The paper reports on a long-wave infrared (cut-off wavelength ~ 9 μm) HgCdTe detector operating under nbiased condition and room temperature (300 K) for both short response time and high detectivity operation. The ptimal structure in terms of the response time and detectivity versus device architecture was shown. The response time of the long-wave (active layer Cd composition, xCd = 0.19) HgCdTe detector for 300 K was calculated at a level of τs ~ 1 ns for zero bias condition, while the detectivity − at a level of D* ~ 109 cmHz1/2/W assuming immersion. It was presented that parameters of the active layer and P+ barrier layer play a critical role in order to reach τs ≤ 1 ns. An extra series resistance related to the processing (RS+ in a range 5−10 Ω) increased the response time more than two times (τs ~ 2.3 ns).

A theoretical analysis of the mid-wavelength infrared range detectors based on the HgCdTe materials for high operating temperatures is presented. Numerical calculations were compared with the experimental data for HgCdTe heterostructures grown by the MOCVD on the GaAs substrates. Theoretical modelling was performed by the commercial platform SimuAPSYS (Crosslight). SimuAPSYS fully supports numerical simulations and helps understand the mechanisms occurring in the detector structures. Theoretical estimates were compared with the dark current density experimental data at the selected characteristic temperatures: 230 K and 300 K. The proper agreement between theoretical and experimental data was reached by changing Auger-1 and Auger-7 recombination rates and Shockley-Read-Hall carrier lifetime. The level of the match was confirmed by a theoretical evaluation of the current responsivity and zero-bias dynamic resistance area product (R0A) of the tested detectors.

This paper presents examples of infrared detectors with mercury cadmium telluride elaborated at the Institute of Applied Physics, Military University of Technology and VIGO Photonics S.A. Fully doped HgCdTe epilayers were grown with the metal organic chemical vapour deposition technique which provides a wide range of material composition covering the entire infrared range from 1.5 µm to 14 µm. Fundamental issues concerning the design of individual areas of the heterostructure including: the absorber, contacts, and transient layers with respect to their thickness, doping and composition were discussed. An example of determining the gain is also given pointing to the potential application of the obtained devices in avalanche photodiode detectors that can amplify weak optical signals. Selected examples of the analysis of current-voltage and spectral characteristics are shown. Multiple detectors based on a connection in series of small individual structures are also presented as a solution to overcome inherent problems of low resistance of LWIR photodiodes. The HgCdTe detectors were compared with detectors from III-V materials. The detectors based on InAs/InAsSb superlattice materials achieve very comparable parameters and, in some respects, they are even superior to those with mercury cadmium telluride.

The temperature dependence of photoluminescence spectra has been studied for the HgCdTe epilayer. At low temperatures, the signal has plenty of band-tail states and shallow/deep defects which makes it difficult to evaluate the material bandgap. In most of the published reports, the photoluminescence spectrum containing multiple peaks is analyzed using a Gaussian fit to a particular peak. However, the determination of the peak position deviates from the energy gap value. Consequently, it may seem that a blue shift with increasing temperature becomes apparent. In our approach, the main peak was fitted with the expression proportional to the product of the joint density of states and the Boltzmann distribution function. The energy gap determined on this basis coincides in the entire temperature range with the theoretical Hansen dependence for the assumed Cd molar composition of the active layer. In addition, the result coincides well with the bandgap energy determined on the basis of the cut-off wavelength at which the detector response drops to 50% of the peak value.