Comment on 'modelling the optical constants of GaAs:: excitonic effects at E1, E1+Δ1 critical points'

Djurisic and Li [1] have recently presented a calculation of the optical constants of GaAs in which they do not take into account excitonic effects at either E-1, E-1 + Delta (1) or E-0, E-0 + Delta (0) critical points (CPs). They employ band-to-band single-particle expressions with an energy-dependent pseudo-Gaussian broadening function. Their paper states that including excitonic effects at the former CPs has 'dubious physical interpretation'. It is also claimed that (a) 'excitonic effects are usually more pronounced at E-0, E-0 + Delta (0) than at the E-1, E-1 + Delta (1) CPs', (b) 'at room temperature the excitons are severely broadened and should not contribute significantly to the dielectric function' and (c) 'excitonic effects at E-1, E-1 + Delta (1) critical points do not represent a significant contribution to the dielectric function at room temperature for materials with low exciton binding energy'. These statements are completely incorrect, as is their fitting scheme.