Silicon Photonics
Lasing in Indirect Gap Semiconductors
In earlier widely cited work, W.P. Dumke considered optical transitions in indirect semiconductors for laser applications. His results, published in the late ‘60s, have so far been generally accepted as the theoretical proof that optical gain is not possible in silicon. Recent theoretical work by centre researchers however, now clarified that these conclusions are based on a flawed theory [T. Trupke et al., J. Appl. Phys. 93, 9058, (2003)].
It could be shown that, in indirect semiconductors, the absolute value of the negative absorption coefficient, which describes optical gain, can be orders of magnitude larger than the equilibrium absorption coefficient, contrary to the case for direct semiconductors (lower curve in figure).
Normalised absorption coefficient for indirect transitions as a function of the separation of the quasi-Fermi energy Δµ at photon energies ħω = Eg ± ħΩ , with ħΩ the phonon-energy. Upper curve: photon absorption with phonon emission (ħω=1.15eV). Lower curve: photon absorption with phonon absorption (ħω= 1.05eV).
Note that these results do not necessarily mean that Dumke’s conclusions are wrong, but that the question, whether fundamentally optical gain via band-band transitions is possible in bulk silicon or not should be checked using the corrected theory. This is currently work in progress.