Photoneutralization and slow capture of carriers in quantum dots probed by resonant excitation spectroscopy

We investigate experimentally and theoretically the resonant emission of single InAs/GaAs quantum dots in a planar microcavity. Due to the presence of at least one residual charge in the quantum dots, the resonant excitation of the neutral exciton is blocked. The influence of the residual doping on the initial quantum dots charge state is analyzed, and the resonant emission quenching is interpreted as a Coulomb blockade effect. The use of an additional nonresonant laser in a specific low power regime leads to the quantum dots neutralization and allows an efficient optical gating of the exciton resonant emission. A detailed population evolution model, developed to describe the quantum dot neutralization and the optical gate effect, perfectly fits the experimental results in the steady-state and dynamical regimes of the optical gate with a single set of parameters. We deduce that ultraslow Auger- and phonon-assisted capture processes govern the quantum dot neutralization with relaxation times in the 1–$100\mu \mathrm{s}$ range. We conclude that the optical gate acts as a very sensitive probe of the quantum dots population relaxation in an unprecedented slow-capture regime.

Figure 1

(a) Typical spectrum of the nonresonantly excited photoluminescence of a single quantum dot, showing the neutral exciton $X$ and the positive trion $X^{+}$. (b)–(e) Resonant emission spectra of $X$ ($•$) and $X^{+}$ ($\circ$) as a function of the detuning between the laser energy $E_L$ and the exciton (trion) energy $E_X$ ($E_{X^{+}}$) when the optical gate is switched off, (b) and (c), and on, (d) and (e).

Figure 2

(a) Resonant emission (RE) and nonresonant photoluminescence (PL) intensities of the neutral exciton ($X$) as a function of the optical gate power $P_{\text{gate}}$. (b) Resonant emission intensity of the neutral exciton ($X$) and the positive trion ($X^{+}$) as a function of the optical gate power $P_{\text{gate}}$. Only the region where the nonresonant photoluminescence is completely negligible is displayed. The experimental data are fitted by the population evolution model described in Sec. 4 (solid lines).

Figure 3

Energy of the neutral exciton (a) and positive trion (b) resonant emission as a function of the optical gate power $P_{\text{gate}}$. The experimental data are obtained from the same set of measurements as Fig. 2b and are fitted by the model described in Sec. 4 (solid lines).

Figure 4

(a) Scheme of a QD with a fundamental hole state $S_h$ quasiresonant with a slightly confined interfacial defect state $\Phi$. The QD is filled with two holes and the absorption of a photon at the neutral exciton energy is impossible. (b) The capture and escape processes are schematized with the corresponding rates in a diagram showing the intrinsic hole reservoir induced by the residual doping of the sample (see Appendix), the defect, and the QD.

Figure 5

(a) Sketch of the optical gate effect: Carriers are created in the GaAs barrier and captured by the QD or the defect. (b) The capture and escape processes induced by the optical gate are schematized in a diagram showing the optical reservoir formed by the charges photocreated in the GaAs barrier by the optical gate.

Figure 6

Diagram of the population evolution model.

Figure 7

(a) Results of the simulation showing the resonant emission (RE) and the nonresonant photoluminescence (PL) intensities of the neutral exciton as a function of the optical gate power $P_{\text{gate}}$. (b) Electron and hole capture rates, ${\gamma }_1$ and ${\gamma }_2$, as a function of the optical gate power $P_{\text{gate}}$.

Figure 8

(a) Temporal evolution of the neutral exciton resonant emission when the optical gate is switched off at $t=0$ ($P_{\text{gate}}=3$ $\mathrm{n}$$\mathrm{W}$). The detuning between the resonant excitation laser and the exciton energies is equal to $\delta =-1$ $\mu$$\mathrm{e}\mathrm{V}$. (b) Time constants $t_{\text{shift}}$ and $t_{\text{off}}$ as a function of the optical gate power $P_{\text{gate}}$.

Figure 9

(a) Temporal evolution of the neutral exciton resonant emission when the optical gate is switched on at $t=0$ ($P_{\text{gate}}=10$ $\mathrm{n}$$\mathrm{W}$). The detuning between the resonant excitation laser and the exciton energies is equal to $\delta =-1$ $\mu$$\mathrm{e}\mathrm{V}$. (b) Time constant $t_{\text{on}}$ as a function of the optical gate power $P_{\text{gate}}$. The theoretical fit (solid line) is given by the power laws (11) of the capture rates ${\gamma }_1$ and ${\gamma }_2$.

Figure 10

Intensity autocorrelation measurements of the optically gated resonant emission of the neutral exciton, for four optical gate powers (taken from a complete study over 12 gate powers). The experimental data correspond to the number of coincidences as a function of the time delay while the theoretical curves, results of the population evolution model, are plotted on a dual scale by adjusting the experimental and theoretical maxima. The parameters used in the theoretical model are $\gamma =3030$ $\mu$${\mathrm{s}}^{-1}$, $r_X=500$ $\mu$${\mathrm{s}}^{-1}$. ${\gamma }_1$ and ${\gamma }_2$ are given by the power laws (11).

Figure 11

(a) Sketch showing the fluctuation of the quantum dot charge state and its resonant emission when the resonant laser and the optical gate are simultaneously switched on. (b) Trend of the second-order correlation function of the exciton resonant emission characterized by the bunching time constant ${\tau }_B$. (c) Experimental values of the time constant ${\tau }_B$ characterizing the bunching decay as a function of the optical gate power $P_{\text{gate}}$ (symbols) and theoretical curve corresponding to Eq. (21) (solid line).

Figure 12

(a) (Top) Sketch along the growth direction $\left(0z\right)$ of the valence band of an empty QD surrounded by neutral acceptors from the unintentional residual carbon doping. (Bottom) Sketch of the same QD after the acceptor ionization and the resulting hole transfer towards the QD. $l_A$ is the thickness of the depletion region around the QD plane and $\mu$ the chemical potential fixed to the Fermi energy of the neutral acceptors outside the depletion region. (b) Hole occupancy and depletion length $l_A$ as a function of the bulk doping density $n_A$, for a QD density $n_{\text{QD}}=5\times {10}^7$ $\mathrm{c}$$\mathrm{m}$${}^{-2}$, at $T=7$ $\mathrm{K}$.