The relaxation mechanisms of an array of 10 vertically coupled layers of InGaAs/AlGaAs quantum dots were studied by time-resolved photoluminescence. Both resonant and non-resonant excitation were employed and the photoluminescence (PL) intensity in the non-resonant case is a factor of 200 larger than the intensity with resonant excitation. The results obtained in the non-resonant pumping experiment were analysed with a rate equation model. It was found that the PL decay time increases rapidly with the wavelength of detection. Large carrier capture cross-sections [(2.5±0.9)×10 -5 cm 3 /s] were deduced, resulting in a capture time of 1 ps for a carrier concentration of 4×10 16 cm -3 . A very fast PL risetime was observed with resonant pumping, ruling out a phonon bottleneck effect in these samples. The decay times at a given wavelength are always shorter for resonant than for non-resonant excitation and their difference increases rapidly with wavelength. This is interpreted in terms of a state filling effect for the non-resonant case.

The processing and analysis are described of range data in a time-of-flight imaging system based on time-correlated single photon counting. The system is capable of acquiring range data accurate to 10 mu m at a standoff distance in the order of 1 m, although this can be varied substantially. It is shown how fitting of the pulsed histogram data by a combination of a symmetric key and polynomial functions can improve the accuracy and robustness of the depth data, in comparison with methods based on upsampling and centroid estimation. The imaging capability of the system is also demonstrated.

The design and operation of a noncontact surface profilometry system based on the time-correlated single-photon-counting technique are described. This system has a robust optomechanical design and uses an eye-safe laser that makes it particularly suitable for operation in an uncontrolled industrial environment. The sensitivity of the photon-counting technique permits its use on a variety of target materials, and its mode of operation does not require the continual presence of an operator. The system described has been optimized for a 1-25-m standoff, has a distance repeatability of <30 microm, and has a transverse spatial resolution of approximately 60 microm at a 2-m standoff and approximately 400 microm at a 13-m standoff.

The nature of lasing threshold in passively Q-switched GaAs/AlGaAs lasers with saturable absorbers formed by heavy ion implantation is investigated in this article. After studying various laser characteristics, including threshold current density, differential quantum efficiency, spectral output, and picosecond time-resolved emission, we conclude that the origin of the Q-switching is unlikely to be caused by spontaneous emission or mode locking, and that collective coherent radiation effects may contribute to the onset of lasing.

A time-resolved photoluminescence study of strained and unstrained InGaAsP/InP double heterostructures has been performed at low photogenerated carrier densities (i.e. ≤1016 cm−3) using a novel high-efficiency germanium photon-counting detector. The photoluminescenee decay times are observed to decrease with increasing strain. Samples grown on substrates with lattice orientation (3 1 1)B are shown to have shorter excess carrier lifetimes than those grown on lattices orientated (0 0 1) or (3 1 1)A.

A method for acquiring range data based on time-correlated single-photon counting is described. This method uses a short-pulse ( approximately 10-ps) laser diode, a detector based on a silicon single-photon avalanche diode, and standard photon-counting timing electronics. The accuracy of the technique has been measured as approximately +/-30 microm in a laboratory experiment and corresponds closely to the results of a theoretical simulation.

Time-resolved photoluminescence has been used to study the cross-well carrier dynamics in a biased multiple quantum well p-i-n structure at temperatures in the range 5–350 K and for electric fields <200 kV cm−1. The photoluminescence decays have been parameterized using reconvolution analysis with a coupled rate equation model and this has provided strong evidence for the successive recapturing of carriers in adjacent wells. For temperatures <100 K, and for electric fields <60 kV cm−1, the carrier escape appears to be strongly affected by resonant tunneling between hole subbands in adjacent wells. At higher temperatures an increase in the carrier escape rate is observed which corresponds to a field dependent thermal activation energy.

High spatial resolution time-resolved photoluminescence has been used to study GaInAs/GaInAsP quantum-well structures selectively intermixed using the pulsed photoabsorption-induced disordering technique. Photoluminescence decay measurements at wavelengths ≳1.3 μm were obtained using novel high-efficiency photon-counting detectors and were found to correlate spatially with the observed luminescence blue shift in these structures. Results indicate a reduction in the nonradiative recombination time of nearly two orders of magnitude as a result of this intermixing technique.

In this paper we present photoluminescence decay measurements from n-type, iodine-doped, ZnSe and nominally undoped Zn0.25Cd0.75SeZnSe single quantum well structures grown by molecular beam epitaxy. In the temperature range 70–500 K, the iodine-doped material shows evidence of the trapping and re-emission of carriers from three deep acceptor levels at 80, 120, and 350 meV above the valence band. Recombination in the quantum well material is dominated by radiative processes for temperatures < 100 K, whilst at room temperature measurements indicate that the non-radiative processes are considerably worse than in bulk material.

A commercially available germanium avalanche photodiode operating in the single-photon-counting mode has been used to perform time-resolved photoluminescence measurements on InGaAs/lnP multiple-quantum-well structures. Photoluminescence in the spectral region of 1.3-1.48 µm was detected with picosecond timing accuracy by use of the time-correlated single-photon counting technique. The carrier dynamics were monitored for excess photogenerated carrier densities in the range 10(18)-10(15) cm(-3). The recombination time is compared for similar InGaAs-based quantum-well structures grown by use of different epitaxial processes.

Time-resolved photoluminescence was used to study exciton recombination in deep CdTe/Cd0.5Mn0.5Te single quantum well. The width of the investigated well was 100 Angstrom. The study was performed at room temperature. The lifetime of the exciton determined in this work has a value comparable to that observed in shallow CdTe/Cd0.85Mn0.15Te quantum wells. A strong enhancement of the photoluminescence decay time with increasing intensity of the exciting laser beam is observed which is indicative of saturation of the non-radiative recombination centers.

Temperature dependent time-resolved photoluminescence has been used to study the excess carrier recombination in Zn0.75Cd0.25Se/ZnSe single quantum well structures grown by molecular beam epitaxy. For temperatures <100 K radiative excitonic recombination appears to dominate, and the photoluminescence (PL) decay time follows the linear dependence on temperature over the range 50–120 K. At higher temperatures the reduction in PL efficiency and decay time indicate that nonradiative processes associated with the ZnCdSe/ZnSe interfaces dominate the recombination. The results are consistent with theoretical predictions.

Time-resolved photoluminescence measurements have been used to study temperature dependent carrier sweepout in biased multiple quantum well p-i-n diodes.

Temperature dependent photoluminescence decay measurements have been used to study the minority carrier dynamics in iodine-doped ZnSe grown by molecular beam epitaxy. The existence of three deep acceptor levels with energies at 80, 120, and 350 meV above the valence band has been established. The 80 and 120 meV levels have a density dependence directly related to the iodine doping density whilst the level at 350 meV does not. Significant broadband donor–acceptor emission is observed from this material and appears to be associated with the acceptor level at 350 meV.

Time-resolved photoluminescence has been used to study carrier recombination in n- and p-type doped ZnSe at room temperature. A band-edge photoluminescence decay time of ∼240 ps has been measured for heavily doped n-type material together with a relaxation time of a few microseconds for the associated deep-level emission. The band-edge photoluminescence decay time for p-type doped material was ≤11 ps and is indicative of a high level of nonradiative Shockley–Read recombination.

While the speed at which a SEED device may be switched is limited fundamentally by carrier vertical transport times and RC effects, in a practical parallel processing system it is likely to be limited by the ratio of incident optical power and device switch energy. Since high speed SEED operation is desirable for a high processing rate, it is important that the device performance is not degraded at high incident power. One of the important mechanisms which result in this degradation is the build-up of a carrier density in the quantum wells (QWs). This carrier density weakens the excitonic absorption feature, resulting in a reduction of the reflectivity contrast ratio. In references [1] and [2] several QW structures were compared and the saturation effects were found to decrease as the sweep-out time [3] was reduced by either lowering or thinning the barrier between wells. An MQW consisting of 100Å GaAs wells and 35Å Ga 0.7 Al 0.3 As barriers was therefore developed in [1] and for this structure, at the intensities used in that work, no saturation effects were observed. We have extended the results by examining a similar structure and measuring the saturation of the responsivity and reflectivity at incident intensities up to 40kW/cm 2 (1400μW in a spot with 2.1μm 1/e 2 -1/e 2 intensity). Thermal effects were also observed. Using time resolved photoluminescence (TRPL) we have directly measured carrier sweep-out times and so have been able to calculate the carrier densities generated in the saturation experiments.

Time-resolved photoluminescence measurements are taken as a function of bias-voltage for MQW p-i-n diodes with different barrier thicknesses and mesa sizes. The two devices studied had different barrier thicknesses and were on standard S-SEED and L-SEED chips. The former device consists of 60.5 periods of 100 Å GaAs wells and 65 Å Al0.3Ga0.7As barriers whilst the latter consists of 70.5 periods of 100 Å GaAs wells and 35 Å Al0.3Ga0.7As barriers

A novel, entirely solid-state, instrument has been developed for the study of time-resolved photoluminescence in a wide variety of bulk semiconductors, low-dimensional structures, and other materials. This system uses a frequency-doubled GaAlAs diode laser (pulse width 30 ps) as the 415-nm pump source and a silicon single-photon avalanche diode for detection of photoluminescence. The time-correlated single photon counting technique allows measurement of photoluminescence decays in the temporal region of 10 ps to ≳500 ns with high statistical accuracy. In addition, the combination of a microscope with a small-area detector provides a spatial resolution of <5 μm. This system is currently being used for the measurement of picosecond time-resolved photoluminescence from II-VI semiconductors.

The authors present a technique for the measurement of S-SEED switching characteristics based on time-correlated single-photon counting (TCSPC). This allows the measurement of switching characteristics over a large temporal dynamic range; from that of the detector response time (<20 ps-or faster with reconvolution analysis techniques) to that of the repetition period of the pulsed laser (in this case, >1 mu s). The technique uses only one pulsed pump source on the S window to switch the device, and a probe-reset beam (at an appropriate wavelength to observe switching) on the R window, which is monitored by the detector. The probe-reset beam has a low power level, but sufficient to reset the device to its original state before the onset of the next pump pulse.< >