Superfluorescence from photoexcited semiconductor quantum wells: Magnetic field, temperature, and excitation power dependence

Superfluorescence (SF) is a many-body process in which a macroscopic polarization spontaneously builds up from an initially incoherent ensemble of excited dipoles and then cooperatively decays, producing a delayed pulse of coherent radiation. SF arising from electron-hole recombination has recently been observed in ${\mathrm{In}}_{0.2}{\mathrm{Ga}}_{0.8}\mathrm{As}/\mathrm{GaAs}$ quantum wells [G. T. Noe et al., Nature Phys. 8, 219 (2012) and J.-H. Kim et al., Sci. Rep. 3, 3283 (2013)], but its observability conditions have not been fully established. Here, by performing magnetic field $\left(B\right)$, temperature $\left(T\right)$, and pump power $\left(P\right)$ dependent studies of SF intensity, linewidth, and delay time through time-integrated and time-resolved magnetophotoluminescence spectroscopy, we have mapped out the $B-T-P$ region in which SF is observable. In general, SF can be observed only at sufficiently low temperatures, sufficiently high magnetic fields, and sufficiently high laser powers with characteristic threshold behavior. We provide theoretical insights into these behaviors based primarily on considerations on how the growth rate of macroscopic coherence depends on these parameters. These results provide fundamental new insight into electron-hole SF, highlighting the importance of Coulomb interactions among photogenerated carriers as well as various scattering processes that are absent in SF phenomena in atomic and molecular systems.

Figure 1

Schematic diagram of the experimental geometry. Spontaneous emission (SE) is emitted in all $4\pi$ spatial directions while superfluorescence (SF) is emitted in the quantum well plane. A $\mu$ prism is placed at one edge of the sample to redirect the in-plane emission (SF) into the edge fiber, while SE is collected through the center fiber placed behind the sapphire plate [23].

Figure 2

Magnetic field dependence of time-integrated PL collected with the (a) center fiber and (b) edge fiber at 4 K with an average excitation laser power of 2 mW. Magnetic field dependence of the (c) linewidth and (d) intensity of the center and edge collected PL emission of the (00) interband transition.

Figure 3

Color surface plots of time-resolved photoluminescence spectra at different magnetic fields at a temperature of 4 K and an excitation laser power of 2 mW. Here, the time ‘zero’ on the time axis corresponds to the laser pulse excitation time when the $e-h$ pairs are created. See Supplemental Material for more information [31].

Figure 4

Magnetic field dependence of SF delay time for different $(N_e,N_h)$ transitions at 4 K and 2 mW.

Figure 5

The magnetic field dependence of the excitonic enhancement $|R_m\left(0\right)|$ of the interband dipole matrix element for different Landau levels.

Figure 6

(a) Temperature dependence of time-integrated PL collected through the edge fiber at 17.5 T with an excitation power of 4 mW. (b) Emission spectra of the (11) transition at different temperatures. (c) The linewidth and intensity of the (11) emission as a function of temperature.

Figure 7

Color surface plots of time-resolved photoluminescence spectra at different temperatures at a magnetic field of 10 T and an excitation laser power of 2 mW.

Figure 8

(a) Temperature dependence of SF delay times for different $(N_e,N_h)$ at 10 T and 2 mW. (b) Temperature dependence of spectrally and temporally integrated peak intensities at 10 T and 2 mW for all emission peaks, the (00) peak and the (11) peak.

Figure 9

(a) Excitation laser power dependence of time-integrated edge PL spectra at 17.5 T and 4 K. (b) (11) emission spectra with different laser excitation powers at 17.5 T and 4 K. (c)–(d) The linewidth and intensity of the (11) emission as a function of average excitation laser power in (c) log scale and (d) linear scale.

Figure 10

Color surface plots of time-resolved photoluminescence spectra at different excitation laser powers at a magnetic field of 17.5 T and a temperature of 4 K.

Figure 11

(a)–(f): Power dependence of spectrally and temporally integrated emission intensity at different magnetic fields at 4 K, showing saturation behavior at high powers. (g)–(l): Power dependence of delay times at different magnetic fields for different $(N_e,N_h)$ peaks at 4 K.

Figure 12

(a) Magnetic field dependence of the critical laser power, $P_c$, of SF for the (11) transition at 4 K. (b) Magnetic field dependence of the critical temperature, $T_c$, of SF under different excitation powers for the (11) transition.