Ka. Mader et al., ELECTRONIC CONSEQUENCES OF RANDOM LAYER-THICKNESS FLUCTUATIONS IN ALAS GAAS SUPERLATTICES/, Journal of applied physics, 78(11), 1995, pp. 6639-6657
We study the effects of a few types of atomic disorder on the electron
ic and optical properties of AlAs/GaAs (001) and (111) superlattices:
(i) atomic intermixing across the interfaces; (ii) replacing a single
monolayer in a superlattice by one containing the opposite atomic type
(isoelectronic delta doping); and (iii) random layer-thickness fluctu
ations in superlattices (SL). Type (i) is an example of lateral disord
er, while types (ii) and (iii) are examples of vertical disorder. Usin
g three-dimensional empirical pseudopotential theory and a plane-wave
basis, we calculate the band gaps, electronic wave functions, and opti
cal matrix elements for systems containing up to 2000 atoms in the com
putational unit cell. Spin-orbit interactions are omitted. Computation
ally much less costly effective-mass calculations are used to evaluate
the density of states and eigenstates away from the band edges in ver
tically disordered SLs. Our main findings are: (i) Chemical intermixin
g across the interface can significantly shift the SL energy levels an
d even change the identity (e.g., symmetry) of the conduction-band min
imum in AlAs/GaAs SLs; (ii) any amount of thickness fluctuations in SL
s leads to band-edge wave-function localization; (iii) these fluctuati
on-induced bound states will emit photons at energies below the ''intr
insic'' absorption edge (red shift of photoluminescence); (iv) monolay
er fluctuations in thick superlattices create a gap level whose energy
is pinned at the value produced by a single delta layer with ''wrong'
' thickness; (v) (001) AlAs/GaAs SLs with monolayer thickness fluctuat
ions have a direct band gap, while the ideal (001) superlattices are i
ndirect for n<4; (vi) there is no mobility edge for vertical transport
in a disordered superlattice, because all the states are localized; h
owever, the density of states retains some of the features of the orde
red-superlattice counterpart. We find quantitative agreement with expe
riments on intentionally disordered SLs [A. Sasaki, J. Cryst. Growth 1
15, 490 (1991)], explaining the strong intensity and large red shift o
f the photoluminescence in the latter system. We provide predictions f
or the case of unintentional disorder. (C) 1995 American Institute of
Physics.