Advanced Architectures and Emerging Materials for High‐Operating‐Temperature Infrared Photodiodes

Abstract

High‐operating‐temperature (HOT) mid‐wavelength and long‐wavelength infrared photodetectors have emerged as critical enablers for eliminating bulky cryogenic cooling systems, offering transfromative potential in developing compact, energy‐efficient infrared technologies with reduced size, weight, power, and cost. Focusing on infrared photodiodes, this review first discusses the fundamental mechanisms limiting performance at elevated operating temperatures. Subsequently, the progress in conventional epitaxial semiconductors, such as HgCdTe, InAsSb, and III‐V type‐II superlattice is reviewed, highlighting the evolution of device architectures designed to effectively suppress dark currents and approach background‐limited performance. The review then surveys recent advancements in emerging material systems for HOT infrared photodiodes, including colloidal quantum dots, 2D materials, and amorphous or polycrystalline thin films. Finally, a comparative analysis of the high‐temperature performance of devices from both conventional and emerging material systems is presented to enable benchmarked evaluation, followed by an outlook on future research directions.

This paper reviews the technology of high‐operating‐temperature mid‐ and long‐wavelength infrared photodiodes, focuses on strategies for suppressing dark current to elevate operating temperatures, and summarizes the current development status and future directions in various material systems, including epitaxial semiconductors, quantum dots, 2D materials, and amorphous or polycrystalline thin films.

1. Introduction

Infrared detection serves as a critical enabling technology across multidisciplinary domains ranging from defense systems to industrial monitoring to environmental sensing, medical diagnostics, and astronomical observation.^{[ 1 ]} The mid‐wavelength infrared (MWIR, 3–5 µm) and long‐wavelength infrared (LWIR, 8–14 µm) spectral bands hold particular strategic significance, corresponding to atmospheric transmission windows and encompassing the spectral regime of room‐temperature thermal radiation.^{[ 2 ]} This makes them uniquely suited for thermal imaging, gas sensing, and target recognition, driving significant research efforts. For high‐performance focal plane arrays (FPAs), photovoltaic photodiodes are typically preferred over photoconductors and phototransistors due to their superior characteristics, including high detectivity, linear response, fast speed, and low noise.^{[ 3 ]} Continuous technological advancements drive a growing need for cost‐effective, portable, high‐performance infrared imaging systems capable of meeting the requirements of emerging and increasingly sophisticated applications.

However, state‐of‐the‐art MWIR/LWIR photodiodes predominantly utilize narrow‐bandgap semiconductor materials. The intrinsically high carrier concentrations inherent to these materials lead to substantial thermally generated dark current and associated noise near room temperature, fundamentally limiting detector sensitivity. Consequently, achieving high detectivity necessitates suppressing this dark current, typically requiring deep cryogenic cooling via Stirling coolers and Dewars. This cooling requirement imposes significant system‐level penalties: Stirling coolers exhibit limited operational lifetimes (2000 h), high power consumption, and considerable cost, while Dewars severely impede portability and integration due to their inherent bulk and weight.^{[ 4 ]}

Reducing, and ultimately eliminating, the reliance on cryogenic cooling has become an inevitable direction for the development of infrared technology. Operating above 200 K enables thermoelectric coolers (TECs) with compact form factors, extended lifetimes (20 years), and Dewar‐free hermetic packaging.^{[ 4 ]} The ultimate objective is to realize high‐performance detectors that can operate uncooled at room temperature. This shift away from cryogenic systems yields substantial reductions in system size, weight, power, and cost (SWaP‐C), coupled with enhanced reliability and adaptability.^{[ 5 ]} In turn, these SWaP‐C advancements dramatically expand the applicability of infrared technology, facilitating its integration into emerging platforms formerly limited by cooling requirements,^{[ 6 ]} including autonomous vehicles, unmanned aerial vehicles, wearables, smart cities, and Internet of Things sensor networks (Figure 1 ).

Figure 1.

The evolution of HOT photodetectors and motivation for higher operation temperature.

As early as 1999, Elliott et al. theoretically demonstrated that achieving room‐temperature, background‐limited infrared photodetection is fundamentally possible for MWIR/LWIR photon detectors.^{[ 7 ]} The key to achieving background‐limited performance is the suppression of internally generated current to a level below the photocurrent generated by background radiation in the detector field of view. Benefiting from continuous advancements in materials science and revolutionary breakthroughs in device architecture (Figure 1), researchers have made significant progress in suppressing thermally generated dark current. These innovations have substantially raised the operating temperatures of conventional photodiodes from the liquid nitrogen range (≈77 K) to above 150 K for MWIR and above 100 K for LWIR. A milestone achievement was the development of a fully‐depleted p‐i‐n structure using HgCdTe with an extremely low carrier concentration, which enabled a single‐element detector to achieve background‐limited performance at a high temperature of 250 K.^{[ 8 ]} In parallel, a new frontier has opened with emerging material systems such as colloidal quantum dots (CQDs),^{[ 9} , ¹⁰ , ^{11 ]} 2D materials,^{[ 12} , ¹³ , ^{14 ]} and amorphous or polycrystalline thin films^{[ 15 ]} (Figure 2 ). These materials offer new avenues to enhance operating temperatures, leveraging their tunable properties and amenability to novel device architectures. Although impressive near‐background‐limited performance has been reported for some single‐element devices at room temperature, the development of a HOT FPA capable of background‐limited performance above 250 K—a temperature accessible by single‐stage TECs—remains a central challenge for the field.

Figure 2.

Conventional and emerging materials for MWIR and LWIR detection.

First, the fundamental principles limiting the performance of HOT infrared photodiodes are discussed. Subsequently, established device engineering strategies employed in conventional epitaxial semiconductors to enhance operating temperatures are summarized. A detailed survey follows, covering recent progress in emerging material systems such as CQDs, 2D materials, and amorphous or polycrystalline thin films for HOT photodiodes and FPAs. These strategies are then compared and evaluated. Finally, a summary of current key challenges and an outlook on future research avenues are presented.

2. Fundamental Performance Limitations of Infrared Photodiodes

To establish the fundamental principles limiting infrared photodiode performance with respect to operating temperature, this discussion examines carrier recombination processes, dark current and noise mechanisms, analyzed within the context of a reverse‐biased p‐n junction photodiode featuring a lightly doped n‐type absorber. Furthermore, the background‐limited performance, along with the temperature dependence of imaging system sensitivity, are explored.

2.1. Carrier Recombination Process

Carrier recombination processes dictate the carrier lifetime, subsequently impacting device dark current and noise. In narrow‐bandgap semiconductors, three primary mechanisms dominate: radiative recombination, Auger recombination, and Shockley‐Read‐Hall (SRH) recombination as depicted in Figure 3a. The overall lifetime is determined by their combined rates.

Figure 3.

Fundamentals of HOT infrared photodiodes. a) Key recombination mechanisms. b) Diffusion and G‐R dark currents. c) Tunneling dark current. d) Surface leakage current. e) Geometry for background flux collection. f) D_BLIP ^*​ and background photocurrent versus cutoff wavelength. g) Relative NETD (NETD/NETD@100 K) versus temperature for an MWIR FPA, including fitted curve.

Radiative recombination involves the direct recombination of an electron and hole accompanied by photon emission. In an n‐type absorber, radiative lifetime τ_R​ depends on the radiative coefficient B and is approximately inversely proportional to the doping concentration N_d.^{[ 16 ]}

$${\tau }_R=\frac{1}{B{N}_d}$$ (1)

Auger recombination, a three‐carrier non‐radiative process, occurs when electron‐hole recombination energy excites a third carrier to a higher energy state. Among various Auger types, the dominant Auger‐1 lifetime τ_A in n‐type semiconductors is often approximated by:^{[ 1 ]}

$${\tau }_A=\frac{2n_i^2{\tau }_A^i}{N_d^2}$$ (2)

where n _i​ is the intrinsic carrier concentration, τ _A ^i​ represents the intrinsic Auger lifetime, which is sensitive to temperature and bandgap. Due to its strong dependence on carrier concentration, Auger recombination frequently limits the carrier lifetime in narrow‐bandgap materials, especially under conditions of high temperature or significant doping.

SRH recombination proceeds non‐radiatively via defect levels within the bandgap. The SRH lifetime τ _SRH​ in n‐type material can be expressed as:^{[ 17 ]}

$${\tau }_{SRH}=\frac{{\tau }_p\left(n+n_i\right)+{\tau }_n\left(p+p_1\right)}{n+p}$$ (3)

where ${\tau }_n=\frac{1}{({\sigma }_n{V}_t{N}_T)}$ and ${\tau }_p=\frac{1}{({\sigma }_p{V}_t{N}_T)}$ are the fundamental electron and hole capture lifetimes, determined by the trap density N_t​, thermal velocity V_th​, and capture cross‐sections σ _n and σ _p ​. The terms n₁​ and p_1​ are related to the trap energy level E_t. Thus, the SRH recombination rate is governed directly by the material's defect properties: trap density, energy level, and capture cross‐sections. Auger and SRH recombination processes predominantly govern the carrier lifetime at high operating temperatures.

2.2. Dark Current Mechanisms

Dark current, which fundamentally limits photodiode sensitivity and operating temperature, typically comprises four main components: diffusion current, space‐charge region generation‐recombination (G‐R) current, tunneling current, and surface leakage current.^{[ 16 ]}

Diffusion current in a reverse‐biased p‐n junction photodiode (Figure 3b) arises from the concentration gradient of minority carriers diffusing from the neutral region into the space‐charge region, where their concentration is below equilibrium. For an n‐type absorber, the diffusion current density J _diff​ is expressed as:^{[ 18 ]}

$$J_{diff}=\frac{q{n}_i^2{D}_h}{L_h{N}_d}\left(e^{\frac{qV}{kT}}-1\right)=\sqrt{\frac{{\mu }_{h}kTq}{{\tau }_h}}\frac{n_i^2}{N_d}\left(e^{\frac{qV}{kT}}-1\right)$$ (4)

where V is the applied reverse bias, D _h​ is the hole diffusion coefficient, L _h is the hole diffusion length, q is the elementary charge, k is the Boltzmann constant, T is the absolute temperature, µ _h​ is the hole mobility and τ _h​ is the minority carrier lifetime. The diffusion current's primary dependence is on carrier lifetime and intrinsic carrier concentration (n _i​). Because n _i​ increases exponentially with temperature ($n_i\propto e^{-\frac{E_g}{2kT}}$), where E_g​ is the bandgap,^{[ 18 ]} diffusion current is highly temperature‐sensitive and typically dominates the total dark current near room temperature.

G‐R current arises within the space‐charge region where carrier concentrations are significantly depleted below their equilibrium values. In this non‐equilibrium condition, the thermal generation rate of electron‐hole pairs exceeds the recombination rate. These net‐generated carriers are then swept out by the reverse electric field, constituting the G‐R current. This process is typically dominated by SRH mechanisms. The resulting current density J_gr​ can be expressed as:^{[ 18 ]}

$$J_{gr}=\frac{qW{n}_i}{2{\tau }_{SRH}}\left(e^{\frac{qv}{kT}}-1\right)$$ (5)

where W is the space‐charge region width. Since J _gr ​is proportional to n _i, its temperature dependence is weaker than that of the diffusion current. Consequently, G‐R current often dominates the total dark current at lower operating temperatures or in materials with higher defect densities.

Under high electric field conditions (Figure 3c), carriers can also contribute to dark current via quantum tunneling effects. Two primary mechanisms are relevant: band‐to‐band tunneling (BTBT) and trap‐assisted tunneling (TAT).^{[ 1 ]} BTBT involves the direct tunneling of electrons from the valence band to the conduction band. Its generation rate is strongly dependent on the electric field strength, bandgap​, and effective mass, but exhibits very weak dependence on temperature. Consequently, BTBT often becomes the dominant dark current component at ultra‐low temperatures, under high doping concentrations, or at large reverse biases. TAT, in contrast, is an indirectly tunneling process by way of intermediate trap sites in the junction region. The characteristics of TAT current are closely linked to the properties of the defect states.

Surface leakage current originates from the device surface, an interruption of the crystal lattice prone to defects like dangling bonds that create surface states. These states often serve as efficient SRH recombination centers (Figure 3d), increasing surface recombination velocity and potentially enhancing bulk diffusion current near the junction. Additionally, surface potential, influenced by passivation quality, processing, adsorbates, and fixed charges, can induce surface accumulation or depletion, forming a surface space‐charge region. This surface space‐charge region contributes directly to dark current through surface G‐R and tunneling processes. The magnitude of surface leakage is thus highly sensitive to device geometry (perimeter‐to‐area ratio) and the effectiveness of surface treatment and passivation, representing a key control point in fabrication. Consequently, once surface leakage currents are effectively suppressed through passivation, the performance of HOT photodiodes is typically limited by diffusion and G‐R currents.

2.3. Noise Mechanisms

Noise denotes the random fluctuations limiting a detector's signal discernment. Key noise sources in the photodiode encompass thermal noise, shot noise, 1/f noise, and readout noise.^{[ 16 ]} Thermal noise results from the random thermal motion of carriers within the device resistance, existing even at zero bias without illumination. The noise current (I _n ) per unit bandwidth (Δf) is given by:^{[ 1 ]}

$$I_n^2=\frac{4kT}{R_0}\mathrm{\Delta }f$$ (6)

where R ₀​ is the zero‐bias dynamic resistance. The resistance‐area product R _0​A is a crucial figure of merit. Higher R ₀​A signifies better junction quality and lower thermal noise.

Shot noise is a result of the statistical process of carriers flowing over a potential barrier. Its magnitude depends on the average current flow, with the shot noise current including contributions from dark current (I _dark​), and photocurrent (I _ph​):^{[ 1 ]}

$$I_n^2=2q\left(I_{dark}+I_{ph}\right)\mathrm{\Delta }f$$ (7)

Shot noise often represents the dominant noise mechanism in photodiodes operating under reverse bias conditions.

1/f noise is another important component, particularly at low frequencies. It is characterized by a noise power spectral density inversely proportional to frequency, becoming prominent typically below 1 kHz. Its complex origins are often linked to material defects, surface states, and contacts.^{[ 1 ]} However, for high‐speed photodiode applications, 1/f noise is often not the primary limiting factor.

Finally, readout noise originates from the readout integrated circuit (ROIC) connected to the detector array and components such as reset noise and amplifier noise.^{[ 1 ]}

2.4. Background‐Limited Performance

The sensitivity of a photodetector is typically quantified by the specific detectivity (D ^*), defined as the area‐ (A) and bandwidth‐ (Δf) normalized ratio of current responsivity (R​) to noise current:^{[ 19 ]}

$$D_{\lambda }^{\ast }=\frac{R\sqrt{A\mathrm{\Delta }f}}{i_n}$$ (8)

The R is given by^{[ 19} , ^{20 ]}

$$R=\frac{q\beta \eta \lambda }{hc}$$ (9)

where β is gain (β = 1 for ideal photodiodes), η​ is external quantum efficiency, λ is wavelength, h is Planck's constant, and c is the speed of light. A photodetector's theoretical performance ceiling is set by the background‐limited condition, where thermal and dark current shot noise is insignificant relative to the background photon shot noise. In this limit, D ^* attains its maximum value:^{[ 1 ]}

$$D_{BLIP}^{\ast }\left(\lambda \right)=\frac{\lambda }{hc}\sqrt{\frac{\eta }{2Q_B}\mathrm{\Delta }f}$$ (10)

The background flux Q_B for blackbody radiation at temperature T​ viewed within a specific field of view angle θ (Figure 3e) can be calculated by integrating the Planck distribution over the detector's spectral response range:

$$Q_B={\sin }^2\frac{\theta }{2}{\int }_0^{\lambda }\frac{2\pi c}{{\lambda }^4\left(e^{\frac{hc}{\lambda k{T}_B}}-1\right)}d\lambda$$ (11)

Assuming ideal parameters (η​ = 1, θ = 180°, and T = 300 K), the background photocurrent and $D_{BLIP}^{\ast }$ as a function of wavelength can be plotted (Figure 3f). The background‐limited temperature (T_BLIP) defines the operating temperature where the dark current equals to the background photocurrent. A higher T _BLIP enables near‐background‐limited sensitivity at higher operating temperature.

2.5. Temperature Dependence of Imaging Sensitivity

The thermal sensitivity of imaging systems is principally characterized by the noise equivalent temperature difference (NETD). Lower NETD values correspond to higher sensitivity, allowing for the resolution of smaller thermal variations. For FPAs utilizing a ROIC for charge integration, NETD is given by:^{[ 21 ]}

$$\mathrm{NETD}=\frac{1+\frac{J_{dark}}{J_B}}{\sqrt{N_W}C}$$ (12)

where N_W denotes the ROIC well capacity, and C is the scene contrast.

The sharp increase in dark current with operating temperature causes the NETD to deteriorate,^{[ 22 ]} as depicted in Figure 3g. In diffusion‐limited HgCdTe FPAs, $J_{dark}\propto n_i^2\propto e^{-\frac{E_g}{kT}}$. Incorporating this into the formula 12 leads to an approximate temperature dependence:

$$\mathrm{Relative}\mathrm{NETD}=a+b{e}^{-\frac{E_g}{kT}}$$ (13)

where a and b are constant numbers. This model aligns well with the experimental data, which typically show a threshold temperature above which NETD rises rapidly. This threshold signifies the physical transition from background‐limited performance to a regime dominated by internally generated dark current. To quantify this transition, the T _BLIP​ is formally defined as the temperature at which the dark current equals the background photocurrent, which causes the NETD to double relative to its base background‐limited performance. Consequently, T _BLIP is considered the maximum practical operating temperature, and devices are generally operated slightly below this threshold to ensure optimal performance.

3. Evolution of Architectures in Conventional HOT Infrared Photodiodes

A core strategy for enhancing the operating temperature of conventional epitaxial infrared photodiodes, such as those based on HgCdTe and III–V materials, involves optimizing the device structure to suppress dark current. The evolution of device engineering within these mature material systems is reviewed herein, focusing on how various architectures specifically target and mitigate different dark current components, thereby enabling higher operating temperatures.

3.1. HgCdTe

Widely considered an ideal infrared detection material, HgCdTe offers a compositionally tunable bandgap (1–30 µm) and high optical absorption, yielding high‐performance photodiodes.^{[ 21 ]} Initial HgCdTe photodiodes typically used an n‐on‐p homojunction design (Figure 4a). This involved creating the n‐region via boron‐ion implantation and relying on Hg vacancies from post annealing as acceptors in the p‐type absorber. Hg vacancies for p‐doping in HgCdTe introduces SRH centers that degrade carrier lifetimes, restricting operation to 80 K.^{[ 23 ]} Replacing Hg vacancies with extrinsic p‐dopants (As, Au, Cu) improves lifetimes and allows higher operating temperatures in n‐on‐p devices.^{[ 23} , ^{24 ]} Consequently, controlled Au doping (10¹⁶ cm⁻³) significantly lowered dark currents in AIM's n‐on‐p MWIR FPAs, raising their effective operating temperature from 80 K to 140 K.^{[ 25 ]}

Figure 4.

HgCdTe photodiodes. Schematics of device architectures: a) n‐on‐p, b) p‐on‐n, and d) p‐i‐n. c) Temperature dependence of NETD for MWIR FPAs with varying device structures.^{[ 22 ]} e) Dark current benchmarks: Rule 07 and Law 19 curves. f) Comparison of T_BLIP for Rule 07 and Law 19.

The p‐on‐n heterojunction architecture was introduced to further suppress dark current and elevate operating temperatures in HgCdTe photodiodes. It typically utilizes an In‐doped n‐type absorber, leveraging its longer minority carrier lifetime, with an extrinsic As‐doped p‐type layer (Figure 4b). The wider bandgap of the cap layer significantly reduce the leakage currents.^{[ 23 ]} P‐on‐n technology has facilitated high‐performance MWIR array operation up to 160 K.^{[ 26} , ^{27 ]} Moreover, reducing the n‐type absorber doping below 1 × 10¹⁵ cm⁻³ can further mitigate Auger recombination, potentially pushing operating temperatures above 180 K (Figure 4c).^{[ 22 ]} Similarly, Sofradir (now Lynred) reported operating temperature increases from 80 to 150 K for their MWIR FPAs.^{[ 28 ]}

Representing a state‐of‐the‐art approach for HOT performance, the fully‐depleted p‐i‐n heterojunction architecture (Figure 4d) was developed to minimize dark current.^{[ 8} , ²⁹ , ^{30 ]} This architecture features a near‐intrinsic absorber positioned between wider‐bandgap n‐buffer and p‐cap layers. Precise control over the i‐layer's ultra‐low doping below 5 × 10¹³ cm⁻³ via molecular beam epitaxy, is critical. It minimizes Auger recombination, thus suppressing the diffusion current. With diffusion current mitigated, the G‐R current becomes the limiting factor. However, the excellent material quality also ensures remarkably long SRH lifetimes (up to 10 ms for MWIR and 0.5 ms for LWIR devices^{[ 8 ]}), keeping this G‐R component well below the background photocurrent at high temperatures. As a result, p‐i‐n structured HgCdTe photodiodes achieve exceptionally high operating temperatures approaching the background‐limited performance.^{[ 30 ]} For instance, when tested with a surrounding dewar cold shield, MWIR and LWIR devices with cutoff wavelengths of 5.9 and 10.2 µm, respectively, demonstrated that their dark current at 250 K was dominated by background radiation.^{[ 8 ]}

Empirical benchmarks aid in quantifying and comparing HgCdTe photodiode dark currents. The most prominent empirical benchmark is Rule 07,^{[ 31} , ^{32 ]} derived by Tennant et al. from p‐on‐n devices. This rule benchmarks dark current density against the product of cutoff wavelength and operating temperature, and is generally applicable for 3–16 µm cutoff wavelength and 40–200 K temperature (Figure 4e). Its $J_{dark}\propto e^{-1.16\frac{E_g}{kT}}$ dependence suggests a baseline limited by diffusion current and Auger recombination lifetime.^{[ 31 ]} Notably, actual device dark currents can deviate from Rule 07, particularly at lower temperatures where tunneling may dominate, or at shorter wavelengths where G‐R currents can be more significant. Extended empirical rules Rule 22, have been proposed to encompass broader operational ranges.^{[ 33 ]} For p‐i‐n devices with background‐limited performance, Law 19 was introduced to describe the achievable performance.^{[ 8 ]} In this regime, intrinsic dark current is negligible relative to background photocurrent, and Law 19 represents the ultimate performance level imposed by the background flux under the specific operating conditions.

The substantial increase in operating temperature enabled by improving dark current from the Rule 07 toward the Law 19 is illustrated in Figure 4f. Specifically, for a standard F/1 imaging scenario viewing a 300 K scene (assuming η = 1, J _dark​ = J _scene​ limit), the achievable operating temperature is projected to rise from 202 K to 266 K for a 5 µm MWIR detector, and from 130 K to 245 K for a 10 µm LWIR detector. Such significant enhancements highlight the potential for realizing high‐performance MWIR and LWIR imaging systems cooled by practical single‐stage TECs.

Due to its outstanding material properties, HgCdTe remains the dominant technology for high‐performance infrared detectors. The fully‐depleted p‐i‐n heterojunction architecture, based on ultra‐low carrier concentrations, positions HgCdTe as a strong candidate for achieving background‐limited performance in FPAs cooled by single‐stage TECs. However, a significant practical challenge remains in achieving the required low carrier concentration with high uniformity over large‐area wafers, a key requirement for producing high‐yield, affordable FPAs. To push operating temperatures even higher, suppressing the internal radiative current becomes essential. Although a reverse‐biased photodiode exhibits negative luminescence and does not significantly radiate outwards, blackbody radiation from neutral regions and the underlying substrate can emit photons with energy greater than the bandgap.^{[ 34 ]} These photons can be reflected from the substrate‐air interface back into the detector, increasing the internal dark current.^{[ 8 ]} Therefore, the next frontier in device engineering involves advanced optical and structural designs. Strategies such as replacing neutral regions with wider‐bandgap materials or isolating the active layer with a highly reflective mirror are promising avenues to mitigate this internal radiative flux and further elevate the operating temperature of HgCdTe detectors.

3.2. InAsSb

The InAsSb ternary alloy, similarly featuring a compositionally tunable bandgap, is widely employed in MWIR and LWIR photodetectors.^{[ 4 ]} However, compared to HgCdTe, InAsSb alloys often exhibit shorter SRH lifetimes.^{[ 35 ]} This characteristic leads to the dark current in conventional p‐n junctions based on these materials being dominated by G‐R processes at elevated temperatures, rather than diffusion current. Consequently, the barrier detector architecture was developed specifically to suppress this dominant G‐R current and facilitate operation at higher temperatures.

The nBn detector architecture, first reported for III–V materials by Maimon et al. in 2006^{[ 36 ]}, strategically inserts a wide‐bandgap electron barrier layer AlAs_0.18Sb_0.82, between an n‐type absorber and contact layer (Figure 5a). Engineered for minimal valence band offset but large conduction band offset, the barrier permits unimpeded transport of photogenerated carriers, while blocking majority carrier dark current from the contact. It also passivates the surface, reducing leakage current. Crucially, by replacing the p‐n junction's depletion region, the nBn structure eliminates the associated G‐R dark current. As demonstrated by Arrhenius plot comparisons^{[ 37 ]} (Figure 5b), the dark current in nBn detectors is primarily diffusion‐limited, in contrast to G‐R dominated p‐n junctions which exhibit a $J_{dark}\propto e^{-\frac{E_g}{2kT}}$ dependence. Consequently, nBn detectors can achieve higher operating temperatures for equivalent dark current levels below the crossover point. This concept spurred the development of related unipolar barrier structures like xBn and xBp.^{[ 38} , ^{39 ]}

Figure 5.

InAsSb nBn photodetector. a) Schematic energy‐band structure and carrier transport process in nBn barrier detector. b) Schematic Arrhenius plot of the dark current in a standard diode and in an nBn device. c) Filtered current components result from the barrier placement. d) R₀A product of a conventional photodiode and a comparable n‐side barrier photodiode.^{[ 37 ]} Reproduced (Adapted) with permission. Copyright 2013, Elsevier.

Barrier layers can also be integrated into traditional p‐n photodiode structures. As depicted in Figure 5c, the placement of the barrier determines which dark current components are blocked.^{[ 37 ]} Since G‐R and tunneling currents originate primarily within the p‐n junction's depletion region, placing an electron barrier on the n‐side is particularly effective. It truncates the extension of the depletion region into the n‐type absorber, thereby suppressing these dominant currents. In contrast, a p‐side barrier has minimal impact on the core junction region and primarily serves to block surface leakage currents. Crucially, as the barrier is engineered to be transparent to minority carrier transport, neither placement is intended to affect the collection of photogenerated carriers or the fundamental diffusion current. Savich et al. reported an n‐side barrier photodiode using InAs with an inserted AlAs_0.18Sb_0.82 electron barrier.^{[ 37 ]} As shown in Figure 5d, the R₀A values for these devices were significantly higher than conventional p‐n photodiodes at the same temperature, exhibiting performance approaching the Rule 07^{[ 37 ]}. Furthermore, nBn InAs_0.915Sb_0.085 MWIR detectors (cutoff 4.5 µm) using an AlAs_0.1Sb_0.9 barrier, reported by NASA's jet propulsion laboratory, achieved background‐limited performance at 200 K, demonstrating potential for room‐temperature operation.^{[ 40 ]} The Israeli company SCD has successfully developed several series of HOT FPA products, including Pelican D, Kinglet, Hercules, and Crane, based on InAsSb nBn structures. These typically feature cutoff wavelengths around 4.2 µm and operating temperatures of 150 K.^{[ 41} , ^{42 ]}

A key advantage of the InAsSb material system compared to HgCdTe lies in its band engineering flexibility. Through compositional tuning, InAsSb alloys can be grown to form a nearly ideal unipolar electron barrier with a near‐zero valence band offset. This allows for the unimpeded transport of photogenerated holes. In contrast, creating such a barrier in HgCdTe is more challenging, as changes in composition affect both bands, often creating an undesirable impediment to photocarrier collection. However, this architectural success reveals a more fundamental material limitation. While barrier structures masterfully suppress G‐R currents and expose the diffusion current as the performance limit, InAsSb's intrinsically shorter carrier lifetime compared to high‐quality HgCdTe becomes a critical bottleneck. Consequently, even when G‐R current is eliminated, the resulting diffusion current floor in InAsSb remains stubbornly high. Therefore, the central challenge for elevating the operating temperature of InAsSb detectors has shifted from device architecture back to fundamental materials science: optimizing epitaxial growth processes to significantly enhance carrier lifetime is the essential next step.

3.3. III–V Type‐II Superlattice

Barrier detector concepts have been effectively extended to Type‐II Superlattice (T2SL) systems. T2SLs, especially materials like InAs/GaSb and InAs/InAsSb from the 6.1 Å family, feature engineered band structures (Figure 6a). This allows for continuously tunable effective bandgaps (0–250 meV), granting advantages over bulk InAsSb, particularly for LWIR and VLWIR detection.^{[ 43 ]} Their potential for near‐zero valence band offset with Al(As)Sb‐based layers makes InAs/GaSb and InAs/InAsSb T2SLs highly suitable for barrier integration. Electron barriers can be realized, for instance, by thinning InAs layers or using GaSb/AlSb superlattices.^{[ 43 ]} Additionally, sophisticated unit cell designs like the “W”, “M”, and “N” structures have been employed to create engineered unipolar barriers for electrons or holes.^{[ 44} , ⁴⁵ , ⁴⁶ , ^{47 ]}

Figure 6.

T2SL photodetector. a) Band diagram of electron and hole minibands in InAs/GaSb T2SL and InAs/InAsSb T2SL. Developed barrier structures including b) nBn, c) pBp, and d) CBRID. e) Plasmonically enhanced structure. f) Interband cascade structure. g) Schematic illustration of the interband cascade infrared photodetectors.

By inserting an InAs/AlAsSb superlattice as an electron barrier between the n‐type absorber and contact layers in an InAs/InAsSb T2SL structure (Figure 6b), an LWIR nBn detector achieved a D^* of 4.72 ×10¹¹ cm·Hz^1/2·W⁻¹ at 77 K (10 µm cutoff) and attained background‐limited performance at 110 K^{[ 48 ]}. As depicted in Figure 6c, the pBp architecture employs an M‐structure superlattice barrier engineered to impede dark current from majority holes in the p‐type regions while facilitating the transport of photogenerated minority electrons.^{[ 49 ]} This approach enabled InAs/GaSb T2SL MWIR detectors to reach a background‐limited operating temperature of 165 K (4.9 µm cutoff).^{[ 50 ]} Subsequent FPA development using this structure, such as the 320 × 256 array by Northwestern University, yielded excellent performance at 120 K, achieving an NETD of 23 mK.^{[ 51 ]} The complementary barrier infrared detector (CBIRD) enhances dark current suppression by integrating both electron and hole barriers on opposite sides of the absorber (Figure 6d). This design has enabled detectors with a 10 µm cutoff wavelength to reach background‐limited performance at 87 K, achieving performance close to Rule 07.^{[ 52 ]}

The long carrier lifetime of Ga‐free InAs/InAsSb T2SLs is advantageous for high‐temperature LWIR detection, but this benefit is offset by weaker optical absorption compared to traditional InAs/GaSb, creating a trade‐off between photon absorption and carrier collection. To overcome this limitation, plasmonic enhancement has emerged as a compelling strategy to boost light‐matter interaction. Leland et al. demonstrated a successful implementation of this approach by employing an array of square metal patch antennas (3 µm period, 1.5 µm width) to efficiently couple incident light into surface plasmon polariton modes at the N⁺⁺/absorber interface (Figure 6e).^{[ 53 ]} This design achieved a 50% external quantum efficiency at 10.4 µm in an ultra‐thin 300 nm absorber, a performance comparable to HgCdTe. Furthermore, by tailoring the effective bandgap of the absorber at elevated temperatures to match the spectral position of the plasmonic mode, they demonstrated operation up to 230 K, achieving a specific detectivity of 2.29 × 10⁹ cm·Hz^1/2·W⁻¹ at 9.6 µm with dark current below the Rule 07 benchmark.^{[ 54 ]} This work highlights the immense potential of plasmonic for developing HOT T2SL detectors.

In parallel with these optical enhancements, effective solutions have also been developed at the device‐architecture level. Interband cascade infrared photodetectors (ICIPs) based on T2SL absorbers, circumvent this limitation by utilizing multiple short, discrete absorber stages connected in series, allowing the total thickness to ensure high absorption while keeping each stage thinner than the diffusion length for efficient carrier collection.^{[ 43 ]} As shown in Figures 6f,g, each stage typically comprises the absorber placed between wider‐gap electron and hole barriers. Smooth inter‐stage transport is achieved via engineered tunnel junctions based on Type II broken‐gap alignment. This architecture maintains the low‐noise benefits of unipolar barriers within each stage while effectively decoupling the absorption and collection length requirements.

A seven‐stage interband cascade detector reported by Tian et al.^{[ 55 ]} used InAs/GaSb T2SL absorbers, with InAs/AlSb quantum wells forming an energy staircase relaxation region and AlSb/GaSb/AlSb quantum wells acting as the tunneling electron barrier. The device exhibited a thermal noise limited D^* of 2.5 × 10¹¹ cm·Hz^1/2·W⁻¹ at 150 K (4 µm cutoff). The viability of this architecture for imaging was further demonstrated by the development of a 320 × 256 FPA employing a five‐stage MWIR interband cascade design.^{[ 56 ]}

The superior design flexibility of T2SLs over bulk materials facilitates advanced device concepts to raise operating temperatures including barrier, cascade, and plasmonic structures. The success of any such advanced design, however, is predicated on overcoming the immense challenge of near‐perfect material epitaxy. Achieving atomic‐level control over the material's growth to prevent performance‐killing dislocations remains the central hurdle. Ultimately, future progress depends on the synergistic co‐optimization of device structures, material systems, and fabrication processes.

4. Emerging Materials in HOT Infrared Photodiodes

Emerging material platforms, including CQDs, 2D materials, and amorphous or polycrystalline thin films, have recently opened new avenues for HOT MWIR and LWIR photodiode development. Progress involving these emerging materials in HOT infrared photodiodes and FPAs is subsequently reviewed.

4.1. Colloidal Quantum Dots

CQDs are semiconductor nanocrystals, typically ranging in size from 2 to 20 nm. Unlike conventional semiconductor materials that necessitate complex epitaxial growth, CQDs can be chemically synthesized via solution‐based methods.^{[ 57 ]} This approach offers significant advantages, including low production cost and facile deposition over large areas using techniques like spin‐coating or printing. Furthermore, due to the quantum confinement effect, the effective bandgap of CQDs increases as their size decreases below the material's bulk exciton Bohr radius (Figure 7a).^{[ 58 ]} This size‐dependent bandgap allows for precise tuning of their absorption spectrum across different infrared wavebands simply by controlling nanocrystal size during synthesis.^{[ 59 ]}

Figure 7.

CQD photodiodes. Tunability of CQD a) bandgap and b) electronic properties.^{[ 9 ]} Reproduced (Adapted) with permission. Copyright 2023, Springer Nature. c) p‐i‐n homojunction HgTe CQD photodetector and d) its temperature‐dependent performance.^{[ 9 ]} Reproduced (Adapted) with permission. Copyright 2023, Springer Nature. e) Planar p‐n homojunction HgTe CQD photodetector and f) its dark current characteristics.^{[ 10 ]} Reproduced (Adapted) with permission. Copyright 2023, American Association for the Advancement of Science. g) Barrier heterojunction HgTe CQD photodetector and h) its temperature‐dependent detectivity performance.^{[ 11 ]} Reproduced (Adapted) with permission. Copyright 2025, Wiley.

Quantum confinement in CQDs has also been shown to strongly suppress Auger recombination. Notably, the Auger recombination coefficient for HgTe CQDs is about 10³ times lower than that of bulk HgCdTe with a comparable bandgap.^{[ 60 ]} Consequently, the extended Auger lifetimes lead to reduced diffusion currents, offering a key theoretical advantage for realizing HOT detectors based on CQDs. Carrier transport within CQD solids predominantly occurs via inter‐dot hopping, yielding mobilities significantly lower than bulk counterparts.^{[ 61} , ⁶² , ^{63 ]} This low mobility presents a design trade‐off: it can suppress diffusion current but may also impede efficient photocarrier collection, thus lowering external quantum efficiency.

Key challenges for CQDs involve detrimental trap states, which strongly promote SRH recombination. These trap states originate from surface dangling bonds as well as from imperfect separation and subsequent agglomeration during processing.^{[ 64 ]} Surface passivation using organic or inorganic ligands is standard practice to mitigate defects in CQDs. Moreover, these ligands critically modulate key electronic properties, including energy levels,^{[ 65 ]} carrier mobility,^{[ 66 ]} and even doping type and concentration.^{[ 67 ]} As exemplified in Figure 7b, controlled doping can be achieved through ligand exchange, using HgCl₂ for p‐type and (NH₄)₂S for n‐type doping in HgTe CQD films.^{[ 9 ]}

Mercury chalcogenide CQDs (HgS, HgSe, HgTe) are prominent candidates for MWIR and LWIR detection owing to their tunable bandgaps, with HgTe CQD photodiodes being the most developed.^{[ 68} , ⁶⁹ , ^{70 ]} Xue et al.^{[ 9 ]} demonstrated such devices using p‐i‐n homojunctions formed by ligand exchange doping control (Figure 7c). These devices showed a higher detectivity than p‐i/i‐n counterparts, raising the T _BLIP from 120 to 200 K. Notably, detectivity remained above 1 × 10¹⁰ cm·Hz^1/2·W⁻¹ at 3.5 µm even near room temperature (280 K) (Figure 7d). Despite this promising performance, the poor uniformity and control associated with ligand exchange doping hinder the fabrication of high‐quality FPAs via this homojunction method.^{[ 10 ]}

As a solution to the doping control challenges in CQD films, Qin et al.^{[ 10 ]} introduced a controllable in situ electric‐field‐activated doping technique for creating lateral p‐n junctions in HgTe CQD films (Figure 7e). This method utilizes polarity‐dependent electric fields to drive metal ion migration, forming localized p‐ and n‐doped regions that yield devices with distinct rectifying current‐voltage behavior (Figure 7f). The resulting planar structure is simple, obviates the need for a top transparent electrode, and exhibits good FPA process compatibility. High‐performance short‐wavelength infrared (SWIR) and MWIR detectors have been fabricated using this approach, notably including 640 × 512 SWIR FPAs achieving a room‐temperature D^* of 3 × 10¹⁰ cm·Hz^1/2·W⁻¹ at a 2.5 µm cutoff wavelength.

To further optimize performance, particularly by suppressing tunneling current and improving carrier collection, Mu et al.^{[ 11 ]} refined the ligand exchange process for HgTe CQDs using HgBr₂ instead of HgCl₂ to achieve lower doping concentration and designed a barrier heterojunction structure (Figure 7g). This architecture employed ZnO as an electron transport layer and P3HT/MoO₃ as a hole transport layer, creating a favorable band alignment that facilitates efficient photogenerated carrier separation and transport while blocking dark current injection. MWIR photodiodes based on this structure, with a 3.7 µm cutoff wavelength, achieved a T _BLIP up to 190 K and a high room‐temperature D^* of 1.26 × 10¹⁰ cm·Hz^1/2·W⁻¹ (Figure 7h). Utilizing this architecture, the authors also successfully fabricated 640 × 512 MWIR FPAs exhibiting a low NETD of 25 mK at 200 K. In contrast, photoconductive HgTe CQD FPAs typically require cooling to around 80 K for optimal performance.^{[ 71} , ⁷² , ⁷³ , ^{74 ]} These advancements underscore the significant potential of CQDs for realizing low‐cost HOT infrared imaging.

CQDs are emerging as a powerful competitor to traditional epitaxial materials, primarily due to their unique fabrication advantages. Their solution‐processability allows for the monolithic integration of detectors onto readout circuits, bypassing the constraints of flip‐chip bonding. This offers a distinct pathway to low‐cost, high‐resolution FPAs. From a performance perspective, CQDs present a complex set of trade‐offs. Their low intrinsic Auger recombination rates and low carrier mobility can, in principle, contribute to a low diffusion current. However, these benefits are often overshadowed by material defects. The high surface‐area‐to‐volume ratio makes them exceptionally sensitive to surface passivation, where defects can introduce excessive G‐R currents and shorten carrier lifetimes. Furthermore, the low mobility creates a direct conflict with quantum efficiency, as it limits the practical absorber thickness for efficient charge collection, leading to lower quantum efficiency compared to bulk materials. Advancing CQD technology toward high‐temperature operation therefore requires a synergistic effort: using ligand engineering to passivate defects and enhance mobility, and band engineering to improve the collection of photocurrents while suppressing dark currents mediated by SRH and TAT processes.

4.2. 2D Materials

2D materials are atomically thin (single or few layers) nanomaterials with large lateral dimensions.^{[ 75 ]} For MWIR/LWIR detection, emerging 2D candidates demonstrate exceptional spectral tunability, including black phosphorus (bP), black arsenic phosphorus (b‐AsP), tellurene, and noble‐metal dichalcogenides (PtSe₂, PdSe₂).^{[ 76} , ^{77 ]} These materials hold significant promise for HOT infrared detectors due to advantageous characteristics such as intrinsically passivated surfaces potentially suppressing surface leakage currents, high carrier mobility, strong optical absorption, and layer‐dependent bandgap tunability.^{[ 21 ]}

The carrier concentration in 2D materials is readily modulated by electric fields, enabling the formation of p‐n homojunctions within a single flake via local electrostatic gating (Figure 8a). Pioneering work by Buscema et al. utilized dual gates with a hexagonal boron nitride dielectric on bP, demonstrating full electrostatic control over its electrical characteristics from metallic to rectifying (Figure 8b), though infrared response was not explored.^{[ 78 ]} Building on this, Yan et al. investigated gated bP photodetectors (Figure 8c),^{[ 79 ]} reporting a D^* of 8.5 × 10⁸ cm·Hz^1/2·W⁻¹ at 77 K (2.2 µm) and importantly showcasing the polarization sensitivity unique to anisotropic 2D materials, unlike conventional bulk semiconductors or CQDs.^{[ 80 ]}

Figure 8.

Photodiodes based on 2D materials. Homojunctions: a) Planar p‐n structure schematic. b) Optical image and output characteristics of a locally gated bP p‐n junction.^{[ 78 ]} Reproduced (Adapted) with permission. Copyright 2014, Springer Nature. c) D^∗ dependence on dual gate voltages for a bP device.^{[ 79 ]} Copyright 2024, Wiley. Heterojunctions: d) p‐n heterostructure schematic. e) D^∗ of a bP/MoS₂ heterojunction photodiode.^{[ 82 ]} Reproduced (Adapted) with permission. Copyright 2018, Springer Nature. f) Carrier transport pathways (vertical versus lateral) in 2D heterostructures.^{[ 83 ]} Reproduced (Adapted) with permission. Copyright 2025, Springer Nature. Barrier Devices: g) Unipolar barrier concept. h) Schematic of a WS₂/h‐BN/PdSe₂ nBn vdW barrier photodetector.^{[ 12 ]} Reproduced (Adapted) with permission. Copyright 2021, Springer Nature. i) Arrhenius plot of dark current for the device indicating high‐operating‐temperature potential.^{[ 12 ]} Reproduced (Adapted) with permission. Copyright 2021, Springer Nature. Interlayer Excitons: j) Device concept utilizing interlayer excitons. k) Schematic and l) D^∗ for a WS₂/HfS₂ device.^{[ 13 ]} Reproduced (Adapted) with permission. Copyright 2020, Springer Nature.

The ability to form van der Waals (vdW) heterostructures by stacking different 2D layers offers a significant advantage, circumventing the stringent lattice‐matching constraints of epitaxy and enabling fabrication through simple exfoliation and transfer methods.^{[ 81 ]} Bullock et al.^{[ 82 ]} utilized this to create bP/MoS₂ vdW heterojunction MWIR photodiodes. Their Type II band alignment (Figure 8d) facilitates efficient charge separation, yielding devices (Figure 8e) with room‐temperature rectification, blackbody response, and a peak D^* of 1.1 × 10¹⁰ cm·Hz^1/2·W⁻¹ (3.8 µm). Performance was further enhanced by Wu et al.^{[ 83 ]} who introduced a vertical graphene (Gr) electrode (Figure 8f). By reducing the vertical transport distance to tens of nanometers, this architecture minimizes recombination losses and boosts external quantum efficiency, resulting in a high room‐temperature D^* of 2.38 × 10¹¹ cm·Hz^1/2·W⁻¹ (3.8 µm). The imaging potential of this platform was demonstrated by an 8 × 1 bP/MoS₂ MWIR line array achieving room‐temperature push‐broom passive imaging.^{[ 14 ]}

Mimicking conventional device designs, unipolar barrier detectors can also be realized using vdW heterostructures (Figure 8g). For instance, Chen et al.^{[ 12 ]} fabricated both nBn (PdSe₂/h‐BN/WS₂, Figure 8h) and pBp (bP/MoS₂/Gr) vdW barrier devices, finding their dark current became diffusion‐limited above 255 K (Figure 8i). While these initial devices showed high room‐temperature detectivity primarily at shorter wavelengths (nBn: 2.7 × 10¹² cm·Hz^1/2·W⁻¹ 520 nm; pBp: 2.5 × 10¹⁰ cm·Hz^1/2·W⁻¹ 2 µm), Zhang et al.^{[ 84 ]} demonstrated the approach in the MWIR range. Their pBp detector using a b‐AsP/MoS₂/bP structure exhibited response up to 4.6 µm with a room‐temperature blackbody D^* of 2.7 × 10¹⁰ cm·Hz^1/2·W⁻¹.

Strong quantum confinement and reduced dielectric screening in 2D semiconductors result in tightly bound excitons.^{[ 85 ]} In Type II heterostructures formed by stacking different 2D materials, photoexcited carriers separate into adjacent layers, creating spatially separated interlayer excitons (Figure 8j). The interlayer excitons' energy, governed by band alignment and typically red‐shifted from intralayer excitons, enables possibilities for tunable infrared detection, potentially into the LWIR regime. Significantly, since interlayer excitons absorption bypasses the need for intrinsically narrow‐bandgap materials, this mechanism offers a promising route to overcome the high thermal dark currents limiting conventional HOT detectors. This potential was demonstrated by Lukman et al.^{[ 13 ]} with a WS₂/HfS₂ heterostructure LWIR photodetector based on interlayer excitons (Figure 8k). This device achieved a remarkable peak D^* of 3 × 10¹⁰ cm·Hz^1/2·W⁻¹ at 7 µm operating at an elevated temperature of 340 K (Figure 8l).^{[ 86} , ^{87 ]}

Notably, the surfaces and interfaces play a critical role in the performance of 2D material‐based devices. These include the 2D material's surface, heterojunction interfaces, and contacts with metal electrodes and underlying substrates. To ensure optimal device operation, these interfaces must be free from airborne impurities and contaminants. Furthermore, optimizing substrate roughness is critical to mitigate the impact of charged impurities on carrier transport and recombination.^{[ 88 ]} A significant challenge arises during the fabrication of electrodes. High temperatures and rapid deposition rates in conventional processes like sputtering or evaporation can introduce thermal stress and physical damage to the underlying 2D semiconductor. This damage often creates atomic disorder and a high density of interface states, which can cause Fermi‐level pinning.^{[ 88 ]} Consequently, this leads to the formation of undesirable, high‐resistance Schottky contacts rather than the required low‐resistance Ohmic contacts. To overcome these issues, it is essential to carefully control the deposition temperature and rate. Alternative strategies include using low‐melting‐point metals for the contacts^{[ 89 ]} or transferring other 2D materials to establish clean, damage‐free vdW contacts.^{[ 90 ]}

Furthermore, studies have shown that vdW interfaces themselves can possess a physical gap, creating a potential barrier that impedes carrier transport.^{[ 91 ]} Addressing this, Liu et al. demonstrated a novel solution by introducing a 0.75 nm intercalated layer of water molecules within a WSe₂/H₂O/PdSe₂ heterostructure. This layer creates a precisely oriented polarization field that depletes electrons in WSe₂, significantly suppressing dark current to sub‐pA levels, while enabling efficient tunneling of photogenerated holes. This innovative interface engineering approach ultimately allowed the device to achieve near‐background‐limited detectivity at room temperature.^{[ 92 ]}

vdW integration moves beyond simple 2D material systems, enabling the combination of materials with different lattice structures and dimensionalities. This approach offers a strategy to synergize complementary properties. Among these combinations, 2D/3D heterojunctions have emerged as particularly promising for developing high‐performance photodiodes suitable for HOT detection.

Zeng et al.^{[ 93 ]} reported a PtTe₂/Si Schottky junction photodetector using graphene as a transparent top electrode (Figure 9a). The work function difference between semi‐metallic PtTe₂ and Si creates a built‐in interfacial electric field (Figure 9b), facilitating rapid photocarrier separation. This device displayed good rectification (Figure 9c) and photovoltaic responses to both MWIR (4.55 µm) and LWIR (10.6 µm) laser illumination, yielding room‐temperature D^* of 2.98 × 10⁹ and 0.93 × 10⁹ cm·Hz^1/2·W⁻¹, respectively.

Figure 9.

Photodiodes based on 2D/3D heterojunctions. PtTe₂/Si schottky device:^{[ 93 ]} a) Schematic illustration, b) Working mechanism, and c) Current‐voltage curves under illumination at specific infrared wavelengths (3.04 µm, 4.55 µm, 10.6 µm). Reproduced (Adapted) with permission. Copyright 2020, Wiley. HgCdTe/Gr Device:^{[ 94 ]} d) Schematic diagram, e) Energy band diagram, and f) Corresponding spectral D ^*. Reproduced (Adapted) with permission. Copyright 2022, Wiley. HgCdTe/bP Device:^{[ 95 ]} g) schematic diagram, h) energy band diagram, and i) corresponding spectral D ^*. Reproduced (Adapted) with permission. Copyright 2022, American Association for the Advancement of Science.

To overcome the limited optical absorption inherent to atomically thin 2D materials, another promising strategy involves combining them with traditional high‐absorption narrow‐bandgap semiconductors like HgCdTe. Wang et al.^{[ 94 ]} pioneered this by constructing vdW heterojunctions on epitaxial HgCdTe (Figure 9d), specifically forming a Gr/HgCdTe Schottky junction. The resulting interfacial built‐in field, combined with HgCdTe's high infrared absorption (Figure 9e), enabled a room‐temperature peak blackbody D^* up to 2 × 10¹⁰ cm·Hz^1/2·W⁻¹ (Figure 9f). In related work, Jiao et al.^{[ 95 ]} reported an HgCdTe/bP vdW heterojunction MWIR photodiode (Figure 9g). Analysis of its band structure (Figure 9h) suggested a p‐i‐n configuration formed by p‐HgCdTe, n‐bP, and an interfacial native oxide (PO_x). Crucially, the high potential barrier presented by this PO_x layer effectively suppressed dark current, leading to a high room‐temperature peak blackbody D^* of 7.93 × 10¹⁰ cm·Hz^1/2·W⁻¹ at a 4.4 µm cutoff wavelength (Figure 9i).

While 2D/3D heterojunctions offer intriguing possibilities, the quality of the interface between the two disparate material systems requires critical attention. The surface of any bulk semiconductor is inherently imperfect, featuring dangling bonds that act as potent generation‐recombination centers, leading to increase dark current. It is well‐established in conventional LWIR photodetector technology that controlling these surface defects through sophisticated passivation techniques is essential for achieving high performance. Therefore, 2D/3D heterojunction platforms face a key challenge: whether the 2D material can directly and effectively passivate the surface dangling bonds of the 3D bulk material. If not, the introduction of an additional passivation layer could adversely affect the band alignment and carrier transport between the 2D and 3D materials, making the design of this interlayer critically important. Systematic investigation into the electronic properties of the 2D/3D interface is essential to determine the ultimate viability and potential of this heterointegration strategy.

Despite promising reports of high room‐temperature detectivity from 2D material heterojunctions in the MWIR and LWIR, significant challenges currently limit their practical viability. A critical issue is the nascent understanding of their fundamental properties. The key parameters such as absorption coefficients and carrier lifetimes, as well as the background‐limited temperature and the underlying dark current mechanisms, remain insufficiently studied. This is compounded by a physical limitation: the atomic thinness of these materials severely restricts photon absorption, creating a major hurdle for achieving high quantum efficiency. Moreover, these impressive results are almost exclusively demonstrated on small‐area, single‐element devices, and a clear pathway to scalable, large‐area FPAs has not yet been established, representing a major technological gap.

4.3. Amorphous or Polycrystalline Thin Films

Amorphous or polycrystalline thin films offer a distinct alternative to single‐crystal materials, defined by their disordered structure lacking long‐range atomic order or comprising randomly oriented nanocrystallites. Their versatile synthesis via diverse deposition methods provides fertile ground for innovation, driven by a key motivation: developing low‐temperature fabrication techniques compatible with complementary metal‐oxide‐semiconductor (CMOS) back‐end‐of‐line (BEOL) processes. This compatibility is a critical pathway toward mass‐producible, low‐cost, high‐performance monolithic infrared FPAs.^{[ 96 ]} Current research in this area predominantly focuses on lead salt compounds like PbS, PbSe, and PbTe for MWIR applications.^{[ 97 ]}

PbSe has garnered significant attention within this group due to several advantageous properties. These include a high absorption coefficient, a relatively low Auger recombination coefficient, and a large exciton Bohr radius (46 nm).^{[ 98 ]} The large Bohr radius implies lower recombination rates, aiding dark current reduction. This favorable combination makes PbSe one of the few infrared materials suitable for high‐performance room‐temperature detection. Despite a long research history, development efforts for PbSe detectors initially focused predominantly on photoconductors.^{[ 99} , ¹⁰⁰ , ¹⁰¹ , ^{102 ]}

The MWIR photodiodes based on CdS/PbSe heterojunctions were reported by Weng et al.^{[ 103 ]} (Figure 10a). These devices intriguingly showed peak response at 200 K (Figure 10b), likely due to temperature‐dependent band alignment effects hindering electron collection at lower temperatures. Their room‐temperature peak D^* was 5.48 × 10⁸ cm·Hz^1/2·W⁻¹ (4.7 µm). Qiu et al. later optimized PbSe processing, identifying optimal annealing conditions (673 K) for minimal dark current (Figure 10c).^{[ 104 ]} In 2023, they demonstrated an improved n‐CdSe/p‐PbSe heterojunction photodiode^{[ 15 ]} (Figure 10d,e) employing vapor physical deposition (VPD) for the PbSe layer. VPD provides advantages over chemical bath deposition, including better uniformity and reproducibility, no sensitization requirement, and enhanced ROIC process compatibility. Supported by developed etching and ZnS passivation techniques, the dark current was found to be limited by G‐R or leakage current. Significantly improved peak detectivities of 7.5 × 10⁹ cm·Hz^1/2·W⁻¹ (298 K) and 3 × 10¹⁰ cm·Hz^1/2·W⁻¹ (220 K) were achieved with these optimized devices (Figure 10f).

Figure 10.

Thin‐film photodiodes based on lead salt heterojunctions. PbSe/CdS heterojunction: a) Device configuration schematic.^{[ 103 ]} Reproduced (Adapted) with permission. Copyright 2014, AIP Publishing. b) Normalized peak photoresponse and cutoff wavelength versus temperature.^{[ 103 ]} Copyright 2014, AIP Publishing. c) Dark current density‐voltage characteristics for devices annealed at various temperatures.^{[ 104 ]} Reproduced (Adapted) with permission. Copyright 2019, Royal Society of Chemistry. PbSe/CdSe Heterojunction:^{[ 15 ]} d) Band alignment at zero bias. e) Device configuration schematic and tilted scanning electron microscope images. f) Blackbody D ^* performance. Reproduced (Adapted) with permission. Copyright 2023, American Chemical Society.

The primary performance hurdle for amorphous and polycrystalline thin films is their intrinsic structural disorder, which creates a high density of bulk defects that act as potent recombination centers and produce substantial dark currents. Overcoming this limitation requires a two‐pronged research effort: enhancing carrier lifetimes through advanced deposition process control, and engineering heterostructures to block dark current pathways. Moreover, expanding research beyond the currently favored lead salt compounds to explore other material systems presents a significant and largely untapped opportunity for future breakthroughs in this area.

5. Summary and Perspectives

Research on HOT MWIR and LWIR photodiodes has advanced significantly through optimizing traditional device architectures and investigating emerging material systems. Here, we synthesize the performance characteristics and core challenges of current technological pathways and offer perspectives on potential future breakthroughs and development directions in the field.

Table 1 provides a comparative summary of key physical properties for various infrared materials, along with a qualitative assessment of their viability for HOT FPA applications. The performance of these materials is dictated not only by their intrinsic properties but also by their practical reliability and potential for scalable manufacturing.

Table 1.

Comparison of conventional and emerging materials for HOT detector development.

Materials	Bandgap [eV]	Carrier lifetime [s]	Mobility [cm2 V−1 s−1]	Absorption coefficient [cm−1]	Auger coefficient [cm6/s]	Reliability	ROIC compatibility	Large‐area uniformity
HgCdTe	0.07–0.75	0.5–10 × 10−3	1–5 × 10 5	2.2–3.2 × 10 3	1 × 10−25	★★	★	★★
InAsSb	0.1–0.36	1–30 × 10−8	3–5 × 10 4	2.5–3 × 10 3	2.1 × 10−26	★★★	★	★★
InAs/GaSb T2SL	0.1–0.25	1–2 × 10−8	2–8 × 10 3	1.6–2.4 × 10 3	1 × 10−27	★★★	★	★★
InAs/InAsSb T2SL	0.1–0.25	5–25 × 10−6	2–8 × 10 3	0.8–1.2 × 10 3	7 × 10−26	★★★	★	★★
HgTe CQD	0.1–0.75	1–10 × 10−9	0.35–8	5000	0.7–40 × 10−29	★★	★★★	★★★
bP	0.3–2	7 × 10−10	1000	1.8 × 10 4	1 × 10−28–1 × 10−24	★	★★★	★
Tellurene	0.3–1	3 × 10−11	700	NA	1.6 × 10−27	★	★★★	★
PtSe2	0–0.72	2.5 × 10−10	210	NA	1.2 × 10−29	★★★	★★★	★
PdSe2	0.03–1.3	2.1 × 10−10	1000	5000	1 × 10−28	★★★	★★★	★
PbSe	0.27	1–6 × 10−6	1000	5000	5 × 10−28	★★★	★★	★★

Data reference: HgCdTe,^{[ 115 ]} InAsSb,^{[ 116} , ¹¹⁷ , ^{118 ]} InAs/GaSb T2SL,^{[ 115} , ^{119 ]} InAs/InAsSb T2SL,^{[ 115} , ^{120 ]} HgTe CQD,^{[ 60} , ¹²¹ , ^{122 ]} bP,^{[ 82} , ¹²³ , ^{124 ]} Tellurene,^{[ 125} , ¹²⁶ , ^{127 ]} PtSe₂,^{[ 93} , ¹²⁸ , ¹²⁹ , ¹³⁰ , ^{131 ]} PdSe₂ ^{[ 132} , ¹³³ , ¹³⁴ , ^{135 ]} and PbSe.^{[ 15} , ¹³⁶ , ^{137 ]} Performance rating is indicated by asterisks (★), with more stars signifying higher performance.

First, the fundamental performance potential is primarily governed by carrier lifetime. HgCdTe sets the industry benchmark with the longest lifetimes, enabling its superior HOT performance. In contrast, III–V materials like InAsSb and Type‐II T2SLs have shorter lifetimes and thus rely on barrier structures to suppress G‐R currents. The challenges for emerging materials are more pronounced. HgTe CQDs are hindered by both low carrier lifetimes and poor mobility, leading to high dark currents and low quantum efficiency. Similarly, 2D materials like bP and tellurene suffer from very short reported lifetimes and fundamentally limited optical absorption due to their atomic‐scale thickness. Polycrystalline PbSe thin films, however, offer a promising compromise with high mobility and absorption coupled with moderate lifetimes, suggesting a viable path toward higher background‐limited temperatures.

Second, practical reliability and stability are paramount for real‐world applications, with thermal and ambient stability being key differentiators. In HgCdTe, elevated temperatures above 100 °C can induce the diffusion of Hg atoms.^{[ 105 ]} For HgTe CQDs, the main thermal degradation pathway is sintering. At temperatures around 130 °C, the desorption of surface ligands can occur, allowing neighboring quantum dots to fuse together and altering the material's properties.^{[ 106 ]} Both of these failure mechanisms make the materials less robust compared to III–V compounds or PbSe. In terms of ambient stability, HgTe CQDs can exhibit long‐term durability, with reports showing their physical properties remaining unchanged after 900 days of storage in an ambient environment.^{[ 107 ]} This stands in stark contrast to certain 2D materials, such as bP and tellurene, which are highly susceptible to oxidation in air, which necessitates encapsulation schemes to ensure long‐term operational reliability.

Finally, the pathway to low‐cost, large‐format FPAs depends on ROIC integration compatibility and large‐area uniformity. Traditional epitaxial materials like HgCdTe and III–V compounds require expensive, lattice‐matched substrates and rely on costly, low‐throughput flip‐chip bonding, which limits scalability. In contrast, emerging materials offer routes to monolithic integration. CQDs can be solution‐processed directly onto large ROIC wafers, with demonstrated 12‐inch compatibility.^{[ 108 ]} 2D materials can be integrated via mechanical transfer, offering good process compatibility, but the synthesis of uniform, high‐quality, large‐area films remains a major technological hurdle.^{[ 109 ]} PbSe thin films can be fabricated using vacuum deposition, enabling highly uniform, large‐area films directly on ROIC substrates.^{[ 110 ]}

In conclusion, traditional materials such as HgCdTe, InAsSb and T2SLs have seen their material properties stabilize after decades of development, but their integration with ROIC relies on complex hybrid bonding techniques. Emerging materials like CQDs, 2D semiconductors and thin‐film materials offer better compatibility with ROIC, but their material properties, especially carrier lifetime, need to be further optimized through improved growth and fabrication processes.

Table 2 provides a comparison of key performance metrics, including D ^*, R, η, J _dark​, and λ _c​, for various reported HOT photodiodes. However, establishing a meaningful cross‐platform comparison using peak D ^* is difficult. The D ^* is highly sensitive to inconsistencies in the measurement of noise, the type of light source (laser vs blackbody sources), and the definition of device area. It is a problem particularly acute for 2D materials. The critical need for rigorous, standardized testing protocols for emerging platforms has been extensively highlighted by reviews such as that from Wang et al., which details characterization methods for the effective area, noise, responsivity, and gain of 2D photodetectors.^{[ 111 ]}

Table 2.

Performance of HOT photodiodes using conventional and emerging materials.

Photodiode	Materials	λ c [µm]	η [%]	R [A·W−1]	J dark [A·cm−2]	D * [cm·Hz1/2·W−1]	Year/refs.
HgCdTe p‐i‐n MW	Epitaxial materials	5.9	80	NA	1 × 10−2 (250 K)	NA	2019/[8]
HgCdTe p‐i‐n LW	Epitaxial materials	10	70	NA	4 × 10−2 (250 K) 8 × 10−3 (200 K) 1 × 10−3 (150 K)	NA	2019/[8]
InAsSb nBn MW	Epitaxial materials	4.5	30	NA	1 (300 K) 5 × 10−2 (246 K) 5 × 10−4 (197 K)	1 × 109 (300 K) 5 × 109 (250 K)	2014/[40]
InAs/InAsSb T2SL nBn MW	Epitaxial materials	5	70	2.3	1 (295 K) 1 × 10−1 (253 K) 4 × 10−3 (208 K)	2 × 109 (295K) 6 × 109 (253 K) 5 × 1010 (208 K)	2018/[138]
InAs/GaSb T2SL CBIRD MW	Epitaxial materials	5	40	1.25	3.2 × 10−3 (200 K) 2.6 × 10−5 (150 K) 1.2 × 10−9 (150 K)	4.4 × 1010 (200 K)	2015/[139]
InAs/GaSb T2SL pBp MW	Epitaxial materials	4.9	67	2.4	0.2 (250 K) 3 × 10−3 (200 K) 1 × 10−5 (150 K)	7.7 × 109 (250 K)	2015/[51]
InAs/InAsSb T2SL nBn LW	Epitaxial materials	10	43	2.65	0.1 (130 K) 1 × 10−3 (90 K) 1 × 10−4 (77 K)	2 × 1010 (130 K)	2017/[140]
InAs/GaSb T2SL ICIP LW	Epitaxial materials	9.5	NA	0.17	2 (250 K) 2 × 10−1 (200 K) 1 × 10−2 (150 K)	1.1 × 108 (300 K) 1.1 × 109 (200 K)	2016/[141]
HgTe Barrier MW	CQDs	4	11	0.35	0.2 (300 K) 6 × 10−4 (180 K) 7.5 × 10−7 (80 K)	1.26 × 1010 (300 K) 1.8 × 1011 (190 K)	2025/[11]
HgTe Planar p‐n MW	CQDs	2.5	25	0.4	1 × 10−4 (300 K)	3 × 1010 (300 K)	2023/[10]
HgTe/Ag2Te MW	CQDs	5	45	1.62	1 × 10−3 (85 K)	4 × 1011 (85 K)	2018/[74]
HgTe/Ag2Te MW	CQDs	4.8	32	0.56	1.1 × 10−3 (160 K) 2.4 × 10−4 (135 K) 2.8 × 10−5 (95 K)	1 × 109 (230 K) 3.3 × 1011 (85 K)	2018/[73]
HgTe /Ag2Te MW	CQDs	3.7	15	NA	1 × 10−4 (80 K)	1 × 109 (300 K)	2023/[142]
HgTe p‐i‐n MW	CQDs	4.4	77	2.7	1.7 × 10−2 (300 K) 2.6 × 10−3 (150 K) 2.5 × 10−4 (80 K)	7.9 × 109 (300 K) 2.7 × 1011 (80 K)	2023/[9]
bP Planar p‐n MW	2D	4.2	NA	1.26 × 10−3	2.6 × 10−4 (77 K)	2 × 107 (295 K) 8.5 × 108 (77 K)	2024/[79]
Gr/bP/MoS2/Gr MW	2D	4.2	23	0.66	8.8 × 10−2 (300 K)	2.38 × 1011 (300 K)	2025/[83]
bP/MoS2 MW	2D	4.2	35	0.9	6.3 × 10−2 (300 K)	1.1 × 1010 (300 K)	2018/[82]
bP/MoS2/Gr pBp MW	2D	4.2	NA	NA	3.2 × 10−1 (300 K)	2.5 × 1010 (300 K)	2021/[12]
B‐AsP/MoS2/bP pBp MW	2D	4.6	NA	2.34 × 10−2	1.5 × 10−2 (300 K)	2.7 × 1010 (300 K)	2024/[84]
PdSe2/MoS2 LW	2D	12	10	5	1.27 × 10−4 (300 K)	8.2 × 109 (300 K)	2019/[143]
PdSe2/MoS2 LW	2D	12	NA	0.47	4.6 × 10−6 (300 K)	1.53 × 109 (300 K)	2025/[135]
PdSe2/bP LW	2D	12	30	2.2	5 × 10−1 (300 K)	2.5 × 109 (300 K)	2022/[144]
PtSe2/MoTe2 LW	2D	12	1.4	0.12	1.15 × 10−1 (300 K)	6.3 × 107 (300 K)	2025/[145]
PtSe2/Te MW	2D	4.2	NA	196.8	4.42 (300 K)	4.3 × 109 (300 K)	2024/[146]
WS2/HfS2 Exciton LW	2D	12	NA	>1 × 10 3	1 × 10−3 (300 K)	3 × 1010 (340 K)	2020/[13]
Gr/PdTe2/Si LW	2D/3D	12	NA	1.96 × 10−3	1.75 × 10−5 (300 K)	3.9 × 109 (300 K)	2023/[147]
PtSe2/AlOx/Ge LW	2D/3D	12	NA	1 × 10−3	2 × 10−3 (300 K)	2 × 109 (300 K)	2024/[148]
HgCdTe/bP MW	2D/3D	4.2	NA	0.168	4.8 (300 K)	1.8 × 1010 (300 K)	2022/[95]
HgCdTe/Gr MW	2D/3D	4.3	NA	3	37 (300 K)	2 × 1010 (300 K)	2022/[94]
PbSe/CdS MW	Thin films	4.7	20	0.055	1 × 10−4 (300 K)	5.5 × 108 (300 K)	2014/[103]
PbSe/CdS MW	Thin films	4.2	NA	0.36	3 × 10−7 (300 K)	8.5 × 108 (300 K)	2019/[104]
PbSe/CdSe MW	Thin films	4.7	18	40	2 (300 K)	7.5 × 109 (300 K) 3 × 1010 (220 K)	2023/[15]

Note: The principle for comparing device performance is to ensure consistency and avoid underestimation. The J _dark values are standardized to those measured under a small bias of −0.1 V. For devices where J _dark​ was not explicitly provided (particularly 2D photodetectors), the effective area was estimated from the published device images. “NA” denotes values not reported in the original literature. The η for some devices was converted from the reported responsivity via Equation (9), assuming a β of 1. If the calculated η exceeded 100%, indicating the presence of internal gain, the value was capped at 100% for this comparison to assess the intrinsic photoconversion efficiency. For 2D semiconductors where the cutoff wavelength has not been experimentally determined, a value of 4.2 µm is assigned to bP. Correspondingly, a cutoff wavelength of 12 µm is assigned to the LWIR 2D materials PdTe₂, PdSe₂, and PtSe₂.

For HOT photodetectors, T _BLIP is a crucial benchmark for evaluating their high‐temperature performance. The T _BLIP of a device can be determined by measuring its dark current as a function of temperature while exposed to a 300 K blackbody source within a 2π field of view. As the device is cooled, its internal dark current is suppressed to below the constant photocurrent generated by the background radiation. Experimentally, the T _BLIP is observed as the inflection point where the total current stops decreasing and plateaus, assuming the photocurrent remains constant with temperature.

However, the majority of publications on HOT detectors do not report the T _BLIP. This makes it exceedingly difficult to fairly compare the HOT potential across different material platforms. To address this gap, we employ a more broadly applicable metric: the ratio of dark current density to quantum efficiency (J _dark​/η) as a key metric. This ratio is a powerful indicator because the background‐limited performance requires J_dark < qQ_B η, meaning lower values of this ratio directly indicate superior HOT potential. Accordingly, Figure 11 compiles the reported J _dark​/η values for diverse materials, benchmarking them against the established Rule 07 and Law 19 limits to provide a more equitable assessment of performance.

Figure 11.

Comparison of reported HOT photodiodes performance across traditional and emerging materials, based on data from Table 2.

State‐of‐the‐art performance, reaching the Law 19 background limit, remains dominated by HgCdTe with fully‐depleted p‐i‐n structures. Other traditional HOT detectors, like barrier‐based InAsSb or T2SLs, generally operate near the Rule 07. MWIR HgTe CQD detectors demonstrate promising performance approaching Rule 07 at higher operating temperatures. However, their relatively higher dark current at lower temperatures, coupled with a shallow temperature‐dependence slope illustrated by the green line, suggests limitations imposed by G‐R current, indicating a need for further reduction in material or interface defects. Similarly, some heterojunction devices based on 2D materials and thin films have shown encouraging results, with reported performance near or even surpassing the Rule 07, suggesting significant potential for HOT applications. However, interpretive caution is crucial for these emerging results, particularly for 2D materials and thin films. Some reported data for these materials falls below the fundamental Law 19 limit. This discrepancy is likely an artifact stemming from an overestimation of the cutoff wavelength or quantum efficiency, which can be attributed the lack of complete absorption and responsivity spectra, the omission of standard blackbody response measurements,^{[ 76 ]} the presence of high internal gain^{[ 19 ]} or inaccurate device area definition.^{[ 111 ]} This underscores the vital need for standardized testing protocols to enable reliable assessment and comparison of these novel platforms relative to established technologies.

HOT infrared photodiodes have advanced significantly through synergistic improvements in both materials and device architectures. Within conventional platforms, HgCdTe p‐i‐n structures with ultra‐low doping have reached background‐limited performance, though scaling these to large, uniform FPAs requires further stabilization of epitaxial growth processes. For barrier detectors based on InAsSb and T2SLs, achieving performance substantially beyond the Rule 07 benchmark necessitates continued refinement of fabrication techniques to enhance carrier lifetimes.

Emerging material systems offer compelling potential to surpass current operating temperature limits while simultaneously enabling cost reductions and broader application reach. Nevertheless, unlocking the full capabilities of these novel materials necessitates focused research and key technological breakthroughs in the following areas:

For CQD photodiodes, future advancements require focused efforts on defect passivation and interface engineering. Optimizing ligand strategies is key to increasing carrier lifetimes and mobility, thereby suppressing the dark current and enhancing quantum efficiency. Concurrently, enhancing carrier collection efficiency demands precise control over interfacial energy level alignment and charge transfer dynamics between the CQD active layer and adjacent charge transport layers.

For 2D material photodiodes, future progress depends on addressing three key areas. First, enhancing light absorption in these atomically thin layers is critical, potentially using mixed‐dimensional heterojunctions, cascade, and plasmonic architectures to improve quantum efficiency. Second, exploring unique 2D physics, such as interlayer exciton absorption, offers pathways to novel detection mechanisms potentially circumventing dark current limitations. Finally, translating laboratory results to practical devices requires overcoming major materials challenges: achieving wafer‐scale, high‐quality synthesis and transfer, alongside developing stable encapsulation, remains essential for viable 2D material FPAs.

For thin film photodiodes, key future directions include expanding the material palette, innovating device structures, and ensuring CMOS compatibility. A critical step is to explore alternatives to conventional lead salts, focusing on promising polycrystalline, amorphous, and organic/hybrid thin‐film semiconductors. The ideal candidates for infrared applications should possess not only suitable bandgaps but also a combination of high absorption, high carrier mobility, and long carrier lifetimes. The use of AI‐driven material discovery is poised to significantly accelerate the identification of such materials, paving the way for the rapid development of HOT thin‐film photodiodes. Simultaneously, adapting established device engineering principles, such as barrier structures and heterojunction designs, will be vital for suppressing dark current in these material systems. Critically, continuous optimization of fabrication processes for compatibility with CMOS BEOL technologies is essential to enable the ultimate goal of low‐cost, large‐scale monolithic FPA integration.

1D nanowires and nanorods, similar to CQDs, leverage quantum confinement effects for tunable optoelectronic properties. However, they offer a distinct advantage in charge transport: their high aspect ratio provides a direct axial pathway for rapid and efficient carrier collection at the electrodes. This significantly minimizes the probability of carriers being captured by material defects or lost to recombination during transit. Furthermore, the inherent geometric anisotropy of nanowires endows them with unique polarization sensitivity, giving rise to a series of optical phenomena advantageous for photodetection. While several studies have investigated III–V semiconductor nanowires for use as MWIR/LWIR photoconductors, research into photodiode architectures is comparatively scarce.^{[ 112} , ¹¹³ , ^{114 ]} Consequently, the challenge of how to construct photodiode structures from nanowires that are compatible with FPAs remains a key area for future exploration.

In summary, HOT photodetector technology is currently undergoing a period of rapid advancement. This progress is primarily attributed to improvements in material quality, and the implementation of innovative device architectures. Consequently, these developments have enabled a systematic increase in the T _BLIP. This technological trajectory is poised to significantly alter the cost structure and application scope of conventional infrared systems. The future development direction will be to continue increasing the operating temperature, with the ultimate goal of developing high‐performance photon detectors that operate at room temperature without any active cooling.

Furthermore, the advancement of HOT technology provides a foundation for the exploration of novel sensing paradigms. These include infrared sensing technologies based on flexible or conformal substrates, as well as neuromorphic sensing systems inspired by biological vision. These emerging technologies are anticipated to have a substantial impact on diverse fields such as defense and security, industrial process control, consumer applications, and scientific research.

Conflict of Interest

The authors declare no conflict of interest.

Biographies

Yunxiang Di is a postdoctoral fellow at the college of integrated circuits & micro‐nano electronics, Fudan University. He earned his Ph.D. in Integrated Circuit Science and Engineering from Fudan University in 2024. His research interests center on the development of novel infrared photodetectors and optoelectronic devices.

Jianlu Wang is a Professor and Ph.D. supervisor at Fudan University. He earned his Ph.D. from the Shanghai Institute of Technical Physics, Chinese Academy of Sciences, in 2010, and served as Assistant, Associate, and Full Researcher there until 2021. His research interests include novel optoelectronic materials and infrared photodetectors.