Explanation of quantum dot blinking without the long-lived trap hypothesis

A simple model explaining the experimental data on QDs luminescence blinking is suggested. The model does not assume the presence of the long-lived electron traps. The bleaching of the QD luminescence is a result of the Auger assisted radiationless relaxation of the excitation through the deep surface states. Possible ways of the experimental verification of the model are discussed.

Figure 1

A schematic picture of the electronic levels of the CdSe quantum dot. $1S_e,1P_e$, and $1D_e$ are the delocalized levels in the conduction band (electron states). $1S_{3∕2},1P_{3∕2}$, and $2S_{3∕2}$ are the delocalized levels in the valence band (hole states). The dashed lines represent the energies of the valence band edge and the conduction band edge in the bulk semiconductor. The dotted line represents the Fermi level.

Figure 2

The kinetic scheme of the QD relaxation after the light-induced excitation. The trapping of the hole is assisted by simultaneous electron $1S_e\to 1P_e$ excitation (Auger process) with consequent relaxation of the excited electron and hole. The energy difference between $1P_e$ state and $1S_e$ state is denoted by $\epsilon$.

Figure 3

The trapping reaction rate dependence on the coordinate $\epsilon$ in the vicinity of ${\epsilon }_1$ given by Eq. (2.8). The parameters are $A_i={10}^{12}{\mathrm{s}}^{-1}$, $\Gamma =0.2\mathrm{meV}$. The surface states are placed equidistantly ${\epsilon }_i={\epsilon }_1+(i-1)\times 1\mathrm{meV}$.

Figure 4

A schematic picture of the mechanism of the QD blinking. The $1S_e-1P_e$ energy gap $\epsilon \left(t\right)$ is the stochastic Markovian process. The thin curve shows the stationary distribution of $\epsilon$. The dependence of the QD luminescence intensity on $\epsilon$ has a sharp threshold at ${\epsilon }^{*}$. So, the luminescence is “off “ when $\epsilon >{\epsilon }^{*}$ (hatched area) and “on” otherwise.

Figure 5

The averaged single dot spectral line width as a function of the integration time (Ref. 29) fitted by the formula (4.7). The experimental measurements were taken with the excitation intensity $I=85\mathrm{W}∕{\mathrm{cm}}^2$. The parameters of the fits are ${\Delta }_E=1.02\mathrm{meV}$, ${\tau }_E=130\mathrm{s}$, for 10 K (squares); ${\Delta }_E=1.6\mathrm{meV}$, ${\tau }_E=109\mathrm{s}$, for 20 K (triangles); ${\Delta }_E=1.8\mathrm{meV}$, ${\tau }_E=99\mathrm{s}$, for 30 K (circles); ${\Delta }_E=2.8\mathrm{meV}$, ${\tau }_E=82\mathrm{s}$, for 40 K (squares).