The Ultimate Performance of HOT HgCdTe Photodiodes

Infrared photon detectors are typically operated at cryogenic temperatures to decrease the noise of the detector arising from various mechanisms associated with the narrow bandgap. The cooled technologies are expensive and, for many applications, are unattractive due to their prohibitive size, weight, and power signature. There have been considerable efforts to decrease system size, weight, and power consumption (SWaP) - in consequence, reducing the system’s cost - to increase the operating temperature in high-operating-temperature (HOT) detectors. The ultimate goal is the fabrication of a detector with the dark current less than the system background flux current and with insignificant l/f noise relative to the shot noise of the background flux. In 1999, the famous British scientist Tom Elliott and his co-workers wrote “that there is no fundamental obstacle to

TABLE 3.1 General Properties ofThermal Detectors








r,j, (ms)



Silicon bolometer





Metal bolometer


IX 10s


Thermistor bolometer





Germanium bolometer





Carbon bolometer





Superconducting bolometer (NbN)







(2-10) X10'°


0.1 X 1 to 0.3X3








(2-5)x 10s



Golay cell






obtaining room temperature operation of photon detectors at room temperature with background-limited performance, even in reduced fields of view” [9]. In this section, we attempt to reconsider the performance of HOT photodetectors in the infrared spectral range.

In 2007, the Teledyne research group published an empirically derived equation, known as “Rule 07”, for estimation of the dark current of P-on-n HgCdTe photodiodes versus the normalized wavelength-temperature product (ACT) [10]. This equation predicts the dark current density within a factor of 2.5 over a 13 order-of-magnitude range and is approximately (exact equation given in the reference [10]):


where Xc is the cutoff wavelength in pm, T is the operating temperature in K, q is the electron charge, and к is Boltzmanns constant (both of the latter being expressed in SI units). Rule 07 was developed for operating temperature cutoff wavelength products between 400 pmK and ~1700 pmK, and for operating temperatures above 77 K.

The Rule 07 metric is closely related to an Auger 1 diffusion-limited photodiode with n-type extrinsic doping concentration in the active region close to 1015 cnr3. Any detector architecture that is limited by Auger 7 p-type diffusion, or by depletion currents, will not behave according to Rule 07. Rule 07 is also an excellent tool for a quick comparison of an RirA product of other material systems with HgCdTe. However, caution should be taken when expanding it to other parameters, such as detectivity and lower operating temperatures.

In the past decade, the Rule 07 metric has become very popular with the IR community for other technologies (especially to III-V barrier and type- II superlattice devices) as a reference level. However, at the present stage of technology, the fully depleted background-limited HgCdTe photodiodes can achieve the level of room-temperature dark current considerably lower than that predicted by Rule 07. The discussion below explains this statement exactly.

SRH Carrier Lifetime

The Shockley-Read-Hall (SRH) generation-recombination mechanism determines the carrier lifetimes in lightly doped n- and p-type HgCdTe, in which SRH centers are associated with residual impurities and native defects. From data gathered by Kinch et al. in 2005 [11], the measured values of carrier lifetimes for LWIR n-type HgCdTe range from 2 up to 20 ps at 77 K, regardless of doping concentration, for values below 1015 cm'3. The values for mid-wave infrared (MWIR) material are typically slightly longer, in the range of 2-60 ps. However, several papers published in the past decade have shown the values of rSRH to be considerably larger in the low-temperature range and at low doping concentrations, above 200 ps up to even 50 ms, depending on the cutoff wavelength [12] (Table 3.2). The lowermost range of low doping that can be reproducibly generated in Teledyne growth HgCdTe epilayers by molecular beam epitaxy (MBE) is about 1013 cm 3. In a recently published paper by Gravrand et al. [13], it has been shown that, for most tested MWIR photodiodes from LETI and Sofradir, the estimated SRH carrier lifetimes [from direct measurements (photoconductive or photoluminescence decay) as well as indirect estimations from current-voltage (I-V) characteristics], are in the range between 10 and 100 ps. These values are lower than previously estimated by US research groups; however, the latter were estimated for photodiodes with higher doping in the active region, above 1014 cm'3. From recently published results [14], Teledyne has confirmed fabrication of depletion layer-limited P-i-N HgCdTe photodiodes, with SRH recombination centers having lifetimes in the range 0.5-10 ms.

All SRH lifetimes estimated for HgCdTe are usually carried out for temperatures below 300 K. Their extrapolation to 300 K, in order to predict the photodiode operation behavior, is questionable. In our estimation

TABLE 3.2 Summary of the SRH Carrier Lifetimes Determined on the Basis of I-V and FPA Characteristics (after Ref. [12])

x composition

TSRH (ps)



>100 at 60 К



>1000 at 110 К



-50000 at 89 К



>3000 at 180 К

P-i-N photodiode

FIGURE 3.6 P-i-N photodiode: (a) energy band diagram under reverse bias, (b) heterojunction architecture.

we assume rSRH to equal 1 ms, which is supported by experimental data achieved by DRS and Teledyne research groups.

Figure 3.6(a) shows a schematic band diagram for a reverse-biased P-i-N heterostructure photodiode. The active region consists of an undoped i-region (v region, low n-doping) sandwiched between a wider bandgap cap (P) and buffer (N) region [see Fig. 3.6(b)]. Very low doping in the absorber region (below 5x 1013 cm"3) is required to allow full depletion at zero or a low value of reverse bias [15]. The surrounded wide-bandgap contact layers are designed to suppress the dark current generation from these regions and to suppress tunneling current under reverse bias. Moreover, fully depleted absorbers, surrounded by wide-bandgap regions, potentially reduce Ilf and random telegraph noise. As previously mentioned, the fully depleted P-i-N structure is compatible with the small pixel size, achieving low crosstalk thanks to the built-in vertical electric field [12,15].

In P-i-N design, the choice of absorber thickness should be a trade-off between the response speed and quantum efficiency (or responsivity). To achieve short response times, the absorber thickness should be thin and fully depleted. For high quantum efficiency, the absorption region should be thick enough to effectively collect photogenerated carriers. However, to enhance quantum efficiency, while maintaining high response speed, an external resonant microcavity has been proposed. In this approach, the absorber is placed inside a cavity, so that a large portion of the photons can be absorbed, even with a small detection volume.

< Prev   CONTENTS   Source   Next >