Quantitative interpretation of the excitonic splittings in aluminium nitride (Vol. 42, No. 3)

image locus of the values of the hole effective masses compatible with the 38meV experimental value of the 1s-2s splitting (purple line). Similar plots for alternative values of the dielectric constant (blue and green plots). The predictions of several first principle calculations are plotted using red dots.

Properties of free exciton are used to address the first self consistent, all-optical determination of hole effective masses in aluminium nitride.

AlN is a wurtzite semiconductor, which appears to be very promising as a substrate or as a buffer layer in many heteroepitaxial growths of devices like UV light emitting diodes, solar blind light detectors, high power and/or high frequency operating field-effect transistors. Its thermal conductivity is superior to that of GaN, which is today, word widely used as a substrate for pulling forward the solid state lighting and the Blue Ray laser technologies.

The growth of AlN was only very recently achieved as bulky bowls or as hetero-epitaxial films. The band gap of AlN changes with the growth conditions, which is interpreted in terms of residual strain fields existing in hetero-epitaxies. Here, the model leading to the quantitative interpretation of the evolution of the band gap of GaN under strain (B.Gil et al., PRB, 52, R17028, (1995)) is extended to AlN.

Improved crystalline quality and residual doping now allow high-resolution optical spectroscopy. The origin of the experimental value of the 1s-2s excitonic splitting is analyzed using a model of H atom adapted to anisotropic masses and dielectric constants. This analysis permits to extract from the experimental data, the couple of relevant values of the dielectric constant, to find the locus for values for the on-axis and in-plane hole effective masses (purple line in the figure) fully compatible with the measured value of the 1s-2s splitting.

Quantitative interpretation of the excitonic splittings in aluminum nitride
B. Gil, et al., Eur. Phys. J. Appl. Phys 53, 20303 (2011)