Comment on "Photoluminescence ring formation in coupled quantum wells: excitonic versus ambipolar diffusion".

In a recent Letter [1], an inner ring [2] in the photoluminescence (PL) pattern from GaAs=AlGaAs coupled quantum wells is attributed to the Mott transition in a system of initially photoexcited electron-hole (e-h) pairs and secondary excitons. In contrast, in Ref. [3] the inner ring has been explained and modeled in terms of in-plane transport and thermalization of indirect excitons. In our Comment we show that (i) Stern et al. considerably overestimate the density of photoexcited carriers, (ii) the density n of secondary indirect excitons absolutely dominates over the density of free e-h pairs, ne;h 1⁄4 ne 1⁄4 nh, and (iii) no Mott transition occurs in the system. Equation (2) for the factor f, used in Ref. [1] in order to improve the mean field approximation for exciton-exciton interaction Eint [see Eq. (1) in Ref. [1]] and to evaluate n by the energy shift of the exciton PL line, cannot be applied for densities relevant to [1–3]. The authors completely disregard the screening [4] of the exciton potential U 1⁄4 UddðrÞ they analyze. The correct expression for Eint is given by Eint 1⁄4 R n‘ðrÞUðrÞdr, where the local density n‘ 1⁄4 n‘ðr;n; TÞ of excitons in quasiequillibrium [5] is 1 e T0=T 1⁄4 ð1 e Tð0Þ 0 =TÞe ðUþu0n‘ u0nÞ=ðkBTÞ (1)