Exciton lifetime in $\mathrm{InAs}∕\mathrm{GaAs}$ quantum dot molecules

The exciton lifetimes $T_1$ in arrays of $\mathrm{InAs}∕\mathrm{GaAs}$ vertically coupled quantum dot pairs have been measured by time-resolved photoluminescence. A considerable reduction of $T_1$ by up to a factor of $\approx 2$ has been observed as compared to a quantum dots reference, reflecting the inter-dot coherence. Increase of the molecular coupling strength leads to a systematic decrease of $T_1$ with decreasing barrier width, as for wide barriers a fraction of structures shows reduced coupling, whereas for narrow barriers all molecules appear to be well coupled. The coherent excitons in the molecules gain the oscillator strength of the excitons in the two separate quantum dots halving the exciton lifetime. This superradiance effect contributes to the previously observed increase of the homogeneous exciton linewidth, but is weaker than the reduction of $T_2$. This shows that as compared to the quantum dots reference pure dephasing becomes increasingly important for the molecules.

Figure 1

(Color online) Gray scale contour plots of the photoluminescence emission decay from the QDs reference sample (right panel) and the QDM sample with a 4 nm wide barrier (left panel) at $T=10\mathrm{K}$. The signal was integrated in both cases for 60 s. Symbols are characteristic traces at the center of the emission band. The peak intensity has been normalized to unity. Note that there is some intensity reduction around 900 ps because of reduced sensitivity of the streak screen.

Figure 2

Photoluminescence decay traces for the QDs reference and the 8, 6, and 4 nm barrier QDMs samples on a logarithmic scale (vertically shifted for clarity), each taken at energies in the center of the inhomogeneously broadened emission peaks. Symbols give experimental data, lines are monoexponential fits (see the text for details). The inset shows the $d=4\mathrm{nm}$ data on a linear scale to stress the existence of a background of roughly constant intensity (the dotted line).

Figure 3

Temperature dependence of the decay time $T_1$ from the QDMs sample with a 4 nm wide barrier (the up triangle symbols) and the QDs reference sample (the down triangle symbols). The line is a guide for the eye and follows Eq. (1) given in the text.

Figure 4

(Color online) Dependence of different decay rates on the QDM barrier width normalized by the corresponding rate of the QDs reference (plotted on a log-log scale): The upper panel gives the exciton population decay rate $1∕T_1$ (error bars include the errors when evaluating $T_1$ in QDMs and QDs), and the middle panel gives the zero-temperature exciton dephasing rate $1∕T_2$ for $T\to 0$, from Ref. 10. From these quantities the pure dephasing rate $1∕T_2^{*}$ is calculated according to Eq. (2) and shown in the lower panel. For comparison, all the rates are shown on the same vertical scale.