Colloidal Quantum Dot Infrared Detectors

Theoretical estimates carried out by Martyniuk et al. [1] in 2008 indicate that the self-assembled quantum dot infrared photodetectors (QDIPs) are suitable for noncryogenic operation, especially in long-wavelength infrared (LWIR) regions. In practice, however, the reduced performance of QDIPs is the result of nonoptimal band structure and technological problems, such as QD size and density control. More recently, the attractive alternative to self-assembled epitaxial QDs has been colloidal quantum dots (CQDs), with better size tunability of the optical features and lower fabrication cost.

In the past decade, significant progress has been made in the fabrication of CQD photodetectors. Using this approach, an active region is constructed, based on 3D quantum-confined semiconductor nanoparticles, synthesized by inorganic chemistry. These nanoparticles could improve the performance of CQD photodetectors, compared with that of epitaxial QDs, as a result of many aspects, collated in Table 8.1 [2,3].

CQDs have attracted attention as a candidate material for a range of optoelectronic applications, including light-emitting diodes, lasers, optical modulators, solar cells, and photodetectors [2,3]. The first mass market for these nanoparticles appeared around 2015, as they started to be used as phosphors for TV displays [4].

TABLE 8.1 Advantages and Disadvantages of CQD Photodetectors in Comparison with Single-crystal QD Photodetectors

Introduction

CQDs operate in near/short-wave/mid-wave infrared (NIR/SWIR/MWIR) imaging (Fig. 8.1) and offer a promising alternative to the single-crystal IR materials [InGaAs, InSb, InAsSb, and HgCdTe, as well as type-II superlattices (T2SLs)].

CQDs belong to a wider class of low-dimensional materials, such as 2D materials, nanowires, and their hybrid structures. As noted in Chapters 6 and 7, effective active regions of photodetectors are fabricated by hybridization of 2D materials with CQDs. For this reason, this chapter provides a review of the fundamental properties of CQD photodetectors operating in the infrared spectral region.

Table 8.2 collates performance parameters of representative CQD photodetectors, whereas Fig. 8.2 plots representative values of responsivity for nanostructured materials versus their detection wavelengths [5]. As shown, the nanostructured materials are characterized by a broadband spectral response.

FIGURE 8.1 The wavelength ranges that can be detected by materials commonly used in imaging applications.

TABLE 8.2 The Colloidal Quantum Dots and Their Performance Parameters (after Ref. [5])

Brief View

Typically, CQD photodetectors are fabricated using conducting-polymer/ nanocrystal blends, or nanocomposites [6-10]. Nanocomposites often feature narrow-bandgap, II-VI (HgTe, HgSe) [11-13], PbSe or PbS [14-17] colloidal QDs. Usually, the reported IR photodetectors use CQDs embedded in conducting polymer matrices, such as poly[2-methoxy-5-(2-ethylhexyl oxy)-l,4-phenylenevinylene] (MEH-PPV), and exhibit photodetection in the near-IR regime (1-3 pm), corresponding to the semiconductor nanocrystal bandgap energy [6].

Colloidal QD photodetectors typically comprise a single nanocomposite layer deposited on a glass slide by spin-casting, and large-area, two-terminal, vertical devices are fabricated using p-[indium-tin-oxide (ITO)] and n-type

FIGURE 8.2 A summary of the spectral responsivity of infrared photodetectors based on different nanostructured materials (after Ref. [5]).

FIGURE 8.3 Colloidal quantum dot photodetector: (a) schematic diagram of device heterostructure in CQD/conducting polymer nanocomposites; (b) an SEM image of a PV QD detector with transport illustration of a photo-generated charge.

(aluminum) contacts, as shown schematically in Fig. 8.3(a). Figure 8.3(b) illustrates the capture and transport mechanism of a colloidal dot film.

The charge transport mechanisms in colloidal QD nanocomposites exhibit subtle differences, compared with epitaxial QDIPs. As shown in Fig. 8.4, the intra-band transitions are not exploited. Instead, bipolar, inter-band (or excitonic) transitions across the colloidal QD bandgap contribute to the photoresponse of the detector. In addition, since CQDs are electron acceptors and the polymers are typically hole conductors, the photogenerated excitons are dissociated at the QD/polymer interface.

FIGURE 8.4 Schematic diagram of energy vs. position for inter-band transitions in PbS/MEH-PPV colloidal QD-conducting polymer nanocomposites, demonstrating photocurrent generation for IR photodetection (after Ref. [18]).

FIGURE 8.5 Two classes of CQD photodetectors: (a) the photoconductor, and (b) the photodiode.

Thus, photoconduction through the nanocomposite occurs as electrons hop among the QDs and holes transport through the polymer [18].

There are two main classes of CQD photodetectors: photoconductors and photodiodes (Fig. 8.5). A photoconductor typically consists of an active semiconducting material (single-crystalline, polycrystalline, amorphous) and two ohmic metal contacts. In the case of photodiodes, the depletion layer is formed between the nanocrystal film and an ohmic contact. A planar transparent ITO film forms the opposing ohmic contact. The photodiode is backside-illuminated through transparent ITO film.

Early studies of CQD-based detectors started from photoconductors. However, in spite of the simplicity of device architecture, the performance of CQD photoconductors was limited by dark current, Ilf noise, and difficulty in the precise control of doping concentration. Improved performance was achieved using photovoltaic devices, which, in principle, decreased l//noise and dark currents.

Although the performance of IR focal plane arrays (FPAs) remains paramount, especially in defense and security applications, the ability to produce FPAs at low cost is becoming an increasingly integral part of the implementation strategy for existing and emerging infrared imaging systems. It is expected that extension of the application of CQD-based devices will be significant, especially in the area of IR imaging, which is currently dominated by epitaxial semiconductor and hybrid technologies. Hybrid technology, due to the complexity of production stages, reduces yield and increases overall cost. The IR CQD-based photodetectors are an alternative solution without these limitations.

IR FPAs are usually fabricated from two separate wafers (Section 4.2), one bearing detector photodiodes and the other containing silicon readout integrated circuits (ROICs), which are physically bonded (hybridized) together via indium bumps. The complexity of these multiple production steps (about 150 steps in total) reduces yield and increases the overall cost. The CQD-based vertical technology, shown in Fig. 8.5(b), permits fabrication of detector arrays directly onto ROIC substrates using a solution process [6,9], as illustrated in Fig. 4.14. Individual pixels are defined by the area of the metal pads arranged on the top of the ROIC surface. The 100-nm thickness of PbS CQD stacks can be used as an efficient absorption region [17]. The traditional hybrid FPAs are typically limited to small arrays (1-megapixel range), due to small detector wafers and low throughput. Pixel pitch, with a size of less than 10-pm, has already been demonstrated (Section 4.5). With a thin-film active layer integrated monolithically directly on top of the ROIC, submicron pixel size (i.e., below l-pm) can be achieved (Fig. 8.6 [17]). At present, the state-of-the-art for CMOS image sensors is 0.9-pm [19].

Malinowski et al. [17,20] described a CMOS-compatible pixel stack, based on a PbS QD tunable absorption peak. As shown in Fig. 8.7(a), the absorption peak depends on the nanocrystal size. Using larger-sized crystals (5.5-nm diameter), the peak absorption occurs at the wavelength of 1440 nm, but, using smaller-sized crystals (3.4-nm diameter), the peak is at 980 nm. This size-dependent tunability can be used in hyperspectral visible image sensors. From a performance standpoint, short-wavelength infrared (SWIR) PbS CQD has achieved detectivities greater than 10¹² Jones at room temperature, which is comparable with commercial InGaAs photodiodes [see Fig. 8.7(c)].

Among all the colloidal nanomaterials, mercury telluride (HgTe) CQDs have demonstrated wide spectral tunability from SWIR [10-13] up to THz

FIGURE 8.6 Positioning of a monolithic CQD array: higher wavelength than monolithic Si and greater resolution than hybrid alternatives (after Ref. [17]).

FIGURE 8.7 PbS CQD photodiode: (a) spectral tunability in dependence on dot size (after Ref. [17]), (b) quantum efficiency and responsivity, and (c) detectivity comparison of a PbS CQD photodiode with a commercial InGaAs photodiode (after Ref. [14]).

region [21]. For example, Fig. 8.8 shows the typical infrared absorption of different-sized thin-film FIgTe dots in SWIR and MWIR regions, with cutoff wavelengths at 2.5 pm and 4 pm, respectively.

Figure 8.9 presents the construction of a HgTe CQD photovoltaic detector on a composite ITO/sapphire substrate. HgTe CQDs, with a final thickness of ~400 nm, are deposited layer-by-layer by drop-casting, and each layer is rinsed with a solution of 1,2-ethanedithiol, hydrochloric acid, and isopropyl alcohol. A solution of ~10-nm Ag₂Te nanoparticles is then spun over the HgTe layer. To create a strongly p-doped HgTe CQD top layer, the Ag₂Te nanoparticles are treated with the HgCl₂ solution, which liberates

FIGURE 8.8 Infrared absorption of HgTe CQDs with cutoff wavelength at 2.5 pm (blue curve) and 4 pm (red curve) (after Ref. [22]).

FIGURE 8.9 HgTe CQDs photovoltaic detector: (a) schematic detector diagram, (b) cross-sectional scanning electron microscopy image of the detector (after Ref. [22]).

the Ag⁺ ions and immobilizes them as AgCl₂. All device fabrication is performed under ambient conditions.

Figure 8.10 shows steady improvement in performance over the past decade for MWIR HgTe CQD photodiodes. D' between 10¹⁰ to 10⁹ Jones at 5-pm was demonstrated for HgTe CQD devices, while maintaining a fast response time at thermoelectric cooling temperatures. The peak responsivity of 0.56 A/W is observed at about 160 К [22]. It is unlikely that CQD IR detectors will ever achieve the performance of the currently popular InGaAs, HgCdTe, InSb, and T2SL photodiodes. However, recent

FIGURE 8.10 Progress in performance improvement during the past decade for HgTe CQD photodetectors at 5 pm. The filled red and black symbols are for detectors with and without optical enhancement of the absorption, respectively (after Ref. [10]).

FIGURE 8.11 Flexible CQD photovoltaic detector: (a) flexible nanocrystal, (b) the device architecture of a flexible detector with a semitransparent top contact, and (c) the CQD layer deposited on flexible and unconventional substrates (after Ref. [22]).

demonstrations of low-cost SWIR and MWIR CQD-imaging arrays have heightened the interest in these devices.

Another unique advantage of CQD-based photodetectors noted in Table 8.1 is the mechanical flexibility. CQDs films can be stretched or compressed due to the fact that they are formed by closely packed nanocrystals linked by ligands (Fig. 8.11). The design and fabrication of flexible photodetectors have been demonstrated in many papers [e.g., 2,9,23].

Nowadays, imaging devices usually rely on planar focal plane arrays (FPAs). As has been described in Chapter 1, evolution of fourth-generation focal plane arrays is inspired by biological eyes. The human eye, despite a simple single-lens imaging system, is able to provide high-resolution imaging with adjustable zoom capability. The strategy of photodetector design, on a flexible substrate, can help in bending or folding, to shape the planar array into a hemispherical configuration.