X-ray detection is important in many areas, such as medicine (diagnosis and treatment), nondestructive product checking, and scientific study. The market is mostly made up of direct X-ray detectors, such as α-Se selenium, and indirect detectors, such as caesium iodide:thallium iodide (CsI:Tl) and gadolinium oxysulfide:terbium (Gd2O2S:Tb) scintillators on α-Si photodiode. On the other hand, these devices have limits because they can’t block hard X-rays well enough for mammograms.
Because they have such great optical qualities, organic-inorganic hybrid perovskites have worked very well in solar cells, light-emitting diodes, and photodetectors. Perovskites have been made into both direct and indirect X-ray detectors that are very sensitive, can cover a big area, and can be scaled up. Because they have a lot of atoms, they can block X-rays very well. When it comes to sensitivity, direct perovskite detectors are at least three orders of magnitude better than α-Se. Their mobility-lifetime product is also three to five orders of magnitude higher. Perovskite nanocrystal scintillators can detect more radiation than a clinic’s image dose. This makes them a good choice for the next generation of medical radiation detection.
Working Mechanism of X-ray Detectors
When they hit an item, X-ray rays are either sent through, absorbed, or scattered by the atoms of the object. X-rays with energies between 15 and 150 keV are used for medical imaging. They can behave in four different ways: they can be transmitted, they can scatter light in Rayleigh or Compton ways, or they can have a photoelectric effect. The direction of the X-ray photon changes with Rayleigh scattering, but no energy is lost. This lowers the quality of the images. If the energy and direction of an X-ray photon change, this is called Compton scattering. It can affect X-ray images up to 1 MeV in energy. The photoelectric effect is a process in which X-ray rays combine with electrons in the inner shell and leave all of their energy to get away from the bound energy.
The photoelectric effect tells the difference between atomic makeup and makes a picture with a Z number contrast. As X-rays pass through the body, they are weakened in different ways by different tissues and organs that are thicker or have different atomic makeup. This creates an X-ray difference picture on the monitor. The X-ray dose difference will be read by the receiver as an electrical or digital signal contrast picture.
Indirect and direct X-ray detectors are the two main types used in X-ray photography. Indirect X-ray detectors use scintillators to turn incoming X-ray photons into visible light. Light sensors like amorphous Si photodiodes, photomultiplier tubes, CMOS detectors, and thin-film phototransistor arrays pick up this light. Photoconductors directly turn the energy from captured X-rays into electrical signs in direct X-ray detectors. These detectors have better spatial clarity and are easier to set up.
Material Properties of Ideal X-ray Detectors
How well X-ray detectors work depends on which photoconductors are used and how they are designed. Some important things that photoconductors should have are a high X-ray attenuation coefficient (𝛀), low electron-hole pair creation energy (W), high mobility and lifetime product (𝼇𝼏) product, low trap density, a large-area and scalable photoconductor, even response and sensitivity across the whole area, and great stability for long-term use.
If you have a good X-ray detector with a high X-ray absorption coefficient (𝛀), more atoms will interact with the X-rays that hit it. This will cause photoelectric effects that produce higher photocurrent or photovoltage readings. For X-ray detection, photoconductor materials with high Z atoms, like Pb (Z = 82), I (Z = 53), and Bi (Z = 83), work better. You can use the National Institute of Standards and Technology (NIST) website to find out the exact X-ray mass attenuation factors of different atoms. This will help you figure out how well the material attenuates X-rays.
For X-ray sensing to work, the creation energy of the electron-hole pair (EHP) must be low. More electron and hole pairs are made by a photoconductor with lower EHP generation energy, which leads to stronger signs. For high EHP generation, smaller bandgap photoconductor materials are better, but they also have more thermally produced carriers, which are noisy and need to be slowed down.
A high photoconductor is needed to make X-ray devices that are sensitive to radiation and charge collection that works well. For digital dynamic X-ray photography, low trap density photoconductors are very sought after.
Indirect X-ray detectors use electron-hole pairs that recombine radiatively instead of being taken out by electrodes, which produces UV-visible light. For a high light output, you want high dielectric scintillations. The spectrum emission of the scintillators should fit the sensor layer’s absorption wavelength range. It is best to use scintillators with the above qualities to find X-rays.
Conventional X-ray Detectors
Digital radiography (DR) X-ray flat panels are usually put together with α-Se (direct detection material) or caesium iodide (indirect detection material) (CsI:TI or Gd2O2S:Tb) on top of the α-Si TFT layer for reading the signal. Arsenic (As) is added to α-Se to keep it stable and stop it from crystallising. This stops dark current and current leaks. But this leads to deep hole traps and a shorter hole drift length. To make the electron and hole bands equal, changes are made to As and Cl in α-Se. This keeps the 𝼏 product between 10−7 and 10−5 cm2/V.
In mammography, where a 20 keV X-ray source is used, α-Se flat-panel imagers are mostly used. The EHP generation energy in α-Se is based on the electric field F instead of the X-ray photon energy E. This could be because there are so many traps. The fact that α-Se has almost no picture lag and very consistent X-ray reaction and pixel-to-pixel sensitivity is a big part of its economic success.
CsI:Tl and Gd2O2S:Tb are the two most common commercial scintillators for DR, especially when it comes to high-energy X-ray medical imaging. X-rays are strongly absorbed by both CsI and Gd2O2S because their atomic numbers are very high. Adding Tl to CsI makes scintillation work better and creates longer range emission spectra that fit the absorption spectra of the α-Si detector. But CsI:Tl detectors have more afterglow than Gd2O2S:Tb detectors, which is better for fast reaction and little afterglow, especially for CT or fluoroscopy tasks that need to scan for a long time.
Perovskite X-ray Detectors
Usually, X-ray image devices are made of artificial semiconductor materials. These have strong electronic and physical qualities, but they are unstable when exposed to high electric fields or when they are made in expensive ways. Organic materials are easy and cheap to work with in solutions, but they don’t absorb X-rays very well. These organic-inorganic combination perovskite materials have the strength of inorganics and the ease of production of organics. Huang and his colleagues suggested perovskite radiation detection in 2015 and found that the electron-hole diffusion length in a MAPbI3 single crystal (SC) could be longer than 175 μm. It was discovered that a MAPbI3 SC that is 3 mm thick could identify cesium-137 gamma radiation and produce a steady current of 36.6 ± 0.3 nA, which means that it could turn light into electrons 3.9% of the time. It has a high atomic number and is a straight bandgap semiconductor. It can absorb X-rays 10 times better than α-Se. Its ability to be successful in X-ray detecting is also due to its low trap density and good ability to handle defects.
Direct Perovskite X-ray Detectors
Heiss and his colleagues were the first to show that perovskites could be used for X-ray photography. They did this by using spray-coating to make a 60 μm-thick layer of polycrystalline MAPbI3. After being covered with a spray, the perovskite film was used to make photovoltaic and photoconductive devices. The photovoltaic device had a p–i–n structure and a 75 keV energy peak. Putting two gold electrodes on a spray-coated perovskite film turned it into a lateral photoconductive device. The dark current of photoconductive MAPbI3 detectors is almost half of the X-ray generated current and rises in a straight line with the applied bias.
The big dark current is mostly because there is no Schottky junction. This can be fixed by adding charge carrier blocking layers as a solar device to lower noise and raise the single noise ratio. The device’s 𝼇𝼏 output is about 10−7 cm2/V, which means the film isn’t very good.
MAPbBr3 has a slightly lower X-ray absorption power than MAPbI3, but it has more benefits, such as a cubic crystal shape, less ion migration, fewer thermally excited charge carriers, and an easy way to check the quality. MAPbBr3 single-crystal devices also have fewer traps because there are no grain boundaries. This means that carriers can move around freely and last a long time.
Wei et al. made MAPbBr3 single-crystal devices using a modified antisolvent method. They found that a stoichiometry ratio of 0.8 for PbBr2/MABr worked best because of the difference in how well PbBr2 and MABr dissolve in water. The clear MAPbBr3 SCs that had just been grown were treated with UV-O3 to improve their ability to remove charge carriers. The MAPbBr3 SC device is 2 mm thick and has a record g-value of 1.4 × 10−2 cm2/V and a mass resistance of 1.7 × 107 Ω cm. It can sense 80 μC/(Gyair cm2) with a small bias of 0.1 V, which is four times better than other α-Se detectors that have been reported.
Huang and his colleagues used a low-temperature solution-processed molecular bonding method to grow MAPbBr3 SC on silicon plates. The chosen brominated (3-aminopropyl) triethoxysilane (APTES) molecules can connect the MAPbBr3 SC to Si substrates that are both physically strong and electrically conductive. This makes it possible to include MAPbBr3 SC on CMOS reading circuits to create active matrix flat-panel imagers (AMFPIs). When 8 keV X-rays hit the Si-integrated device, its sensitivity went up to 2.1 × 104 ¼C/(Gyair cm2), which is over 1000 times higher than the sensitivity of α-Se devices.
Park and his team created the first large-area polycrystalline MAPbI3 AMFPI. They did this by putting an 830 μm-thick layer between two interlayers of polyimide (PI)–perovskite composites. To connect the MAPbI3 film to reading integrated circuits (ROICs), the blade-coating method and tape-automated glueing process are used to place it. With a bias of 10–200 V, the sensitivity goes up to 1000–11,000 µC/(Gyair cm2). The pixel pitch is 70 μm, and the geometry fill factor is 76%.
The blade-coating method and the mechanical sintering method can both be used to make large perovskite X-ray detectors. Gebhard J. Matt and his colleagues showed a small MAPbI3 chip by pressing MAPbI3 power at 0.3 GPa for five minutes. The perovskite chip as it was made has a density of 3.76 g/cm3, which is close to the MAPbI3 SC. The MAPbI3 wafer-based device has a 𝼇𝼏 product of 2.0 × 10−4 cm2/V, which gives it a good sensitivity of 2527 μC/(Gyair cm2) at 0.2 V/μm when exposed to X-rays.
Researchers have tried different types of inorganic perovskites to make the device stable and have a small dark current. These include two-dimensional inorganic perovskites ((NH4)3Bi2I9), inorganic perovskites (CsPbBr3), and double perovskites (Cs2AgBiBr6). Tang and his colleagues mostly studied Cs2AgBiBr6 double perovskite, which has recently become known as a possible and hopeful semiconductor for X-ray imaging because it is stable, doesn’t contain any harmful substances, and has a low detection limit.
Perovskite has also been used to make bendable X-ray detectors by putting it into a sheet with holes in it. The perovskite-filled membrane (PFM) can be made thinner or thicker by putting together different layers of PFMs. It can sense up to 8696±228 μC/(Gyair cm2) at 0.05 V/μm when exposed to X-rays. The high sensitivity is due to the fact that perovskite materials can absorb X-rays well and have high carrier movement and lifetime products. The flat detector doesn’t give as good of a picture as the flexible detector, which can be put into a pipe for imaging.
Perovskite X-ray Scintillators
Halide perovskites can be made into scintillators with different ratios. One example is caesium lead halide perovskite nanocrystal-based scintillators. By changing the cesium-oleate reaction with various lead halides, a group of perovskite CsPbX3 (X = Cl, Br, and I) nanocrystals were made that glow in a narrow, distinct way when exposed to X-rays. Under 10 kV and 5 μGy/s X-ray, the CsPbX3 nanocrystals strongly absorb X-rays and have a high emission quantum yield. This means that a 100 μm-thick layer of nanocrystals can emit X-rays as well as a 5 mm-thick bulk CsI scintillator.
There is 420 times less X-ray dose than the lowest dose rate that can be detected by the perovskite nanocrystal scintillators (13 nGy/s). 44.6 ns is the reaction time to X-ray photons. This makes sure that the X-ray detector responds quickly, which is very important for dynamic X-ray photography. We made the perovskite nanocrystal-based AMFPI by putting the perovskite nanocrystal scintillators on an industrial α-Si photodiode array. At a precision of 2.0 lp/mm, the MTF of the perovskite nanocrystal-based X-ray imager is higher than that of the CsI detector.
Perovskite nanocrystals are more sensitive than industrial α-Se detectors, as shown by the fact that they can take clearer pictures of the Apple iPhone. To make a more in-depth comparison, Tables 14.1 and 14.2 show the material qualities and image performance of scintillators and direct X-ray detectors.
Characterization of X-ray Flat Panels
After the materials are fabricated as AMFPIs, several parameters can be used to estimate the imager quality, including sensitivity, detective quantum efficiency (DQE),
MTF, pixel-to-pixel uniformity, imaging lag, and ghosting.
Sensitivity
X-ray detectors’ sensitivity is based on how much charge they receive per unit area per unit energy that hits them, as shown in the equation: Se = Q
AXE = ¥I A¥D (14.1). The sensitivity of X-ray imagers is linked to quantum efficiency and charge gathering efficiency in a positive way, and the EHP generation energy of the photoconductor in a negative way.
DQE
DQE measures how well X-ray rays are turned into useful picture data (SNRoutput) based on the signal-to-noise ratio (SNRinput). It is often used to guess how good digital radiography (DR) systems are. A high DQE means that the output signal is strong. It is caused by high EHP conversion, high charge collection efficiency, and low noise. The EHP is based on the features of the material or on its W. The ability to receive charges is based on the material’s, the electric field F, and the device’s thickness L. For high charge gathering effectiveness, raise F. But this may cause a lot of noise or a high dark current.
Different measurement factors, like the amount and energy of the incoming X-ray, can change the DQE number. The idea of IEC 62220 was made in 2003 to solve this problem. The following method can be used to figure out the output signal: The SNR2 output is equal to S2MTF2(f)NPS(f), where S is the average signal strength, MTF(f) is the image’s MTF, S × MTF(f) is the signal’s actual output, and NPS(f) is the system’s noise power spectrum.
A two-dimensional Fourier function can be used to change the noise picture into a centrally symmetric image. The input signal to noise ratio (SNRinput) should stay the same if the X-ray quantum noise that comes in is white. The DQE values are linked to picture clarity in a positive way and to the initial X-ray dose in a negative way. When the DQE number is high, it means that the picture quality is good even when the X-ray energy is low.
MTF
A very important way to measure how well an X-ray imager can resolve details is the modulation transfer function (MTF). It is usually measured in line pairs per millimetre and gives an idea of how clear and sharp the picture from the X-ray imager is. Imaging an irregular object is a good way to show how an imaging device changes the quality of a picture. The MTF can be written as MTF = fmax – fmin, where fmax and fmin are values that are not negative and show the image’s highest and lowest intensities, respectively. When everything is perfect, MTF = 1. But in real life, noise and X-ray scattering make the contrast not as good as the object. There are two ways to measure MTF: the curved edge method and the single-slit aperture method. To figure out MTF, most people use the slanted edge method, which involves taking a picture of the sharp edge of a 1 mm thick tungsten plate that is raised at an angle of 1.5° to 3° from the detector’s pixel array. Next, a finite-element convolution is used to separate the edge spread function (EFS) into the line spread function (LSF). The MTF is then found by using fast Fourier transformation (FFT) on the LSF.
Pixel-to-Pixel Uniformity
X-ray imagers need to have the same reaction over a big area so that they don’t get things wrong when they look at pictures. But pixel-to-pixel regularity can be changed by things like the uniformity of the material, random X-ray photon noise, and electrical reading system noises. It is possible to make things more regular by making the photoconductor grain size smaller than the pixel size. As the incoming X-ray dose goes up, pixel-to-pixel differences can get smaller because of processes like charge carrier creation and collection, carrier trapping and detrapping, and absorption that depend on dose. After imaging, changes in the pixels can be fixed, and linear reactions let pictures be taken in the dark and with X-rays without the object.
Imaging Lag
Lag is when the picture charge from earlier X-ray scans moves to later frames. This is mostly noticeable in dark photos. A chopper can be used to measure the photocurrent to test it. Lag is equal to either l = B1 – B0S (14.7) or lx = Bx – B0S (14.7), where S is the X-ray signal, B0 is the dark current measured before exposure, B1 is the dark current measured 0.5 seconds after exposure, and BX is the mean dark current measured during exposure. Lag goes down as time goes on after exposure and is very dependent on frame rate.
Ghosting
X-ray imagers have a feature called “ghosting” that lowers sensitivity because of past scans. It is the difference between the sensitivity measured at the nth X-ray exposure and the sensitivity measured at the first X-ray exposure. When there is ghosting, the photoconductor’s sensitivity changes based on the previous picture. This change can only be seen in the next X-rays. This happens because charge carriers get trapped and may reunite with opposite charge carriers, making it less effective to collect charges in later exposures. This makes it less sensitive and changes how the electric field is spread across the photoconductor, which changes how new carriers are extracted. Ghosting is not the same as lag effect, and you can only see it in the next set of X-rays.
Summary and Outlook
Perovskites with a lot of Z atoms might be able to successfully weaken X-rays, which would make them more sensitive and allow for lower dose rates to be detected than standard α-Se X-ray imagers. Low-temperature solution-growth perovskite SCs make it possible to make single-crystal X-ray imagers that cover a big area and don’t have any ghosting or image lags. Composition-tunable perovskite also makes it easier to make X-ray detectors that are very sensitive, steady, low-trap, and low-dose.
Still, more work needs to be done to turn perovskite into AMFPIs that can be sold. Stability studies on perovskite X-ray imagers are needed because problems with stability have made it hard to sell perovskite solar cells. Once the device is made, it needs to be checked for financially important factors like DQE, picture lag, and ghosting so that it can be compared to commercial detectors and improved even more. To make real things, you need to do a lot of study. In the future, high-quality perovskite X-ray devices will help medical images more.
Leave a Reply