{kcal/mol}

Then recomputing the abstraction enthalpy independently from formation enthalpies:

\Delta H^\circ = \big[H_f^\circ(\text{HCl}) + H_f^\circ((CH_3)_3C^\bullet)\big] - \big[H_f^\circ(\text{Cl}^\bullet) + H_f^\circ((CH_3)_3CH)\big]

\Delta H^\circ = (-22.1 + 12.3) - (29.0 - 32.1) = -9.8 - (-3.1) = -6.7\ \text{kcal/mol}

This matches the Step 3 result (−6.5 kcal/mol) to within rounding of the tabulated formation enthalpies, confirming the two routes (direct BDE subtraction vs. full formation-enthalpy cycle) are the same calculation viewed two ways.

Worked examples

Example 1 — Foundational: rank the four labeled hydrogens of 2-methylbutane by expected BDE. 2-Methylbutane, $(CH_3)_2CHCH_2CH_3$ , has hydrogens of four distinct types: the tertiary C–H (on C2, attached to three carbons), the secondary C–H's (on C3, attached to two carbons), and two inequivalent sets of primary C–H's (the two methyls on C2, and the terminal methyl on C4 — chemically slightly different but both "primary," so grouped together for this ranking). Applying the hyperconjugation argument: the tertiary carbon radical formed by removing the C2–H has three flanking alkyl groups (two methyls + the ethyl chain) supplying σ(C–H) donors; the secondary radical (from C3–H) has two flanking groups; the primary radicals (from either methyl) have only one flanking group. Predicted BDE order:

\text{BDE(3°, C2–H)} < \text{BDE(2°, C3–H)} < \text{BDE(1°, C1/C4–H)}

Using anchor values (96.5, 98.6, 101.1 kcal/mol respectively, as transferable approximations from the simpler alkanes), the tertiary hydrogen is the "softest" target — it is removed with roughly 4.6 kcal/mol less enthalpic cost than a primary hydrogen on the same molecule. This is exactly the reactivity handle exploited in radical chain halogenation to rationalize (though not fully quantify, since statistics and reagent selectivity also matter) which C–H bond reacts preferentially.

Example 2 — Applied: comparing initiator efficiency using BDE(peroxide O–O). Radical initiators such as di-tert-butyl peroxide, $(CH_3)_3CO{-}OC(CH_3)_3$ , work because the O–O bond is unusually weak: BDE(O–O) ≈ 38 kcal/mol, versus a typical C–C BDE of ~85–90 kcal/mol or C–H BDE of ~98–105 kcal/mol. Homolysis,

(CH_3)_3CO{-}OC(CH_3)_3 \longrightarrow 2\,(CH_3)_3CO^{\bullet}

is thermally accessible around 100–150 °C precisely because 38 kcal/mol is a small activation enthalpy compared to breaking any C–H or C–C bond directly. This is a direct, practical payoff of the BDE concept: initiators are chosen by scanning a BDE table for the weakest bond compatible with the desired operating temperature, rather than by trial and error. The resulting tert-butoxy radical, $(CH_3)_3CO^\bullet$ , is itself reactive enough to abstract H atoms from substrate C–H bonds (favorably, since BDE(O–H) in tert-butanol, ~105 kcal/mol, exceeds most alkane C–H BDEs), continuing a chain.

Example 3 — Exam/competition level: quantify the extra stabilization energy of a tertiary vs. a methyl radical using an isodesmic-type BDE comparison, and connect it to a branching energy check. Consider the isodesmic (bond-type-conserving) hydrogen-transfer reaction:

(CH_3)_3C^{\bullet} + CH_4 \longrightarrow (CH_3)_3CH + CH_3^{\bullet}

Because both sides have the same number and rough type of bonds (one C–H broken, one C–H formed on each side, only the identity of the radical-bearing carbon changes), this reaction's enthalpy isolates the differential radical stability cleanly, canceling most systematic errors in absolute BDEs. Using Hess's law directly from BDEs:

\Delta H^{\circ}_{\text{rxn}} = \text{BDE(CH}_3\text{–H, methane, bond broken on RHS as reverse)} - \text{BDE(3° C–H, isobutane)}

More carefully, write it as two homolyses combined:

\Delta H^{\circ}_{\text{rxn}} = \big[\text{BDE(CH}_4) - \text{BDE}\big((CH_3)_3CH\big)\big]

\Delta H^{\circ}_{\text{rxn}} = 105.0 - 96.5 = 8.5\ \text{kcal/mol}

The reaction is favorable (exothermic by 8.5 kcal/mol) in the direction written, because it converts a less-stable methyl radical situation into a more-stable tertiary radical situation while dumping the "expensive" H back onto methane. This 8.5 kcal/mol is, by definition, the extra thermodynamic stabilization of a tertiary radical relative to a methyl radical, isolated as a clean number free of most bond-type bookkeeping ambiguities — exactly the number an exam question would want you to derive from a BDE table rather than quote from memory. As a limiting-case check: if the two radicals were equally stable (hypothetically, if hyperconjugation contributed nothing), the two BDEs would be equal and $\Delta H^\circ_{\text{rxn}} \to 0$ — which is not what is observed, correctly reflecting that the effect is real and finite.

Limiting cases, scope & common traps

Situation	What happens	Why
Comparing BDEs across different bond types (e.g., C–H vs. C–C)	Do not conclude radical stability directly; BDE also reflects the strength of the bond being broken, not just the radical formed	BDE = (energy to separate) reflects both fragments' properties, not radical stability alone
Highly delocalized radicals (allylic, benzylic)	BDE drops sharply below any alkyl value (~88, ~90 kcal/mol)	Resonance delocalization (multi-center) dominates over simple hyperconjugation (a two-center perturbative effect)
Vinyl or aryl C–H	BDE is higher than any alkyl C–H (~110+ kcal/mol)	The radical would sit in an sp² orbital in the molecular plane with poor hyperconjugative overlap, and more s-character bonds are intrinsically stronger
Bredt's-rule-type bridgehead radicals	Hyperconjugation trend can be disrupted by ring geometry preventing σ-alignment	Overlap requirement ( $S \neq 0$ ) fails if rotation/alignment is locked by ring constraints
Statistical vs. intrinsic selectivity	More primary H's exist per molecule than tertiary H's, so product ratios are not simply the inverse BDE ratio	Reaction rate depends on both intrinsic reactivity (BDE-correlated) and the number of equivalent hydrogens — a preview of the full Radical Halogenation of Alkanes: Chain Mechanism treatment
Using BDE tables for ΔH of ionic reactions	Invalid	BDEs are homolytic (radical) quantities; ionic bond-breaking enthalpies (heterolytic) are utterly different (much larger, dominated by charge separation)

The chemist behind it

The systematic use of bond dissociation energies to quantify radical stability grew out of gas-phase kinetics and thermochemistry in the mid-20th century, most closely associated with the American physical chemist Sidney W. Benson, whose group-additivity methods for estimating heats of formation (and hence BDEs) became the standard toolkit still taught today. Benson's key conceptual contribution was insisting that bond energies be treated not as fixed universal constants but as local, substituent-dependent quantities, extractable self-consistently from a large, cross-checked body of gas-phase kinetic and thermochemical data — precisely the "BDE is a measurement, not a lookup constant" philosophy that makes the modern radical-stability order rigorous rather than merely a mnemonic.

Origins · 1 minSidney W. Benson and how this was discovered

Sidney W. Benson

1918–2011 · Physical organic/thermochemical kinetics, USA, mid-to-late 20th century

Benson developed group-additivity ("Benson group") methods for predicting thermodynamic properties of organic molecules and radicals from tabulated fragment contributions, and his compiled BDE and heat-of-formation tables (refined through decades of gas-phase pyrolysis and kinetic studies) remain the standard reference data underlying the modern C–H BDE ladder (105.0 / 101.1 / 98.6 / 96.5 kcal/mol for methyl/1°/2°/3°) used throughout organic chemistry.

History

- 1930s — Early gas-phase pyrolysis kinetics begin providing activation energies interpretable as approximate bond strengths.

- 1950s — Systematic tabulation of radical heats of formation from kinetic and mass-spectrometric data (shock tubes, VLPP) accelerates.

- 1965–1976 — Sidney Benson formalizes group-additivity thermochemistry, enabling consistent, self-checking BDE tables across large families of hydrocarbons.

- 1980s–1990s — Gas-phase acidity/electron-affinity cycles and photoelectron spectroscopy refine radical heats of formation, correcting several older BDE values by a few kcal/mol.

- 2000s–present — High-level computational thermochemistry (composite methods such as G3, CBS-QB3) cross-validates experimental BDEs to within ~0.5–1 kcal/mol for most simple alkanes.

Is BDE the same thing as "bond energy" from an average bond-energy table?

No. Average bond energies (like a generic "C–H bond energy ≈ 99 kcal/mol" used for quick heat-of-reaction estimates) are averaged over many different molecules and environments, smoothing out exactly the substituent effects this article is about. A true BDE is specific to one bond in one named molecule, measured or calculated for that exact structure. The whole 3°>2°>1° trend is invisible if you only use a single averaged C–H value; it only appears once you look at position-specific, molecule-specific BDEs.

Why does hyperconjugation need "alignment," and doesn't rotation average it away?

The overlap integral between a σ(C–H) orbital and the SOMO's p orbital depends on the dihedral angle between them, maximal near 0°/180° (parallel) and zero near 90° (perpendicular). Free rotation about the C–Cα bond means the molecule samples all angles, but the energetically lowest conformer (and the Boltzmann-weighted average) always has substantial population near the aligned geometry, so on average there is always significant hyperconjugative stabilization available to an sp³ neighbor — it's diminished, not erased, by rotation.

Why is the allylic C–H BDE (~88 kcal/mol) so much lower than even a tertiary alkyl C–H (~96.5)?

Allylic stabilization is resonance, not just hyperconjugation: the unpaired electron delocalizes fully into an adjacent π system, spreading spin density over two (or more) carbons via genuine conjugation of a p orbital with a π bond, not just partial mixing with a σ bond. Resonance delocalization is intrinsically a stronger stabilizing effect than hyperconjugative donation, because the interacting orbitals are much closer in energy and much better overlapped (p–p π overlap vs. p–σ hyperconjugative overlap).

Does a lower BDE always mean a faster reaction rate for making that radical?

Usually correlated, but not guaranteed. BDE is a thermodynamic (enthalpy) quantity for the overall homolysis; the actual activation energy for an abstraction step also depends on polar effects (electrophilic radicals like Br• are choosier and prefer electron-rich C–H bonds even beyond the BDE trend), steric hindrance around the target hydrogen, and the specific transition-state geometry. The Bell–Evans–Polanyi/Hammond-type relationship makes lower BDE a very good predictor of lower activation energy within a closely related family of reactions, but it is an empirical correlation, not an identity.

Can BDE values be negative, or can they be zero?

No — a BDE is defined as the energy required to break a real, existing bond into two radical fragments, so it is always positive for any genuine covalent bond (otherwise the molecule would simply fall apart spontaneously at room temperature, which stable alkanes obviously do not). Extremely weak bonds (like O–O in some peroxides, ~35–39 kcal/mol) are still positive; they are just small enough to allow thermal homolysis at accessible temperatures.

How do I use a BDE table to predict which product dominates in a radical halogenation, roughly?

Multiply a rough "reactivity per hydrogen" factor (higher for more substituted, lower-BDE positions) by the number of equivalent hydrogens of each type, then compare the products; this combines the intrinsic BDE-driven preference with the statistical factor of how many such hydrogens exist. Chlorination is much less selective (small BDE-driven preference, Cl• is very reactive and early-transition-state) than bromination (highly selective, Br• is much choosier and BDE differences matter far more), a distinction developed fully in the chain-mechanism treatment of halogenation.

Why do textbook BDE tables sometimes disagree by 1–3 kcal/mol between editions?

Because BDEs are back-calculated from measured heats of formation of the radical and the parent molecule, and radical heats of formation have historically been harder to pin down experimentally (radicals are transient, reactive species) than those of stable closed-shell molecules. As spectroscopic and computational thermochemistry methods improved (photoelectron spectroscopy, high-level composite quantum-chemical methods), several radical heats of formation were revised by a few kcal/mol, shifting the derived BDEs slightly — the qualitative 3°>2°>1° ordering, however, has remained completely robust throughout.

Bond Dissociation Energies and Radical Stability

Where this lives today

The physical picture

The core ideas

Worked examples

Limiting cases, scope & common traps

The chemist behind it

Before this