In junior-high and high school you learn that electrons around a nucleus can take only specific energies. The 1925 Franck–Hertz experiment provided one of the earliest decisive pieces of direct evidence for that quantization. The scientists accelerated electrons in a mercury-vapor vacuum tube and measured the current. Every 4.9 volts the current dropped, showing that electrons lost exactly 4.9 eV when they excited mercury atoms in an inelastic collision. This verified Bohr’s idea that atoms possess discrete, not continuous, energy levels. The concept underpins fluorescent lamps, semiconductor lasers and many optical-electronic devices and strongly influences spectroscopic element identification and space-plasma research.

The standard Franck–Hertz setup uses a triode vacuum tube with a variable accelerating voltage between cathode and grid and a small retarding potential before the collector. Keeping mercury vapor at ≈1 Pa gives a mean free path of a few millimeters, so each electron has about one collision inside the 2 cm drift region. As the accelerating voltage is swept, the collector current shows minima every 4.9 V, matching the 6 ^3P₁ excitation energy of mercury. Additional minima at 9.8 and 14.7 V confirm that electrons can undergo multiple additive excitations. Observing the 254 nm UV line with a spectrometer further proves the quantitative relation E = hν. The method evolved into electron-impact spectroscopy, allowing precise measurement of excitation cross sections and state lifetimes. The results underpin the design of low-pressure mercury lamps, plasma displays, laser sources and inform techniques such as electron-impact ionization and laser cooling.

The original Franck–Hertz tube employed a triode geometry with a variable cathode-to-grid accelerating voltage V_acc and ~1 V grid-to-anode retarding potential V_ret. At a mercury vapor pressure of 0.8–1 Pa the mean free path λ_mfp ≈ 1 cm, satisfying the single-collision condition within the 2 cm drift region. The collector current I_c shows pronounced minima spaced by 4.9 V, coinciding with the 6¹S₀→6³P₁ excitation energy of 4.89 eV. The inelastic scattering cross-section σ(E) rises steeply above threshold and peaks near E≈6 eV, consistent with modern R-matrix calculations. Successive minima demonstrate that electrons undergo integer-multiple excitations before reaching the grid, verifying additivity of energy losses. Detection of 254 nm fluorescence synchronized with current minima establishes a direct correspondence between electrical and radiative channels. Analysis of the second derivative d²I_c/dV² yields threshold energies with <0.02 eV accuracy. Modified experiments with He and Ne probe metastable-state production and test LS-coupling selection rules. The resulting cross-section data serve as benchmarks for plasma modeling and EUV-lithography light-source design. Historically, the observations provided decisive empirical support for the old quantum theory and directly motivated the development of matrix mechanics in 1925–26.