Abstract
An exact solution of the tunnelling current through quantum-well barriers is compared with various approximations. Better agreement between theory and experiment follows in certain règimes.