X-rays penetrating live organisms in small doses can facilitate medical examination. To reduce the radiation dose applied to human bodies for medical imaging and applications, X-ray detectors should have high sensitivity. Hence, the sensitive detection of X-rays has emerged as an essential research area for minimizing radiation doses. There is a need for cost-effective, high-resolution detectors that operate at low-photon fluxes [1]. Recently, the underlying correlations between the properties of halide perovskites and their X-ray detection performance were intensively studied; at present, most of the commercial X-ray detectors are based on inorganic semiconductors such as Si, CdTe, and Ge, which require complex and high-cost fabrication processes [2].

Metal halide perovskites (MHPs) generally possess low trap densities, high charge-carrier mobilities, long minority-carrier diffusion lengths, and high photoluminescence quantum yields (PLQYs). Moreover, because they make possible solution processability for cost-effective synthesis [3], MHPs have garnered much interest recently in the fields of photovoltaic and optoelectronic applications and devices, including solar cells, light-emitting diodes, photodetectors, and irradiation detectors.

Currently, two strategies are available for X-ray detection: direct conversion of X-ray photons into electrical signals in detectors and indirect conversion of X-ray photons into low-energy photons in scintillators. The remarkable properties of perovskite materials make them a promising candidate material for X-ray detectors and scintillators.

A linear increase in photocurrent with the incident X-ray dose was reported in 2018 [4]. An MAPBI₃-based X-ray detector was reported to be 100 times more sensitive than a Si-based X-ray detector in 2020 [5]. In 2021, Glushkova et al. [1] reported that the X-ray dose required to generate an image can be reduced by more than 1,000 times in a perovskite-based X-ray detector designed on graphene. MAPbX₃ (X=Br or I) and CsPbBr₃ are the most studied materials for these detectors, exhibiting large mobility–lifetime (μτ) products (~10⁻² cm² V⁻¹), high resistivity (~109 Ω cm), low detection limits (<100 nGy s^–1), and excellent sensitivity (>10,000 μC Gy⁻¹ cm⁻²) [1,2,6-8].

As scintillators, halide perovskites exhibit high PLQYs, tunable optical bandgaps, and short decay times, leading to excellent scintillation properties, including high light yield (LY), tunable radioluminescence wavelength, and fast response time (nanoseconds). At a low temperature of 10 K, the LY of the MAPbI₃ single crystal can reach ~300,000 photons/MeV [9].

The research of low-dimensional two-dimensional (2D) MHP is in the early stages. Currently, MHP scintillators are still restricted by their small Stokes shift. Low-dimensional perovskite derivatives with self-trapped excitons can increase the Stokes shift, but they have very long luminescence lifetimes [10]. In addition, improvement of the signal-to-noise ratio (SNR) is an unsolved issue.

Meanwhile, in 2019, Sofer et al. [11] reported a quantum-enhanced X-ray detection method based on spontaneous parametric down conversion (SPDC) with practical potential to improve the SNR by tens to thousands of times.

In this review, we present an overview of the advancements in halide perovskite materials with regard to their use for X-ray detection in the recent years pointing out the remaining challenges and our perspective for future research directions toward perovskite-based X-ray applications. Finally, we compare the conventional detection methods with quantum-enhanced detection and discuss the future research directions toward perovskite-based X-ray applications employing SPDC.

Sensitive detection of X-rays is crucial in medical diagnostic imaging. High-dose X-rays increase the risk of cancer in exposed patients. Hence, an ideal X-ray image detector should be able to provide high-quality images at low X-ray doses.

X-ray detectors can be classified as direct- and indirect-type devices. In a direct-type detector, the detecting materials are ionized by high-energy radiation to generate charge carriers, which then are collected to form electric signals in the detector. In an indirect-type detector, the scintillator converts X-rays into low-energy photons that photodetectors can detect during imaging.

Here, we describe the detection mechanisms of both types of X-ray detectors. In addition, we summarize the key elements that determine the performance of both types.

1. Direct X-ray detection

Depending on the operating principle, direct-type X-ray detectors may be classified into pulse-mode and current-mode devices. Pulse-mode (Fig. 1a) detectors are suitable for low-flux or photon-deficient applications. It takes long to analyze and form the acquired electric signals. The output of the pulse-mode detector comprises voltage pulses corresponding to each detected particle. Current-mode (Fig. 1b) detectors receive high-flux, high-energy photons, and many excess free charges are generated. Then, a current signal is generated when the opposite electrodes collect the charges under a particular bias. Current-mode detectors can realize high spatial resolution and are widely used in radiography imaging and computed tomography [3].

Figure 1. Operation principles of direct conversion X-ray detectors. (a) A pulse-mode detector and (b) a current-mode detector. Reused from the article of Zhou et al. (ACS Energy Lett 2021;6:739−768) [3] with original copyright holder’s permission.

In a direct X-ray detector, an X-ray photoelectric material layer (i-layer) absorbs incident X-ray photons to produce electrons and holes and then sends them to the n-layer and p-layer, respectively.

The collected charges need to be stored and processed into signals, where a thin-film transistor (TFT) or complementary metal-oxide-semiconductor (CMOS) array is required. The TFT-based flat panel detector is widely used in X-ray imaging techniques, such as digital radiography and mammography [10].

An ideal X-ray direct conversion detector should have the following properties: (i) high sensitivity to obtain high-quality X-ray images at a given X-ray dose rate, (ii) low detection limit to reduce the X-ray dose exposure, and (iii) long-term stability under high voltage. The active materials used for direct conversion should have the following properties: a sufficiently large μτ product, minimum trap density of the semiconductor, and a large resistivity to maintain stability [3].

The detection limit is used to quantify the minimum signal identified by the X-ray detector and is defined as the equivalent dose rate that produces a signal thrice the noise level. The International Union of Pure and Applied Chemistry (IUPAC) defines the detection limit as the equivalent dose rate required to produce a signal greater than thrice the noise level (SNR=3.53) The noise current is obtained by calculating the standard deviation of the photocurrent [10]. The detection limit is also an essential parameter in an X-ray detector, which decides the lowest detectable dose. The key to obtaining a relatively low detection limit is to achieve a high current signal at a low noise level.

The termμ represents carrier mobility, which is the drift speed of the carrier under 1 V cm⁻¹. Further, τ is the lifetime of the charge carrier. Theμτ product represents the distance that the carrier can drift per electric field unit. Thus, theμτ product directly determines the charge collection efficiency at a given electric field, which is an essential parameter that determines the X-ray sensitivity [10]. Increasing theμτ product or the applied bias can remarkably improve the device sensitivity. However, the large applied bias may induce a high noise level (e.g., dark current), leading to a poor detection limit. The dark current, or called leakage current, is an important parameter for direct detection.

2. Indirect X-ray detection

The scintillator converts X-rays into visible light. Subsequently, using a light detector such as TFT, charge coupled device (CCD), and CMOS sensors, the signal is efficiently converting to a digital image. An ideal scintillator should possess properties such as a high LY, fast scintillation response, high stability, etc. The scintillation process has three main stages (i.e., conversion, transport, and luminescence), as shown in Fig. 2.

Figure 2. Schematic illustration of the scintillation process in inorganic scintillators. Reused from the article of Zhou et al. (ACS Energy Lett 2021;6:739−768) [3] with original copyright holder’s permission.

X-ray luminescence is generated in the luminescence stage mainly by the trapping and radiative recombination of electron−hole pairs, and light is emitted in the ultraviolet (UV)/visible regions. [3]. Currently, the most successful design of radiation image detectors combines a layer of X-ray fluorescent material with the a-Si:H AM readout array. The fluorescent light is converted to an electronic signal in an array of a-Si:H photodiodes [12].

The LY determines the sensitivity and detection limit of the detector. It is defined as the number of emitted photons per 1 MeV of energy absorbed by the scintillator when exposed to ionizing radiation, and it is expressed in photons/MeV units.

A short response time is also essential to achieve dynamic real-time X-ray imaging for specific medical diagnostics such as compu ted tomography [3]. Therefore, scintillators based on indirect X-ray detectors that provide the advantages of fast response time and convenient integration with TFT or CMOS arrays are the mainstream products.

In recent years, halide perovskites have become candidate materials for X-ray detection and imaging because of their low cost, simple solution process, and desirable characteristics such as large absorption coefficient, high PLQYs, and tunable bandgap. In this section, we provide an overview of the material-selection criteria for X-ray detection by comparing the advantages of different types of halide perovskites with those of traditional semiconductors. We also provide a review of the recent advancements in the use of halide perovskites as active layers in direct conversion detectors and as scintillators for X-ray detection.

1. Advantages of halide perovskites

Hybrid organic–inorganic MHPs represent a class of materials with the stoichiometry ABX₃, whereas A denotes an organic cation (e.g., MA⁺=methylammonium and FA⁺=formamidinium) or an inorganic cation (e.g., Cs⁺ and Rb⁺) with charge +1; B is a metal cation (e.g., Pb²⁺ and Sn²⁺) with charge +2; and X is a halide anion (I⁻, Br⁻, Cl⁻) with charge –1 [13]. The crystallographic structure is shown in Fig. 3 [14].

Figure 3. Left: ABX₃ perovskite structure. Right: The same perovskite structure seen from different angle. Left: Reused from the article of Chen et al. (RSC Adv 2018;8:10489-10508) [14] with permission from the Royal Society of Chemistry.

Recently, MHPs garnered extensive research attention because of their attractive photoluminescence (PL) and electroluminescence properties, including wider absorption coefficient and high PLQYs. Remarkably, the broadband emissions of MHPs present a fascinating prospect for application in optoelectronic devices [15].

The bandgap of halide perovskites can be easily tuned by mixing them with halide ions or cations; this makes it possible for their luminescence spectrum to cover the whole visible light region. Thus, CsPbX₃ (X=Cl⁻, Br⁻, I⁻, or mixed halides) nanocrystals exhibit color-tunable and narrow emissions under X-ray beam excitation, as shown in Fig. 4.

Figure 4. Commission Internationale de I’Eclairage (CIE) chromaticity coordinates of the XEL measured for samples 1−12, which are (1) CsPbCl₃, (2) CsPbCl₂Br, (3) CsPbCl_1.5Br_1.5, (4) CsPbClBr₂, (5) CsPbCl_2.5Br_0.5, (6) CsPbBr₃, (7) CsPbBr₂I, (8) CsPbBr_1.8I_1.2, (9) CsPbBr_1.5I_1.5, (10) CsPbBr_1.2I_1.8, (11) CsPbBrI₂, and (12) CsPbI₃. Reused from the article of Zhou et al. (ACS Energy Lett 2021;6:739−768) [3] with original copyright holder’s permission.

Chen et al. [16] reported multicolor, high-efficiency X-ray scintillators fabricated from a series of CsPbX₃ nanocrystals. For cubic nanocrystals, with an average size of 9.6 nm, subjected to X-ray beam excitation, the perovskite quantum dots (QDs) yield narrow and color-tunable emissions. This unique property makes it possible to realize multicolor, high-efficiency X-ray scintillation.

Typically, the response of the CsPbBr₃ nanocrystal film exhibits good linearity within a broad X-ray dose rate range and has a low detection limit of 13 nGy s⁻¹, which is about 400-fold lower than that required for regular medical diagnostics (5.5 μGy s⁻¹). Meanwhile, this scintillator has a fast response (44.6 ns), which is ideal for X-ray imaging. In addition, halide perovskites with a low band gap (1.6–3.1 eV) are expected to yield up to 129,000–250,000 photons/MeV as the LY of the scintillator is negatively correlated to the optical bandgap. CsPbX₃ nanocrystals have achieved near-unity PLQYs in the violet to the near-IR spectral range [3].

The X-ray absorption coefficient is correlated positively with the atomic number Z of constituents of the scintillation materials. For example, the most studied MAPbBr₃ and MAPbI₃ halide perovskites have linear absorption coefficients of 19.41 and 40.61 cm⁻¹, respectively, for 50 keV X-rays; these values are much higher than those of Si (1.02 cm⁻¹) and α-Se (3.86 cm⁻¹) [3]. The existence of high-Z ions provides excellent X-ray absorption capacity for halide perovskites.

Halide perovskites are semiconductors with high defect tolerance, large mobility, and long carrier recombination lifetimes. The MAPbI₃ polycrystalline film has a μτ product of 2×10⁻⁷ cm² V⁻¹, which is comparable to that of α-Se (~10⁻⁷ cm² V⁻¹). In the MAPbI₃ single crystal, the charge carrier diffusion length reaches 175 μm, resulting in a μτ product of 1.2×10⁻² cm² V⁻¹. Single crystals have a much lower defect density than polycrystalline films because of the high crystallinity and absence of grain boundaries.

Operation of detectors under a high electric field can improve the X-ray detection sensitivity. However, the high voltage induces a large dark current. To suppress the dark current, semiconductors should have high resistivity. The dark current can be reduced by using a high-quality single crystal with a low defect density. In addition, the passivation of grain boundaries can remarkably weaken ion migration. Dark current drift is an issue for MHP X-ray detectors. The dark current drift of the MHP-based device is caused by ion migration. The total conductivity of perovskite comprises electronic conductivity and ionic conductivity.

Methods of growing crystals can be classified into vapor growth, solution growth, and solid-state growth. Among them, the most prevalent growth method is solution-phase growth. Thus, solution-processable, low-cost halide perovskites have considerable advantages for X-ray detection.

2. Halide perovskite-based direct X-ray detectors

Over the past few years, direct X-ray detectors based on various perovskite materials were developed, and impressive progress was made in terms of achieving high sensitivity and low detection limits. Such an all-solution-based perovskite detector can enable low-dose X-ray imaging. This section highlights some advanced and meaningful work in halide perovskite-based (direct-type) X-ray detectors.

In 2017, Kim et al. [15] reported a printable organometallic perovskite X-ray detector with high sensitivities of up to 11,000 μC Gy⁻¹ cm⁻². The MAPbI₃-based X-ray detector was built with TFT pixels. The bottom layer of the PI-MAPbI₃ composite forms a hole-transporting layer, while the top PI-MAPbBr₃ layer forms a hole-blocking layer (HBL). The polycrystalline MAPbI₃ photoconductor layer was printed on the coated PI-MAPbI₃. The top electrode (indium tin oxide) was biased to a positive voltage source to operate the detector in the hole-collection mode [17].

In 2018, Gill et al. [4] reported on organolead halide perovskite-based detectors. The photocurrent increased linearly with the X-ray irradiation dose in perovskite-based detectors. These detectors were about 550% more sensitive than the commonly used amorphous Si solar devices.

In 2020, a detector employing a perovskite photodiode was developed. It benefitted from being equipped with CsPbBr₃ NCs, and a record high sensitivity of 54,684 μC Gy⁻¹ cm⁻² was achieved at a dose rate of 8.8 μGy s⁻¹ [18].

Zhang et al. [6] also reported that X-ray detectors made of perovskite CsPbBr₃ single crystals exhibited a high sensitivity of 1,256 μC Gy⁻¹ cm⁻² for 80 kVp X-ray detection under a relatively low electric field of 20 V mm⁻¹; this sensitivity was 60 times that of the commercially used a-Se detectors.

In 2021, Geng et al. [8] reported a process to synthesize single-crystal MAPbI₃ perovskite. They could boost the sensitivity of the single-crystal perovskite X-ray detector to 1,471.7 µC Gy_air⁻¹ cm⁻² at a low electric field of 3.3 V/mm under 39 keV X-ray radiation.

The introduction of MAPbI₃-based detection into the field of medical imaging has remarkably reduced the health hazards related to the strongly ionizing X-ray photons. Glushkova et al. [1] demonstrated X-ray detector units with a record sensitivity of 2.2×10⁸ μC Gy_air⁻¹ cm⁻² when detecting 8 keV photons at dose rates less than 1 μGy/s.

We present the direct cell constituent materials and reported efficiencies in Table 1 [1,2,6-8,13,17,19,20] and Fig. 5. From Fig. 5, we see that the perovskite X-ray detectors have much higher sensitivity than the conventional X-ray detectors.

Table 1 . Performance of the conventional X-ray detectors for medical imaging.

Material	Applied electric field (V μm−1)	Sensitivity (uC Gy−1 cm−2)	Ref.
MAPbI3	0.0033	1,471.7	HP [8]
MAPb3	8	2.2×108	HP [1]
MAPb(I0.9Cl0.1)3	0.226	14,400	HP [2]
CsPbBr3	0.020	1,256	HP [6]
CsPbI3	0.0045	2,370	HP [7]
Si	0.5	8	[13]
a-Se	10	20	[13]
HgI2	10	1,600	[13]
CZT	0.1–1	318	[13]
MAPbBr3	0.05	2.1×104	[19,20]
MAPbI3	0.24	1.1×104	[13]
Cs2AgBiBr6	0.025	105	[17]

Figure 5. Sensitivity vs. X-ray voltage per µm when using the direct perovskite X-ray structure. The open squares indicate halide perovskites.

3. Halide perovskite-based scintillators

Scintillators convert X-ray photons into visible photons that are then detected by a photodiode array. A scintillator should possess a high LY, fast scintillation response, high stability, lack of self-absorption, etc. to achieved improved performance. The following research directions can be promising: (i) avoiding self-absorption effectively, (ii) increasing exciton binding energy to suppress thermal quenching, and (iii) achieving a short PL lifetime to acquire a fast scintillation response [3]. Unlike the conventional scintillator materials, the halide perovskites for X-ray imaging show attractive merits such as easy fabrication, fast response, and good spatial resolution [21]. Herein, we mention some advanced and meaningful work in halide perovskite-based X-ray scintillators. The key scintillation parameters (e.g., LY, maximum emission, and response time) are also summarized.

In 2015, Protesescu et al. [22] synthesized high-quality CsPbX₃ (X=I⁻, Br⁻, or Cl⁻) nanocrystals with PLQYs reaching 90%. The CsPbBr₃ single crystal is a very bright scintillator, with a LY of 50,000±10,000 photons/MeV and a rapid response of ~1 ns at low temperatures (7 K) [3,22]. In 2018, Chen et al. [16] reported multicolor, high-efficiency X-ray scintillators fabricated from a series of CsPbX₃ nanocrystals. Typically, the response of the CsPbBr₃ nanocrystal film has good linearity within a broad X-ray dose rate range, and the film has a low detection limit of 13 nGy s⁻¹, which is about 420-fold lower than that required for regular medical diagnostics (5,500 nGy s^–1). However, because of the high resolution and low detection limit of the CsPbBr₃ nanocrystal film scintillator, high-quality X-ray images were obtained at a low X-ray dose of 15 μGy.

The emission wavelength of the scintillator should be optimized to maximize the LY. The a-Si-based p-i-n photodiode generally reaches its highest responsivity within the wavelengths of 500 to 600 nm. The ideal condition for X-ray detection is when the scintillator’s emission wavelength matches the photodiode’s responsivity peak.

The emission wavelength of halide perovskites can be adjusted continually by tuning the B/X sites. Chen et al. [16] modified the X site of CsPbX₃ (X=Cl, Br, and I) nanocrystals and fabricated different scintillators with emission wavelengths, ranging from the UV to red wavelengths [10]. Because of the self-absorption effect, the created photons can be reabsorbed by the scintillator, thus decreasing the LY. Furthermore, defects in the scintillator cause nonradiative recombination, further decreasing the LY [10]. Recent reports of radiation detection by organometallic perovskites (MAPbX₃, where MA=CH₃NH₃ and X=Cl, Br, or I) suggest that such materials may enable low-dose X-ray imaging [17]. Table 2 summarizes the LY of halide perovskites and conventional scintillators.

Table 2 . Light yield (LY).

Material	Photons/MeV	Refs.
Cs2Ag0.6Na0.4In0.85Bi0.15Cl6	39,000	HP [21]
CsPbBr3	50,000 (7K)	HP [3]
MAPbBr3	90,000 (77K)	HP [3]
CsPbBr3	21,000	HP [23]
Mg2+LuAG:Ce	21,900	[24]
LYSO:Ce	32,000	[25]
CsI:TI	56,000	[25]

In Table 2 [3,21,23-25], we summarize the LY of some halide perovskites and conventional scintillators. Compared with other scintillators used for X-ray detection, halide perovskites have a high theoretical LY of approximately 200,000 photons/MeV because of their defect tolerance and relatively small band gap. However, the actual LY will be reduced because of self-absorption, nonradiative recombination, and non-unity light extraction efficiency.

The X-ray LY of single crystals is as high as ~150,000 photons/MeV at 10 K, but it decreases drastically to less than 1,000 photons/MeV at room temperature. This severe temperature dependence is referred to as the thermal-quenching behavior [3]. The different band gaps and exciton binding energies may cause differences in the quenching behavior.

Scintillators based on indirect X-ray detectors have the advantages of fast response time and convenient integration with TFT or CMOS arrays. The MAPbBr₃ crystal showed an intense sub-nanosecond scintillation response at low temperatures (50–130 K); this response is much faster than that of many conventional scintillators. For MAPbX₃ (X=I⁻, Br⁻, or Cl⁻) single crystals, excellent scintillation performance, i.e., fast response and high LY, were achieved at low temperatures (<100 K), but the performance is limited at room temperature [3]. Previous research showed that the PLQY of the halide perovskite nanocrystal reaches nearly 100% and they exhibit fast-luminescent decay.

As scintillators, halide perovskites exhibit high PLQYs, tunable optical bandgaps, and short decay times, resulting in excellent scintillation properties, including high LY, tunable radio-luminescence wavelength, and nanosecond response time. MHPs realize tunable bandgaps by adjusting the B/X sites to match the photodetector response wavelength [10]. Halide perovskites with high radioluminescence covering all visible wavelengths have been demonstrated. The halide perovskites possess a LY as high as 64,000 photons/MeV at room temperature. As the temperature drops to 10 K, the LY of the MAPbI₃ single crystal can reach ~300,000 photons/MeV [3].

The high stability of layered 2D halide perovskites is attributed to their special structure, which comprises layers of the inorganic semiconductor that are sandwiched between insulating bulky organic layers. The transition from 3D to 2D structures results in strong quantum confinement effects. The quantum wells of 2D perovskites confine the electrons and holes generated from ionizing radiation and enlarge the overlap of the electron and hole wave functions. This, in turn, increases the excitonic oscillator strength and shortens the excitonic radiative lifetime in 2D perovskites.

The 2D layered perovskites play an essential role in perovskite photovoltaics and optoelectronics. In particular, the possibility of combining high performance and stability for perovskite-based X-ray detection has opened up a new avenue of research [26,27].

Herein, we discuss the structure of 2D layered perovskites and their unique properties. Then, we discuss the progress of 2D layered perovskites for X-ray detection. Finally, we summarize and outline the perspectives related to high-performance 2D layered perovskite X-ray detectors.

1. 2D perovskite: structure and properties

The 3D halide perovskite has a general chemical formula of ABX₃. A is a small organic cation or an inorganic cation; B is a metal cation, and X is a halide anion. Reducing the structural dimensions from 3D to 2D perovskites results in a general chemical formula A′₂A_n–1BnX_3n+1 or A′A_n–1BnX_3n+1 (A′=1⁺ or 2⁺ cation, A=MA⁺, FA⁺, or Cs⁺). Further structural reduction of the 2D perovskite may lead the formation of 1D and 0D halide perovskites [3]. Reducing the perovskite dimension by incorporating large organic molecules (i.e., forming a 2D structure) or fabricating QDs/nanocrystals can increase its stability and performance.

Reducing the dimension tends to increase the activation energy against the uncontrollable migration of carriers in halide perovskite, thereby improving the performance and lifetime of the device. The 2D perovskite is introduced with large organic molecules to reduce the 3D structure, thereby increasing the bulk resistance to avoid noise and further improve the sensitivity of the detector.

A strong quantum confinement effect indicates a large exciton binding energy. A large exciton binding energy effectively eliminates the thermal-quenching effect, and hence, the 2D halide perovskite single crystals exhibit a high LY of more than 10,000 photons/MeV at room temperature.

According to the characteristics of octahedral connectivity, as shown in Fig. 6, perovskite-based structures can be classified into different dimensionalities, such as 3D, 2D,nd 0D at the molecular level. Fig. 6. presents the schematic representations showing the connectivity of BX₆ octahedra in different dimensionalities (3D, 2D,nd 0D) at molecular levels. Broadband emissions were observed in corrugated-2D,nd 0D perovskites as a result of efficient exciton self-trapping [17].

Figure 6. Schematic representations showing the connectivity of BX6 octahedra in different dimensionalities (3D, 2D, 1D, and 0D) at the molecular levels. D, dimensional. Reused from Zhou et al. (Mater Sci Eng R Rep 2020;141:100548) [17] with original copyright holder’s permission.

2. 2D perovskite: direct X-ray detection

One of the critical requirements of a high-performance X-ray detector is to minimize the dark current at reverse bias so that the current generated at a low X-ray dosage can be well resolved above the dark noise. Compared to their 3D counterparts, 2D perovskites have higher stability, directional charge transport, low dark current, and fast response time [3,28], in turn increasing the bulk resistance to avoid noise [18]. Direct conversion detectors based on low-dimensional perovskite polycrystalline films and single crystals are being developed over the past few years.

In 2020, Tsai et al. [5] demonstrated a thin-film X-ray detector comprising of crystalline,uddlesden–Popper phase-layered perovskites fabricated in a fully depleted p-i-n architecture. The active layers were composed of 2D layered perovskite BA₂MA₂Pb₃I₁₀. Because of the high crystallinity and preferred orientation in the 2D thin film, the p-i-n junction design had a dark current as low as 10^–9 A cm^–2, thereby making it possible to further decrease the X-ray detection limit. The design showed a high diode resistivity of 10¹² Ω·cm in the reverse-bias regime, leading to a high X-ray detection sensitivity of up to 0.276 C Gy_air^–1 cm^–3 [5].

Such high signals are collected by the built-in potential underpinning operation of the primary photocurrent device with robust operation [5]. Their findings suggested a new generation of X-ray detectors based on layered perovskite thin films for future X-ray imaging technologies.

3. 2D perovskite scintillation

Scintillators downconvert high-energy photons (X-ray) to lower-energy photons (UV, visible, or IR wavelengths). 2D layered perovskites can be used to resolve the shortcomings of the widely used scintillators [17]. MHPs are rapidly emerging because of their unique emission properties that are important for the sensitivity. The quantum wells of the 2D perovskites confine the electrons and holes and enlarge the overlap of the electron and hole wave functions. These unique electronic properties increase the excitonic oscillator strength and shorten the excitonic radiative lifetime in 2D perovskites [28].

The scintillation properties of the 2D lead halide perovskites are mainly governed by the exciton properties [3]. The 3D perovskites hardly show efficient scintillation at room temperature because of thermal quenching. The exciton binding energies of 2D halide perovskites are larger than those of their 3D analogs. These deep excitonic levels ensure that the 2D perovskites can resist thermal quenching at room temperature.

In addition, the 2D halide perovskite single crystals exhibited a high LY of over 10,000 photons/MeV at room temperature. They also have a short radiative lifetime and high environmental stability [3]. The enhanced dopant–carrier exchange interactions can be realized because of the relatively high exciton binding energy, especially for 2D perovskites. All these advantages make 2D halide perovskites excellent candidates for X-ray scintillators.

4. Mn²⁺ doped (2D) perovskite

The Mn²⁺ ion has been widely applied in various luminescent materials. It is well known that electron transition from the ⁴T₁ excited state to the ⁶A₁ ground state is generated in Mn²⁺-doped materials (Fig. 7), and thus, the PL spectrum depends strongly on the crystal field surroundings of the host lattice [29,30]. Mn²⁺-doped engineering was implemented in 3D and low-dimensional MHPs.

Figure 7. Schematic representation of the mechanism of PL excitation, energy transfer, and PL emission in Mn²⁺-doped material. The orange lines represent the energy level diagram of the Mn²⁺ ion in a free-ion state (right) and tetrahedrally coordinated environment (left) in a supertetrahedral nanocluster. The dotted arrows indicate nonradiative transitions. PL, photoluminescence. Reprinted with permission from Lin et al. (J Am Chem Soc 2014;136:4769-4779) [24]. Copyright 2014 American Chemical Society.

Fig. 7 shows a schematic representation of the proposed mechanism for the PL excitation and energy transfer. It is generally accepted that the lowest excited state (⁴G) for a free Mn²⁺ ion will split into four energy levels (⁴T₁, ⁴T₂, ⁴E, and ⁴A₁) in the tetrahedrally coordinated Mn²⁺ ion. With increasing doping, more Mn²⁺ ions serve as trap centers to accept the photoexcited electron, thereby enhancing the efficiency of Mn²⁺-related emission [24]. An energy transfer or a charge transfer may lead to a new population in the states. In the case of an energy transfer, the photoexcited electron–hole pair of the host follows a nonradiative Auger-like recombination, transferring the energy from the host to Mn for exciting the Mn²⁺ ground state ⁶A₁ to its excited ⁴T₁ state. Usually, the octahedrally coordinated Mn²⁺ shows an orange to red emission because of the relatively higher crystal-field strength, whereas the tetrahedrally coordinated Mn²⁺ exhibits green-light emission.

In 2020, Xu et al. [31] reported highly efficient X-ray scintillators based on an organic metal halide (C₃₈H₃₄P₂)MnBr₄. This 0D organic metal halide hybrid exhibits green emission peaked at 517 nm with a PL quantum efficiency of ~95%. Its X-ray scintillation properties showed a linear response to X-ray dose rate, a high LY of ~80,000 photon MeV^–1, and a low detection limit of 72.8 nGy s^–1. The X-ray imaging tests showed that scintillators based on (C₃₈H₃₄P₂)MnBr₄ provide an excellent visualization tool for X-ray radiography [31,32].

In addition, extrinsic Mn(II) dopants were incorporated to serve as extra emitting centers and emission shifters to eliminate the photon loss caused by self-absorption. Mn²⁺-doped double perovskites were also widely investigated. Mn²⁺ doping of these perovskites results in broadband emissions. Mn²⁺ induced a new broad emission peak at 590 nm with a long decay time because of the 4T₁–6A₁ transitions of Mn²⁺. The photon energy of Mn emission is ~2 eV, which is much lower than that of the 2D perovskite host (~3.1 eV). This low photon energy suppresses the photon loss induced by self-absorption [3].

On the one hand, Mn dopants capture the energy from the trap states in perovskites, improving the scintillation efficiency. In 2014, Lin et al. [24] reported a simple structure with a vacant core site with an ordered distribution of Mn²⁺ dopants; this structure precluded the formation of Mn²⁺ clusters in the host matrix. They achieved a uniform distribution of Mn²⁺ dopants in the crystal lattice. The PL, X-ray photoelectron spectroscopy, and electron spin resonance spectra revealed the incorporation of Mn²⁺ ions [24].

In 2020, Yu et supl. [33] reported the use of a 2D halide perovskite as a β-ray scintillator. They developed a type of β-ray scintillator based on Mn-doped 2D bromide perovskites. The scintillator showed good thermotolerance and irradiation hardness because of the 2D halide perovskite [33].

As a typical case, the Mn²⁺-doped 2D perovskites show enhanced energy transfer efficiency from the strongly bound excitons of the host to the d electrons of Mn²⁺ ions, resulting in intense orange–yellow emission. This emission is attributed to the spin-forbidden internal transition (⁴T₁→⁶A₁) [17]. This manganese halide has a high LY of up to ~80,000 photons/MeV and a low detection limit of 72.8 nGy s^–1 [31,32]

Herein, we briefly review the status of quantum-enhanced X-ray detection technology, which is emerging as a new promising high-sensitivity X-ray detection technology. In addition, we examine the SPDC X-ray detection method combined with perovskite-based X-ray detection for implementing an ultrasensitive X-ray detector in the future [13].

1. SPDC

SPDC is a nonlinear optical phenomenon wherein it con­verts a higher energy photon into a pair of entangled lower-energy photons according to the law of conservation of energy and momentum. Depending on the configuration of the conversion process, entanglement can be generated; signal photons and idler photons should be generated in at least a two-mode state, for example, with horizontal and vertical polarizations [34].

2. Quantum-enhanced X-ray detection

In 2019, Sofer et al. [11] presented an experimental demonstration of quantum-enhanced detection at X-ray wavelengths based on the SPDC principle. They showed that the X-ray photon pairs that are generated by SPDC can be used for generating heralded X-ray photons and for obtaining the statistics of the single photons. They utilized the down-converted entangled photons to demonstrate the feasibility of improving the visibility and SNR of an image with a small number of photons in an environment with a noise level higher than the signal by many orders of magnitude [11]. An example of the experimental setup is shown in Fig. 8.

Figure 8. Scheme of the experimental setup. The purple beam indicates the pump beam; the green beams are the signal and idler beams; and the red beams represent the noise. The object has three slits. The detectors are Si detectors. Reused from Sofer et al. (Phys Rev X 2019;9:031033) [11].

Sofer et al. [11] used a pump beam at 22.3 keV to generate the photon pairs via X-ray PDC in a nonlinear diamond crystal. After SPDC, the entangled signal photon and the auxiliary photon were subjected to two entangled photon absorptions by utilizing the property of time–energy correlation. This pair of photons can be used as a predicted X-ray photon (i.e., a signal photon) and an auxiliary photon (i.e., idler). In the experimental device, the auxiliary (idler) photon is collected by the auxiliary detector and is used as the trigger for the second detector. The second photon, which we denote as the signal photon, is collected by a second detector. Si detectors that provide a signal were used; thus, Sofer et al. [11] could resolve the number of detected photons and their photon energies.

The signal photon passing through the imaging transmittance was registered by the data acquisition system by the two entangled-photon absorption processes only when an auxiliary photon was detected. Thus, the signal photon was filled with only the true single photon predicted by the pump photon. Hence, even under low-photon conditions, a huge SNR increase was achieved. This experiment showed the potential for a thousand-fold improvement.

The images obtained by using the quantum detection procedure are shown in Fig. 9a. For comparison, the images obtained by classical measurements with a comparable average number of photons are also shown. The image in Fig. 9b was obtained by illuminating the object with classical radiation and measuring the intensity only at the object detector. They obtained the image shown in Fig. 9a by using coincidence measurements between the ancilla and object detectors with classical radiation. The advantages of the quantum scheme over the classical schemes as indicated from the comparison of Fig. 9a with Figs. 9b and c are prominent.

Figure 9. Reconstruction of the image of the triple-slit object by (a) quantum radiation, (b) classical radiation, and (c) classical coincidence counting. The average number of counts is comparable among the panels. In each of the panels, the horizontal axis represents the relative position of the object, and the vertical axis represents the number of events that are detected by the detection system. Reused from Sofer et al. (Phys Rev X 2019;9:031033) [11].

These methods can be used for measuring weak signals. This work is anticipated to pave the way for more quantum-enhanced X-ray regime detection schemes. The procedure reported by Sofer et al. [11] possesses great potential for improving the performance of X-ray measurements.

Despite the advantages of halide perovskites and the achievements realized in their use for high-energy radiation detection, the development of halide scintillators remains a challenge [21]. Therefore, we highlight some urgent problems and propose some promising directions for further research.

A perovskite solar cell (PSC) deteriorates when exposed to UV light. Likewise, when X-rays are applied to halide perovskite-based X-ray detectors, the stability problem arises. The mechanism of UV-induced PSC performance degradation and recovery has been extensively studied. However, the effects of X-rays are still not well understood; the degradation and possible recovery mechanism in HP-based X-ray detectors should be elucidated.

PSCs are also very sensitive to humidity, though the degradation under humid conditions can be remarkably improved by encapsulation [35].

The challenge is to prepare large-scale, high-quality 2D perovskite single crystals via solution processing [3]. The research on 1D MHP is in the infancy stage.

The extension of quantum optics to the X-ray regime requires overcoming many challenges; for example, a source that can generate entangled photons at high flux at the required wavelengths is necessary.

Quantum-enhanced X-ray detection methods can be beneficial for measuring weak signals. Sofer et al. [11] demonstrated the SPDC experiment for quantum-enhanced detection using X-ray photons from an accelerator. Although synchrotrons can provide line emissions with sufficient brightness, they are big and extremely expensive facilities, and they cannot be used in clinical practice [36]. For realistic SPDC, we need a portable X-ray laser, not synchrotron X-ray.

Finally, we suggest that the quantum-enhanced X-ray detection method combined with perovskite X-ray detectors be adopted to implement a future ultra-sensitive X-ray detector. The use of built-in X-ray lasers will open up additional possibilities. The development of a portable X-ray laser should be prioritized. It is very likely that quantum-enhanced effects can be observed with perovskite sources with much higher output yields. We anticipate that this work will pave the way for more quantum-enhanced X-ray regime detection schemes.

The authors have nothing to disclose.

The data that support the findings of this study are available on request from the corresponding author.

Conceptualization: Nam Joong Jeon, Jung Min Cho, and Jung-Keun Lee. Data curation: Jung Min Cho and Jung-Keun Lee. Investigation: Nam Joong Jeon, Jung Min Cho, and Jung-Keun Lee. Supervision: Nam Joong Jeon and Jung-Keun Lee. Writing –original draft: Jung-Keun Lee. Writing – review & editing: Nam Joong Jeon, Jung Min Cho, and Jung-Keun Lee.