A heavy nucleus that absorbs a neutron forms a compound nucleus that can split ("fission") into two lighter fragments plus a few free neutrons. Energy is released because mid-mass nuclei are more tightly bound per nucleon than very heavy ones — the reaction converts a small amount of mass into energy via E = Δmc².
Binding energy is estimated with the semi-empirical (liquid-drop) mass formula:
B(A,Z) = aV·A − aS·A^(2/3) − aC·Z(Z−1)/A^(1/3) − aA·(A−2Z)²/A + δ
aV=15.75, aS=17.8, aC=0.711, aA=23.7 MeV
δ = +12/√A (even-even), −12/√A (odd-odd), 0 (odd A)
For a compound nucleus (A,Z) splitting into fragments (A₁,Z₁) and (A₂,Z₂) plus ν free neutrons (A₁+A₂+ν = A, Z₁+Z₂ = Z):
Q = B(A₁,Z₁) + B(A₂,Z₂) − B(A,Z)
For U-235 this simplified liquid-drop estimate gives roughly 140–180 MeV depending on the split you choose — in the right ballpark of, but somewhat below, the measured ≈202 MeV per fission (the SEMF is a smooth average model and misses shell corrections that add a few extra MeV in reality). The released Q splits roughly into: fragment kinetic energy (~84%, the dominant share — this is what heats a reactor coolant), prompt neutron KE (~2.5%), prompt gamma rays (~3.5%), and delayed beta/gamma/antineutrino energy from fragment decay (~10%, mostly not recoverable as heat).
Note on asymmetry: real U-235 fission strongly prefers an asymmetric split (~95/140) because of nuclear shell effects the simple liquid-drop model above does not include — the slider lets you explore the full range, but the shape of the real yield curve is a shell-structure effect layered on top of this liquid-drop energetics.
Power generation: a reactor cannot use all of Q as usable heat — antineutrino energy escapes entirely, and 1 gram of fully-fissioned U-235 (idealised, no real reactor reaches 100%) releases roughly a day's output of a small power plant, which is why nuclear fuel has such an extreme energy density compared to chemical fuels like coal.