Carrier relaxation in $\mathrm{GaAs}$ v-groove quantum wires and the effects of localization

Carrier relaxation processes have been investigated in $\mathrm{GaAs}∕{\mathrm{Al}}_x{\mathrm{Ga}}_{1-x}\mathrm{As}$ v-groove quantum wires (QWRs) with a large subband separation $\left(\Delta E\simeq 46\mathrm{meV}\right)$. Signatures of inhibited carrier relaxation mechanisms are seen in temperature-dependent photoluminescence (PL) and photoluminescence-excitation measurements; we observe strong emission from the first excited state of the QWR below $\sim 50\mathrm{K}$. This is attributed to reduced intersubband relaxation via phonon scattering between localized states. Theoretical calculations and experimental results indicate that the pinch-off regions, which provide additional two-dimensional confinement for the QWR structure, have a blocking effect on relaxation mechanisms for certain structures within the v groove. Time-resolved PL measurements show that efficient carrier relaxation from excited QWR states into the ground state occurs only at temperatures $\gtrsim 30\mathrm{K}$. Values for the low-temperature radiative lifetimes of the ground- and first excited-state excitons have been obtained ($340\mathrm{ps}$ and $160\mathrm{ps}$, respectively), and their corresponding localization lengths along the wire estimated.

Figure 1

TEM micrograph cross section showing the crescent-shaped QWR, which is separated from side QW’s by narrow pinch-off regions. The $\mathrm{Ga}$-rich VQW that also forms is seen to consist of three separate structures.

Figure 2

(a) First three QWR electronic states (with pinch-off); (b) SQW $n=4$ state with pinch-off; (c) as (b) without pinch-off barrier included in the calculation. The shaded scales indicate the normalized relative probability density. (d) Calculated electron (circle) and heavy-hole (triangle) confinement energies for the sample under investigation.

Figure 3

Temperature-dependent PL spectra, exciting at (a) $1.78\mathrm{eV}$ and (b) $2.41\mathrm{eV}$. The dotted lines give the theoretical energies for various QWR optical transitions.

Figure 4

(a) PL spectrum at $5\mathrm{K}$, exciting at $1.83\mathrm{eV}$. (b) PLE spectra as a function of temperature, detecting at the $n=2$ QWR peak energy, and (c) detecting at the $n=1$ QWR peak energy.

Figure 5

Temperature dependence of the Stokes shift for the $n=2$ (circles) and SQW (triangles) transitions. The dotted lines are guides to the eye.

Figure 6

Temporal profiles of the (a) $n=1$ and (b) $n=2$ luminescence peaks at three different temperatures. The monoexponential fits used to extract the decay times are shown.

Figure 7

Transient PL spectra (smoothed) at (a) $60\mathrm{K}$ and (b) $5\mathrm{K}$. The inset in (b) is a magnified view of the $n=1$ peak. Note the shift in the $n=2$ peak in (b).

Figure 8

Experimental luminescence decay times versus temperature, for the $n=1$ (triangles) and $n=2$ (circles) QWR states. The dashed line shows the predicted square-root dependence of the free exciton decay time with temperature, and the solid line includes localization effects, as given by Eq. (1).