Heavy doping made tunneling accessible
Esaki’s germanium junction work in 1957, published in 1958, used very heavy doping. The narrow depletion region and degenerate carrier occupations allowed tunneling to compete strongly with ordinary diode conduction.
The measured current first rose and then fell as forward bias increased. This reproducible negative differential resistance was the observation that needed an explanation.
Sources: [1]
Transmission alone does not set the current
An electron needs an occupied starting state and an available final state at the appropriate energy. Bias changes their overlap. Beyond the peak, fewer eligible states can contribute even though the voltage is larger.
Sources: [1]
The valley has other contributions
At larger forward bias, ordinary diffusion current grows. Phonon-assisted and defect-related processes can fill the valley, so a measured peak-to-valley ratio depends on temperature and junction quality. The diode does not supply energy without a bias source.
Sources: [1]
The barrier Lab explains one ingredient
A barrier-transmission calculation shows why a thin barrier can pass an electron. It has no degenerate semiconductor reservoirs, so it cannot on its own reproduce the tunnel diode’s full current curve. Esaki and Giaever share one 1973 award contribution but probed different electronic systems.