Bond energies let you estimate the enthalpy change of a reaction without ever opening a thermodynamics table — useful when the molecules you’re working with don’t have tabulated ΔHf values, which is most of them once you leave gen chem behind. The catch is that you’re trading exactness for coverage: bond energies are averages across many parent molecules, so a calculated ΔH typically lands within 5–15 % of the experimental value rather than dead on. For estimation and intuition that’s fine. For high-precision work, you need formation enthalpies.

The governing relationship reads cleanly:

ΔH_rxn = Σ(bonds broken) − Σ(bonds formed)

Breaking bonds takes energy in (positive); forming bonds releases energy out (negative). If forming the new bonds releases more than was spent breaking the old ones, the reaction is exothermic. Get the subtraction order wrong — formed minus broken — and your ΔH sign flips. That’s the single most common error in this entire calculation.

A reference table to anchor the numbers

Bond	Energy (kJ/mol)	Bond	Energy (kJ/mol)
H–H	436	C–C	347
O–H	463	C=C	614
C–H	413	C≡C	839
N–H	391	C–O	358
H–Cl	431	C=O	799
H–Br	366	C–N	305
Cl–Cl	242	C≡N	891
O=O	498	N=N	418
N≡N	945	O–O	146

A small thing worth noticing: the C=O entry at 799 kJ/mol is an average. In CO₂ each C=O bond is closer to 805 kJ/mol; in formaldehyde it’s about 743 kJ/mol. That spread of ±5–10 % is exactly what gives bond-energy ΔH estimates their error bars.

Worked: methane combustion

CH₄ + 2 O₂ → CO₂ + 2 H₂O

Bonds in the reactants: 4 C–H bonds in methane, 2 O=O bonds in oxygen.

Bonds in the products: 2 C=O bonds in CO₂, 4 O–H bonds across two waters.

Energy in (broken): 4(413) + 2(498) = 1652 + 996 = 2648 kJ.

Energy out (formed): 2(799) + 4(463) = 1598 + 1852 = 3450 kJ.

ΔH = 2648 − 3450 = −802 kJ/mol.

The accepted experimental value is −890 kJ/mol. The estimate is within 10 %, which is exactly what you’d expect from average bond energies — and the sign is right, which is the part that matters for predicting whether the reaction releases heat.

Hydrogenation of ethylene — the trap

C₂H₄ + H₂ → C₂H₆

Easy to mess this one up by counting all the C–H bonds in ethylene as broken. They aren’t — the C–H bonds in ethylene survive into ethane unchanged. Only count bonds that change.

Broken: the C=C double bond (614) and the H–H (436) → 1050 kJ.

Formed: a new C–C single bond where the double bond used to be (347) and two new C–H bonds where the hydrogens stuck (2 × 413 = 826) → 1173 kJ.

ΔH = 1050 − 1173 = −123 kJ/mol. Experimental is −137 kJ/mol — within 10 %.

HCl formation — surprisingly clean

H₂ + Cl₂ → 2 HCl. Broken: 436 + 242 = 678. Formed: 2(431) = 862. ΔH = 678 − 862 = −184 kJ/mol versus experimental −185. That’s about as good as bond-energy estimates ever get, and it’s good because the molecules involved are simple diatomics where the tabulated values are essentially exact.

Nitrogen fixation — why it’s so hard

N₂ + 3 H₂ → 2 NH₃. Broken: 945 + 3(436) = 2253. Formed: 6(391) = 2346. ΔH = 2253 − 2346 = −93 kJ/mol versus experimental −92 kJ/mol.

The interesting thing here isn’t the ΔH — it’s the 945 kJ/mol N≡N bond sitting in the “broken” column. That triple bond is one of the strongest in chemistry, and it’s the entire reason the Haber–Bosch process needs 400–500 °C and 200+ atm to run, and the entire reason biological nitrogen fixation needs an enzyme (nitrogenase) with a complicated metal cluster. The reaction is favorable thermodynamically but the activation energy is ferocious because you have to break that triple bond to do anything.

Why estimates are approximate

The bond energy you look up is an average across many parent molecules. The same C–H bond differs by 10 % depending on the carbon’s substituents:

C–H in CH₄: 439 kJ/mol
C–H in C₂H₆: 423 kJ/mol
C–H in CHCl₃: 401 kJ/mol

The tabulated 413 kJ/mol is roughly the midpoint. This is why bond-energy ΔH estimates are quick and useful but never precise. If the experimental value disagrees with your calculation by 30 %, suspect the underlying numbers, not your arithmetic.

Where this goes wrong

The single biggest error is reversing the subtraction — write ΔH = formed − broken instead of broken − formed and your sign flips, turning exothermic combustion into endothermic absurdity. Second: counting bonds that don’t change. The four C–H bonds in ethylene that persist into ethane are spectators. They don’t appear on either side of the calculation. Third: applying gas-phase bond energies to liquids or solids without correcting for vaporization or fusion enthalpies. The values in tables assume gas-phase reactants and products.

For practice, try these and verify with the Thermochemistry Calculator:

Estimate ΔH for ethanol combustion: C₂H₅OH + 3 O₂ → 2 CO₂ + 3 H₂O. (Ethanol has 1 C–C, 5 C–H, 1 C–O, and 1 O–H bond.)
ΔH for H₂ + Br₂ → 2 HBr (H–H = 436, Br–Br = 193, H–Br = 366).
Estimate ΔH for CH₄ + Cl₂ → CH₃Cl + HCl.
Which is more exothermic per mole — methane combustion or ethane combustion? Estimate both with bond energies.
Why is N≡N (945) so much stronger than N–N single (163)? Calculate ΔH for N₂H₄ → N₂ + 2 H₂.

How to Calculate Bond Energy