From Classic Detectors to Spectral Counting: A Journey Through X-ray Detection Systems for NDT

 

1. Introduction to X-rays and Detection:
    Fundamentals and Context in Non-Destructive Testing

 

The use of X-rays in Non-Destructive Testing (NDT) represents one of the most established and sophisticated techniques for structural inspection of materials.

Industrial radiography, based on the transmission and absorption of X-rays, enables the identification of internal defects without compromising the integrity of the examined component.

1.1 When atoms become radioactive, they can emit different types of radiation,

Technological progress, particularly in the field of detectors, has revolutionized the quality, speed, and reliability of radiographic inspections, evolving from analog technologies to highly specialized digital systems.

This transformation has taken place over decades of technological innovation, leading to the development of detectors that are increasingly sensitive, precise, and adaptable to various industrial needs.

In particular, the evolution of detectors has directly impacted the effectiveness of NDT, contributing to the reduction of detection errors, increased productivity of inspection processes, and the possibility of integrating controls with automated and intelligent systems.

X-rays, discovered in 1895 by Wilhelm Conrad Röntgen, immediately found applications in the medical field, but their potential in NDT emerged strongly in the 20th century, especially in the aerospace, energy, and mechanical industries.

Detectors, initially based on photographic films, have evolved to include solid-state technologies, photon counting systems, and multi-energy solutions capable of providing detailed and analytical readings of the internal composition of materials.

1.2 Monoenergetic X-rays are transmitted through different layers of an attenuator with an attenuation coefficient, μ, of 20% per unit thickness. Attenuation occurs exponentially, as illustrated by the graph (bottom left) showing transmitted primary X-rays as a function of the attenuator thickness. In the semilogarithmic graph (bottom right), the exponential curve appears as a straight line for the mono energetic X-ray beam.

This article aims to explore, with a technical-scientific approach, the entire spectrum of X-ray detection technologies currently available and under development, starting from the fundamental physical concepts and moving towards advanced applications in the most critical industrial sectors.

The NDT context requires reliable, repeatable, and precise solutions capable of adapting to components of varying sizes, different materials, and complex geometries.

The selection of the most suitable detector depends not only on the physical characteristics of the material to be analyzed, but also on the type of defect sought, the spatial and energy resolution required, the necessary acquisition speed, and the environmental conditions during inspection.

1.3 The transmitted bremsstrahlung spectrum changes its energy distribution and photon fluence following X-ray attenuation. (Top) The curve shows the spectrum of 120 kVp transmitted through soft tissue (water) thicknesses of 0, 5, 10, 15, and 20 cm, illustrating the continuous decrease in photon number and the shift of the spectrum toward a higher effective energy. (Bottom) The variation of effective energy is shown more clearly for each transmitted spectrum normalized to the peak energy of the transmitted X-rays in the individual graphs.

In an increasingly digitalized, connected, and automated industrial world, the adoption of innovative detectors allows for improved final product quality, reduced production times and costs, and guarantees compliance with increasingly stringent safety standards.

Modern detectors no longer merely provide a two-dimensional image of the inspected object, but allow for three-dimensional data, spectral information, and even quantitative assessments of the type and severity of detected defects.

Moreover, the integration with artificial intelligence algorithms and advanced analysis software enables intelligent management of acquired data, facilitating automated diagnosis, traceability, and inspection reporting.

In this continuously evolving scenario, understanding the operation, advantages, and limitations of different types of X-ray detectors is essential for researchers, professionals, and industry operators.

 

1.4 Detection of the luminous field from X-rays to visible light using pixelated arrays of perovskite nanocrystals.

The article is structured into ten sections, each addressing a specific aspect of X-ray detection, starting from the physical bases of radiation-matter interaction, through traditional technologies and advanced semiconductors, up to the latest frontiers represented by perovskite sensors, photon counters, and multi-energy systems.

There are also in-depth discussions on practical applications in various industrial and scientific fields, as well as a concluding reflection on the future prospects of this strategic technology.

Supporting the discussion, comparative tables, technical formulas, practical examples, and critical considerations will be presented to help understand the current state of the art and guide the selection of the most appropriate solution for operational needs.

The choice of detector is never neutral but represents a decisive factor for the success of inspections and the overall quality of the production process.

In an industrial context where competitiveness also relies on the ability to innovate and guarantee high-quality standards, investing in advanced detection technologies is a key element to strengthen safety, traceability, and efficiency in non-destructive testing operations.

2. Physical Principles of X-ray Detection:
   From Interaction with Matter to Electronic Signal

  

Understanding the physical principles underlying X-ray detection is a fundamental prerequisite for any technological development in the field of non-destructive testing.

When a beam of X-rays strikes a material, the interactions that occur depend mainly on the energy of the incident photons and the atomic composition of the target.

The three main interaction mechanisms are the photoelectric effect, Compton scattering, and, at very high energies, pair production.

 2.1 Illustrative summary of interactions between X-rays and γ-rays.

(A) The primary unattenuated beam does not interact with the material. (B) Photoelectric absorption causes the complete removal of the incident X-ray photon with energy greater than the electron binding energy in its shell, with the excess energy distributed as the kinetic energy of the photoelectron. (C) Rayleigh scattering is an interaction with an electron (or an entire atom) in which no energy is exchanged; the energy of the incident X-rays equals that of the scattered X-rays, with only a small angular deviation in direction. (D) Compton scattering interactions occur with essentially unbound electrons, with energy transfer shared between the recoil electron and the scattered photon, with energy exchange described by the Klein-Nishina formula.

  

The photoelectric effect predominates at low energies and is particularly useful in high atomic number materials, as the absorption coefficient varies with the cube of the atomic number.

This effect consists of the complete absorption of the photon energy by a bound electron, which is then ejected from the atom.

Its relevance in solid-state detectors is fundamental as it provides high conversion efficiency in high-Z semiconductors.

Compton scattering, dominant in intermediate energy ranges, results in partial photon energy loss and angular deviation, contributing to the generation of a continuous background in detected signals.

2.2 Graph showing the probability distribution of scattering as a function of angle relative to the direction of the incident photon. Three energies (20, 80, and 140 keV) show a relatively isotropic distribution (in all directions) of scattering at low energies, which becomes more peaked (smaller scattering angle) at higher energies.

Finally, electron-positron pair production occurs only for photons with energy above 1.022 MeV and is typically not relevant in standard NDT contexts, although it is important in scientific applications or high-energy environments such as synchrotrons.

Once the photon energy is transferred to matter, it is necessary to convert this event into a readable electrical signal.

There are mainly three energy-to-signal conversion mechanisms: ionization, scintillation, and direct conversion.

Ionization occurs in gas detectors, where the photon passage ionizes gas molecules, generating ion-electron pairs that are collected by electrodes under an electric field.

This principle forms the basis of ionization chambers, Geiger–Müller counters, and proportional detectors, known for their robustness and ability to operate in extreme environments but limited in spatial resolution and energy sensitivity.

Scintillator materials, such as NaI(Tl) or CsI(Tl), emit visible light photons in response to X-ray absorption; the emitted light is then converted into an electrical signal by photomultipliers or solid-state devices like photodiodes.

2.3 Monoenergetic X-rays are transmitted through various layers of an attenuator with an attenuation coefficient, μ, of 20% per unit of thickness. Attenuation occurs exponentially, as illustrated by the graph (bottom left) of primary X-rays transmitted as a function of attenuator thickness. In the semilogarithmic graph (bottom right), the exponential curve is a straight line for the monoenergetic X-ray beam.

This method is widely used for its high detection efficiency and flexibility, although the energy resolution is generally lower compared to direct conversion.

In direct conversion detectors, X-ray photons directly interact with a high-Z semiconductor, generating electron-hole pairs that are collected to form an electrical signal proportional to the absorbed energy.

Materials like cadmium telluride (CdTe) and cadmium zinc telluride (CZT) are paradigmatic examples of this technology, which offers an excellent combination of energy resolution, compactness, and rapid response.

The choice of detection mechanism depends on specific application requirements.

 

2.4 Emission spectra and spectral sensitivity curves of various luminescent materials and optical sensors.

In industrial NDT, for example, high absorption efficiency and good spatial resolution are essential, as well as compatibility with high-radiation environments and the capability to operate in real-time.

Detector performance can be described in terms of quantum efficiency (QE), energy resolution, response time, signal linearity, and temporal stability.

Moreover, electronic background noise, the ability to operate in counting or integration mode, and resistance to saturation or pile-up effects are crucial parameters to assess the suitability of the detection system.

The fundamental equations describing radiation attenuation derive from the Lambert-Beer law, which explains the reduction in the intensity of a beam attenuated according to the thickness of the material traversed and its attenuation coefficient.

The relationship is expressed by the formula:

 

 

where:

• I is the transmitted intensity,
• I₀ is the initial intensity,
• μ is the attenuation coefficient,
• x is the material thickness.

     

"This model forms the foundation for designing all radiographic and tomographic imaging systems, as it quantifies radiation absorption in materials."

Finally, it is important to emphasize that the combination of physical interaction principles and conversion technologies constitutes the basis on which future innovations are built.

 

 

2.5 The relentless research into advanced X-ray detection technologies has been significantly strengthened by the emergence of metal halide perovskites (MHP) and their derivatives, which exhibit remarkable luminescence yield and sensitivity to X-rays. This comprehensive review explores cutting-edge approaches to optimize MHP scintillator performance by enhancing intrinsic physical properties and employing engineering strategies based on radioluminescence (RL), highlighting their potential to develop materials with superior detection capabilities and high-resolution X-ray imaging. Initially, we explore recent research focused on effectively designing the intrinsic physical properties of MHP scintillators, including luminescence yield and response times. Furthermore, we examine innovative engineering strategies involving stacked structures, waveguide effects, circularly polarized chiral luminescence, increased transparency, and fabrication of flexible MHP scintillators, all able to effectively manage RL light for high-resolution, high-contrast X-ray imaging. Finally, we provide a roadmap for next-generation MHP scintillator development, emphasizing their transformative potential in high-performance X-ray detection systems.

With advancements in nanofabrication technologies, readout electronics, and signal processing, modern detectors can not only improve traditional performance but also explore new imaging and analysis modes.

Integration with machine learning algorithms and dynamic calibration based on material type or the incident beam’s energy profile are just some of the directions in which research is moving.

Thus, a deep understanding of physical principles is not merely theoretical but constitutes the indispensable foundation for designing and implementing efficient, reliable solutions tailored to the increasingly complex needs of contemporary non-destructive testing.

 

3. Traditional Detection Technologies:

    Films, Gas Chambers, and Indirect Scintillators

In the landscape of X-ray detection devices, traditional technologies still hold a significant place today, despite the progressive advancement of digital and high-resolution solutions.Radiographic films, gas chambers, and indirect scintillator detectors represent the historical foundations upon which modern detection systems have been developed.

3.1 applications spanning the fields of medical imaging, security surveillance,

scientific research, and non-destructive industrial testing.

Each traditional technology features its own physical architecture, specific mode of interaction with radiation, and consequently a defined performance profile.

Understanding the strengths and limitations of these approaches is essential to evaluate their use in contexts where technological innovation has not yet entirely replaced the proven effectiveness of these systems.

Radiographic films, for example, have been the reference standard in industrial radiography for decades.

3.2 Ideal projection radiography represents the primary fluence of X-rays transmitted from a point source through the object and incident on the detector, as illustrated on the left for a uniform incident fluence, I0, and transmitted fluences I1, I2, and I3 through tissue, air, and bone, respectively. The subject contrast is the difference between object and background signals, e.g., (I1 - I2)/I1 and (I1 - I3)/I1. On the right, a typical situation in the presence of scatter is shown, demonstrating the loss of subject contrast and a reduced difference between incident and transmitted radiation intensities.

  

Their operating principle is based on the interaction between X-ray photons and a photosensitive silver halide layer, which undergoes a latent chemical modification upon radiation exposure.

This effect is later developed through chemical processes, producing a high-resolution visible image of the inspected component.

Films offer excellent spatial resolution, which can exceed 50 line pairs per millimeter, making them ideal for detecting very small or thin defects.

However, they have significant disadvantages in terms of logistics, processing times, storage, and the inability to integrate immediately with digital workflows.

Furthermore, quality control heavily depends on operator skill and consistency in development processes, factors that limit reliability in high-productivity industrial environments.

Gas chambers constitute another important category of traditional detectors.

They operate via ionization of the gas contained in a sealed volume, typically argon or xenon, where X-ray photons generate ion-electron pairs collected by electrodes to produce an electrical signal.

Geiger-Müller counters, proportional counters, and ionization chambers are classical examples of this technology.

Although less precise in spatial resolution compared to other solutions, gas chambers are distinguished by their simple construction, robustness, and low cost, making them suitable for environmental monitoring, dosimetry, and applications in harsh environments.

However, their detection efficiency decreases rapidly with increasing photon energy, making them less suitable for imaging thick or very dense materials.

Additionally, they are sensitive to pressure and temperature variations and require frequent calibration to maintain signal consistency.

Indirect scintillation detectors complete the panorama of traditional technologies, representing an intermediate transition between purely analog and digital systems.
 

 

3.3 Depending on their operating mechanism, X-ray detectors can be indirect or direct.

 

 

These devices use scintillator materials which, upon interaction with X-rays, emit visible light proportional to the absorbed energy.

The produced light is then converted into an electrical signal via photomultiplier tubes or CCD/CMOS sensors, enabling digital image acquisition.

Common scintillator materials include silver-activated zinc sulfide (ZnS:Ag), thallium-activated sodium iodide (NaI:Tl), and thallium-activated cesium iodide (CsI:Tl), each with different light efficiency, decay times, and density characteristics.

This technology offers good absorption efficiency and acceptable image quality, with the advantage of being integrable into digital chains for real-time visualization and storage.

However, energy resolution is lower compared to direct conversion systems, and image quality may be influenced by detector geometry, internal reflection of the light signal, and optical sensor calibration.

3.4 Summary of all introduced VUV photodetector types

The main advantages (or features), disadvantages (or bottlenecks), and application fields of each detector type are listed (from top to bottom). Note that the specific subject of description has been highlighted.

The use of traditional detectors is still justified in numerous application scenarios.

In contexts where priority is robustness, ease of use, and low cost-such as routine inspection of standard metal components or compliance verification of simple elements traditional technologies remain a valid choice.

Moreover, in situations where digital infrastructures are limited, such as in developing countries or isolated industrial environments, relying on analog solutions remains a reliable strategy.

However, it is clear that the processing, automation, and integration potentials offered by digital detectors increasingly overshadow traditional solutions, reserving them for well-defined operational niches.

In-depth knowledge of the performance, limitations, and optimal operating conditions of these devices is nevertheless indispensable to make informed choices, balancing technological effectiveness with economic and operational constraints.

Future evolution could lead to greater hybridization between traditional and digital elements, leveraging the best of both worlds to achieve reliable and scalable performance across a wide range of application scenarios.

Traditional detectors should not be considered a relic of the past but rather a historically significant component and, in certain cases, a technically valid option even in contemporary inspection practices.

4. Direct Conversion Solid-State Detectors:

    CdTe, CZT and the Evolution of Spectral Precision

The advent of direct conversion solid-state detectors has represented a significant technological leap in the field of X-ray detection, especially for applications where energy precision, compactness, and reliability are essential.

These devices exploit directly the interactions between X-ray photons and semiconductor material to generate electric charges, avoiding the intermediate stages typical of scintillation technologies.

Among the most significant materials used for direct detection are cadmium telluride (CdTe) and cadmium zinc telluride (CZT), both characterized by a high atomic number and good charge carrier mobility, factors that ensure remarkable absorption efficiency even at medium-high energies and spectral resolution superior to conventional detectors.

 

 

Properties of different detector materials

4.1 Comparison of properties of different detector materials

 

These materials are distinguished by their ability to operate at room temperature or with moderate cooling, offering good operational stability even in complex industrial contexts.

Their crystalline structure allows efficient collection of generated charges, reducing dispersion phenomena and improving signal linearity.

The intrinsic efficiency of CdTe and CZT is linked to the combination of high density and average atomic number of their constituents, which guarantees strong attenuation power for X-ray photons.

 

4.2 Lanthanide-doped MHP scintillators:

a Partial diagram of the 4f energy level for trivalent lanthanide activators.

b Main luminescent transitions of lanthanide activators in the electromagnetic spectrum.

c RL spectra of CsPbBr3 CG and CsPbBr3:Eu3+ CG under X-ray excitation.

d Calculation of steady-state light yield of CsPbBr3:Eu3+ CG and commercial LuAG:Ce.

e Absorption spectra, visible emission spectra, and near-infrared emission spectra (λ_ex = 365 nm) of lanthanide-doped and undoped CsPbCl3.

f Comparison of light emission between CsPbCl3:Yb3+ scintillators and various typical scintillators.

 

 

This characteristic makes such materials ideal for applications such as spectral computed tomography, medical imaging, quality control in the electronics and aerospace sectors, and even airport security.

In particular, the ability to discriminate photons based on their energy by analyzing the emitted spectrum allows the acquisition of multi-level images capable of distinguishing different materials even when they have similar density or thickness.

 

  

4.3 Spectral matching factors of various combinations of scintillators/optical sensors.

This capability is fundamental, for example, in verifying the integrity of multi-layer electronic circuits or detecting inclusions in high-precision metal alloys.

From an engineering perspective, CdTe and CZT detectors are available as pixelated or strip devices, coupled with integrated readout circuits (ASICs) that allow signal digitization and subsequent processing.

Using pixel arrays not only enhances spatial resolution but also enables three-dimensional imaging, particularly when integrated into rotating tomographic systems.

Another strength is the possibility to operate in single-photon counting mode with configurable energy thresholds, a feature that allows detailed signal quality control and greater immunity to electronic noise.

Despite numerous advantages, semiconductor detectors also present technical and operational challenges.

The crystalline growth process of the materials, especially for CZT, remains complex and costly, and may introduce imperfections in the detector structure that negatively affect response uniformity and spectral resolution.

Moreover, temperature sensitivity often requires active thermal control systems to ensure signal stability, particularly in industrial environments subject to environmental variations.  

From an application standpoint, CdTe and CZT detectors have opened new frontiers in quantitative imaging, offering the possibility to perform real-time spectroscopic analysis during inspection operations.

This feature is particularly useful for automatic material classification and for detecting hidden defects in composite or layered components.

 

 

4.4 Block diagram of the data acquisition board design principle.

 

Applications in the medical field, such as spectral tomography for soft tissue and vascular contrast evaluation, benefit from the high energy resolution of these materials, enabling more accurate diagnosis and lower patient exposure.

In scientific fields, semiconductor detectors are also employed in synchrotrons and high-energy physics laboratories for high-precision X-ray spectrometry.

The integration of these detectors with artificial intelligence and machine learning technologies represents a further step toward intelligent diagnostic systems.

Automatic segmentation algorithms, pattern recognition, and spectral classification can be used to improve anomaly detection, reduce analysis time, and facilitate operators’ work.

Looking ahead, the miniaturization of semiconductor detectors and the reduction of manufacturing costs could lead to wider dissemination of these devices, even in contexts where currently economic investment is a barrier.

The development of new crystal synthesis techniques, optimization of readout circuits, and standardization of calibration protocols represent the main challenges to be addressed in the coming years to ensure greater diffusion and reliability of these technologies.

4.5 Timeline of UV-visible-NIR perovskite photodetectors

 

In conclusion, direct conversion solid-state detectors represent an advanced and high-performance solution for applications requiring precision, compactness, and spectral capability.

Their impact on non-destructive testing is already evident in multiple sectors, and future prospects point to continuous evolution toward increasingly intelligent, integrable, and high-performing systems, capable of responding to the needs of a constantly evolving industry.

 

  

 

5. Perovskite Sensors:
   New Opportunities for Low-Cost, High-Performance Detectors

 

In recent years, interest in perovskites as sensitive materials for X-ray detection has grown exponentially, thanks to their unique combination of optoelectronic properties, ease of synthesis, and low cost of raw materials.

Perovskites - crystalline structures generally represented by the formula ABX₃, where A and B are cations of different sizes and X is a halide anion - initially emerged in the photovoltaic field but have also demonstrated remarkable potential in detectors for ionizing radiation.

5.1 X-ray semiconductor detectors using various micro/nano-perovskite

materials have shown impressive progress in achieving

higher sensitivity and lower detection limits.

 

What makes these materials particularly promising is their high efficiency in converting X-ray energy into electrical signals, combined with the possibility of being deposited as thin films or grown on flexible substrates, greatly expanding design options for portable and customizable devices.

Perovskite sensors can operate either in direct mode, through the generation of electron-hole pairs following X-ray photon absorption, or in indirect mode, coupled with scintillation layers.

However, it is the direct mode that has attracted the greatest scientific attention, as it enables faster response, superior spatial resolution, and simpler device structure.

Among the most studied perovskites for X-ray imaging are methylammonium lead triiodide (MAPbI₃), cesium lead bromide (CsPbBr₃), and formamidinium lead triiodide (FAPbI₃).

These materials exhibit good charge carrier mobility, strong X-ray attenuation due to the presence of lead, and an appropriate bandgap to ensure signal sensitivity and background noise suppression.

A major advantage offered by perovskites is the ability to realize low-cost detectors using manufacturing processes compatible with inkjet printing, spin-coating deposition, or thermal evaporation—all techniques that do not require complex cleanrooms or high temperatures.

This aspect represents a revolution compared to conventional inorganic semiconductor technologies, whose production costs remain a major barrier to widespread adoption.

Moreover, perovskites’ versatility allows engineering of electronic and structural properties through chemical substitution of A, B, or X components, enabling detector customization for specific energy bands, operating environments, and sensitivity levels.

In terms of performance, perovskite detectors have already demonstrated high sensitivities exceeding 10⁴ μC Gy⁻¹ cm⁻² and extremely low background noise, making them suitable for low-dose medical imaging, food safety, material analysis, and industrial inspections.

 

F5.2 Data table of different scintillators

 

 

The integration of these materials into flexible matrices also opens the way to conformable sensors for complex components like welded joints or curved surfaces, which are difficult to inspect with rigid devices.

Promising applications are also emerging in aerospace and military sectors due to the lightweight nature of the sensors and their ability to withstand severe environmental conditions.

However, perovskite technology is not without its limitations and challenges.

One major obstacle concerns the chemical and thermal stability of the materials, especially in the presence of humidity and oxygen, which can degrade the crystal and reduce efficiency.

Various approaches have been explored to mitigate these issues, including encapsulation in polymeric materials, the adoption of fully inorganic perovskites (such as CsPbBr₃), and surface modification through chemical passivation.

Reproducibility of large-scale synthesis also remains an open challenge, as crystal quality crucially influences charge collection properties and signal linearity.

Another critical aspect is the toxicity of lead, a key element for radiative performance but subject to environmental restrictions.

This has driven research toward alternatives based on tin, bismuth, or antimony, which, while promising, require further optimization to reach the sensitivity and stability of lead-based counterparts.

Interdisciplinary collaborations among materials chemists, condensed matter physicists, and electronic engineers are accelerating progress in this field, with numerous publications reporting incremental improvements in device performance and durability.

5.3 vdW 2D/3D heterostructures and their potential functional applications for photodetection

 

 

Perovskites also offer interesting opportunities in advanced spectral detection.

Recent studies have shown that it is possible to engineer structural parameters to achieve energy-selective response, potentially useful for realizing multi-energy detectors capable of distinguishing materials with similar absorption but different chemical composition.

Their fast temporal response further opens the possibility of realizing high-frequency dynamic imaging systems, with applications in industrial real-time monitoring of critical production processes.

The combination of perovskites with advanced digital techniques, such as processing via artificial intelligence algorithms, enables the construction of adaptive and intelligent systems capable of real-time acquisition parameter adjustment based on defect type or observed material.

This synergy between material and software represents one of the main strengths of the new generation of detectors, envisioned not only as measurement tools but also as automatic data interpretation instruments.

In summary, perovskite sensors represent a highly interesting technological frontier for X-ray non-destructive testing.

They offer an ideal balance between production costs, performance, and design flexibility, while still presenting significant challenges related to stability and environmental sustainability.

With ongoing advances in materials chemistry, encapsulation techniques, and fabrication processes, it is plausible to envision that in the coming years perovskites will become one of the dominant technologies in radiographic detection.

Their ability to adapt to specific application needs, their economical scalability, and integration in intelligent electronic devices make them ideal candidates for a new generation of industrial sensors, driving non-destructive testing toward increasingly precise, accessible, and sustainable dimensions.

 

 

6. Photon Counting X-ray Detectors (PCXDs):
   Photon Counting and Precise Spectral Analysis

 

Photon Counting X-ray Detectors (PCXDs) represent one of the most advanced developments in the field of X-ray detection, introducing a paradigm shift compared to traditional integration systems.

Unlike conventional detectors, which measure the total amount of energy deposited by a photon beam over a period of time, PCXDs are capable of detecting and counting each individual incident photon, also recording its individual energy.

6.1 PILATUS 6M was the first large-scale HPC detector routinely used at a synchrotron beamline.

 

This approach enables intrinsically higher spectral resolution, a significant reduction in electronic noise, and an overall improvement in image quality and material discrimination capabilities.

PCXDs operate based on the use of semiconductor materials with high energy conversion efficiency, such as CdTe (cadmium telluride) or CZT (cadmium zinc telluride), coupled with high-speed readout electronics able to discriminate the energy of each photon.

When a photon interacts with the sensitive material, it generates a cloud of charge carriers collected by single or multiple pixel electrodes.

 

6.2 Schematic representation of a sensor readout hybrid used in HPC detectors.

Indium is not the only material used for electrical connections between sensor pixels and readout pixels.

These signals are immediately amplified, filtered, and analyzed by ASICs (Application Specific Integrated Circuits), which digitize and count the events, assigning them to specific energy windows.

This mechanism allows real-time construction of energy histograms, providing detailed information about the composition of the material traversed by the X-rays.

One of the main applications of PCXDs is spectral or multi-energy computed tomography (CT), where the ability to distinguish different energy levels enables signal decomposition based on the sample’s constituent elements.

This approach has proven particularly effective in discriminating materials with similar densities but different chemical compositions, such as separating plastics from light metals in security inspection systems, or characterizing multiple phases in complex metal alloys.

In the medical field, PCXDs allow greater accuracy in bone density measurement, soft tissue differentiation, and reduction of the patient’s dose due to increased detection efficiency.

The implementation of PCXDs also offers numerous operational advantages. First, direct counting eliminates the need for high-resolution analog-to-digital converters in the acquisition chain, reducing complexity and noise.

Moreover, PCXD systems provide a superior dynamic range, as counting is less prone to saturation than integrative systems.

This feature is especially valuable in high-flux environments, such as industrial plants where exposure times must be minimized to avoid productivity losses.

6.3 Side view of the hybrid sensor readout system.

The polarity of the electric field (indicated by three parallel arrows) separating the charge generated by photon absorption

depends on the sensor material requirements and readout electronic.

Other benefits include advanced noise correction algorithms and the ability to perform real-time spectral segmentation, features that make these devices ideal for integration into automated and intelligent analysis systems.

Despite their many strengths, PCXDs also face challenges and limitations.

The primary significant obstacle is the high cost of high-efficiency semiconductor materials and dedicated ASIC circuits, potentially limiting widespread adoption in budget-constrained sectors.

Furthermore, high-speed readout electronics require active cooling and careful design to avoid cross-talk, saturation, and pile-up—i.e., the overlap of multiple events within a time interval shorter than the system’s temporal resolution.

These effects can compromise energy response linearity and must be carefully managed through correction algorithms and optimized pixel design.

6.4 Illustration of the operating principles in a single pixel between direct and indirect conversion cameras.

The photon counting detection principle eliminates all other noise sources present in CCD or flat-panel cameras. This results in a significantly improved signal-to-noise ratio and consequently the ability to detect greater image detail.

Image sharpness, or the effective spatial resolution of the acquired image, is defined by the electric charge in the CMOS readout.

Although the pixel size of direct conversion cameras is larger than conventional indirect conversion cameras, the detected X-ray signal is better focused within the pixels.

Typical direct-conversion pixel sizes range from a few millimeters down to tens of micrometers, with ADVACAM representing the highest pixel density among current industrial X-ray cameras, with a pixel size of 55 µm.

From a technological standpoint, one of the most interesting developments is the miniaturization of readout circuits and the integration of intelligent functionalities directly at the pixel level.

This approach, known as pixel-level intelligence, enables dynamic thresholds, adaptive digital filters, and even on-chip neural networks for real-time signal classification.

AI integrated into PCXDs is opening new horizons in predictive diagnostics and automatic defect identification, allowing systems to learn from previous images and progressively improve recognition capabilities.

PCXDs are also compatible with two-dimensional or three-dimensional matrix configurations, facilitating the construction of scalable modular systems for high spatial and volumetric resolution applications.

These detector arrays are particularly suited for use in synchrotron beamlines, non-destructive analysis systems for aircraft turbines, compressor blades, and critical welds, where sub-micrometer resolution and elemental analysis capability are essentials.

6.5 CMOS and CCD 2D detectors.

HPC detectors, by contrast, directly convert photons into electric charge and count them one by one immediately. Furthermore, they are event-based, meaning they simply count each single photon without accumulating noise. They also incorporate energy discrimination to prevent any possible charge sharing between adjacent pixels, ensuring maximum precision.

 

In industrial contexts, PCXDs are increasingly used in quality verification of components in additive manufacturing (AM), where it is essential to detect internal defects, porosity, or inclusions that could compromise structural integrity.

Fast acquisition and high energy sensitivity allow inline inspections without slowing production.

Similarly, in the food sector, PCXDs are used to identify low-density contaminants such as plastic, glass, or bone fragments in packaged products, thanks to their fine discrimination of absorption energies.

The future evolution of PCXDs will likely focus on greater electronic integration, adoption of more sustainable alternative materials, and expansion of the operational energy range to include high-energy photons up to the MeV range.

Additionally, the development of automatic calibration techniques, self-diagnostics, and predictive maintenance will contribute to making these devices even more reliable and easier to integrate into complex systems.

With the support of edge processing and distributed artificial intelligence, photon counting detectors can evolve into adaptive and cooperative sensors capable of sharing data and learning in networks, leading non-destructive testing into a new era of precision, efficiency, and autonomy.

6.6 Photon counting detectors

 

In conclusion, PCXDs represent a disruptive innovation in X-ray detection, offering unparalleled advantages in information quality, operational flexibility, and automation potential.

Their growing adoption across medicine, industry, and research confirms their key role in the evolution of non-destructive testing, positioning them as a benchmark for the design of future intelligent imaging and quantitative material analysis systems.

 

 

 

 

7. Multi-Energy Imaging and Spectral Tomography:

    Material Discrimination with Advanced Detection

The introduction of multi-energy imaging and spectral tomography has marked a turning point in the field of X-ray non-destructive testing (NDT), offering new capabilities for material discrimination through detailed energy analysis of the radiation beam.

This approach is based on the principle that different materials absorb X-ray photon energy differently depending on their atomic composition and electron density.

7.1 Schematic illustration of the three X-ray imaging technologies:

(A) Single includes one scintillator layer and a grayscale image sensor with energy integration characteristics.

(B) Dual includes two scintillator layers, two grayscale image sensors, and a metallic energy filter with energy integration and energy separation features.

(C) Multi includes three scintillator layers, two optical/energy filters, an energy filter, and a color image sensor with energy integration, energy separation, and spectral information acquisition capabilities. Trichromatic vision (red, green, blue) is produced by three individually stacked scintillator layers (DE1 layer, DE2 layer, E layer). Filter 1 (long-pass 600 nm) and Filter 2 (long-pass 500 nm) block extra photons to avoid spectral overlap and synergistically harden the X-ray beam with Filter 3 (CaF2).

 

The implementation of multi-energy imaging systems thus enables separation, identification, and quantification of the materials present in an object through their energy signature, overcoming the limitations of conventional radiography, which provides only integrated absorption information.

The technical foundation of spectral tomography is the use of detectors capable of distinguishing photons of different incident energies.

This can be achieved through various strategies: sequential acquisitions at different energies, interchangeable spectral filters, or more recently, through the use of photon counting detectors with energy discrimination (PCXDs).

In dual-energy CT (DECT), the simplest configuration, two scans or simultaneous readings at different energies are performed, enabling the generation of compositional maps and separation between high- and low-density materials.

7.2 Conceptual multi-energy X-ray imaging technology for baggage inspection:

(A) Illustration of multi-energy X-ray imaging technology for baggage inspection.

(B) Sketch of simulated baggage containing eight different objects (items a - h).

(C & D) Original multi-energy radiographic images (C) and color-reconstructed images (D) of the simulated baggage.

(E–G) Three regions of interest (ROI) with different emission wavelength ranges from the original multi-energy X-ray image of the simulated baggage.

(H) Modulated transfer function curves of the telescope scintillator under low, medium, and high X-ray tube voltages.

(I–K) X-ray images of a standard line pair phantom (type 39b) captured with the telescope scintillator at (I) low, (J) medium, and (K) high kilovoltages.

More advanced technologies employ spectral detectors with more than two energy channels (multi-energy CT), improving chemical resolution and enabling quantitative decomposition in terms of constituent elements, electron density, or effective atomic number.

A main application of multi-energy imaging in industrial NDT is the discrimination of phases within composite materials or metal alloys.

7.3 Schematic illustration of the multi-energy X-ray detection principle.

 

For example, in aircraft turbines or components produced by additive manufacturing, it is crucial to identify inclusions, segregations, or porosity that could compromise mechanical strength.

With multi-energy analysis, it is possible to obtain three-dimensional maps of material distribution, locating anomalies with accuracy superior to conventional CT.

Similarly, in the automotive sector, this technique is employed to verify the integrity of welds, joints, and plastic-metal bonded materials, enhancing vehicle quality and safety.

In security, multi-energy imaging is widely used in airport and customs inspection systems, allowing the identification of organic materials, explosives, and hazardous substances based on their energy signature.

Unlike single-energy systems, which often fail to distinguish materials with similar densities, spectral analysis enables more reliable automatic classification and a significant reduction in false positives.

Furthermore, multi-energy detectors can be integrated into intelligent systems equipped with deep learning algorithms, capable of learning threat patterns and dynamically adapting detection thresholds.

From an instrumental design perspective, multi-energy systems require rigorous calibration, both in terms of detector energy response and volumetric data reconstruction.

Spectral reconstructions demand complex decomposition algorithms, which may include iterative optimization, nonlinear regression, and Bayesian methods for estimating elemental concentrations.

The combined use of spectral and structural data also enables quantitative imaging, where CT information is no longer purely morphological but also chemophysical, with significant impacts on material characterization.

Technically, a major challenge in spectral imaging is the management of noise and artifacts caused by low photon statistics in each energy channel.

Increasing spectral resolution implies dividing the photon flux into narrower bands, resulting in reduced signal in each band.

To overcome this, high-intensity sources, prolonged acquisition times, or reconstruction algorithms based on denoising and spectral compression techniques are employed.

Other problems include cross-channel calibration, spectral beam hardening correction, and management of detector response nonlinearities.

Looking to the future, the trend is toward greater integration of multi-energy detectors with on-chip processing and artificial intelligence.

Next-generation spectral imaging systems will likely feature real-time automatic classification, material recognition, and predictive anomaly analysis, thanks to convolutional neural networks and supervised learning algorithms.

Additionally, miniaturization of detector modules and the realization of three-dimensional spectral arrays will enable development of compact high-performance scanners, usable even in mobile or field environments.

Finally, multi-energy imaging opens new possibilities for in-service monitoring of materials and components via dynamic high-spectral-resolution imaging techniques.

In combination with actuation technologies such as thermal induction or ultrasonic vibration, it is possible to observe in real time structural and chemical changes within a component under stress, improving understanding of degradation mechanisms and facilitating predictive maintenance.

In summary, spectral tomography and multi-energy imaging represent powerful extensions of traditional radiographic techniques, paving the way for a new generation of non-destructive testing that provides not only images but also quantitative data on material composition.

With advances in detectors, increasing computational power, and integration of artificial intelligence, these technologies will become increasingly central in advanced inspection, industrial quality, and critical safety practices.

8. Advanced Applications of X-ray Detectors:

    Synchrotrons, Medicine, Security, and Beyond

The advanced use of X-ray detectors extends well beyond conventional industrial applications, finding wide employment in highly specialized fields such as scientific research at synchrotrons, medical diagnostic imaging, airport security, and space monitoring.

In each of these areas, requirements in terms of spatial resolution, energy sensitivity, response time, and material discrimination capacity push detectors toward ever higher performance, while posing complex engineering and physical challenges.

8.1 Diagram of multi-energy and energy-integrated X-ray imaging systems:

(a) Diagram of a conventional energy-integrated X-ray imaging system.

(b) Diagram of a large-area, flat-panel multi-energy X-ray imaging system based on stacked multilayer scintillators.

Synchrotrons, in particular, represent one of the most demanding contexts for X-ray detection systems, thanks to the production of highly brilliant, collimated, and monochromatic X-ray beams.

In these environments, detectors are used for diffraction, absorption, and coherent scattering experiments, requiring temporal resolutions on the order of nanoseconds and high dynamic range readout capabilities.

 

8.2 LAMBDA detector module. A 42 mm × 28 mm Cr-compensated GaAs sensor,

connected to 6 Medipix3 chips, mounted on the left side of the detector head.

 

 Among the most common devices at synchrotrons are active pixel detectors (APDs), solid-state photon counting detectors with hybrid architectures (such as DECTRIS PILATUS and EIGER), as well as silicon strip or pixel sensor systems with parallel readout.

These instruments are designed to acquire large data volumes in very short times, with minimal electronic noise and high resistance to ionizing radiation.

Applications include crystallographic studies, ultra-high-resolution tomography, and elemental imaging via X-ray spectroscopy, where it is essential to separate low-intensity signals amid strong photon gradients.

8.3 Raman spectrometer at ID20.

Left: spectrometer portal.

Right inset: single-chip MAXIPIX detector module.

Recent advances include the introduction of time-of-flight detectors and devices integrated with sensor-level processing electronics, enabling real-time analysis of photon-matter interactions.

In medicine, the use of X-ray detectors has undergone a revolution with the advent of spectral computed tomography (CT), digital mammography, and low-dose fluoroscopy.

Photon counting detectors with energy discrimination are now adopted in new generations of clinical scanners, allowing precise tissue segmentation, reduction of patient doses, and early identification of vascular or tumor pathologies.

Multi-energy analysis capability in medical detectors offers significant advantages in contrast visualization, bone density quantification, and advanced cardiovascular imaging.

The future of diagnostic radiology targets hybrid systems that integrate spectral capabilities with artificial intelligence algorithms for automated diagnosis and longitudinal patient monitoring.

In the security sector, both civil and military, X-ray detectors are used in airport screening systems, vehicle and container inspection portals, and mobile detectors for ordnance disposal or nuclear smuggling prevention.

In these contexts, the goal is to discriminate potentially dangerous materials with high reliability and speed.

Modern systems adopt dual- or multi-energy detectors combined with automatic recognition algorithms and artificial intelligence, capable of distinguishing explosives, chemicals, drugs, and fissile materials.

Some devices are also equipped with backscatter X-ray systems, which provide images of the external surface of inspected objects, or portable X-ray tomography for 3D field inspection.

 

8.4 Synchrotron facility

In space applications, X-ray detection is crucial for scientific missions studying high-energy cosmic phenomena such as black holes, neutron stars, and solar flares.

Space missions require detectors with reduced weight, high quantum efficiency, and the ability to operate under extreme temperature and radiation conditions.

Semiconductor detectors such as CdZnTe and silicon microstrip detectors have been employed on various X-ray observation satellites launched by NASA, ESA, and JAXA.

Some detectors are designed to operate in spectroscopic mode, while others prioritize temporal resolution for studying transient events.

Integration with advanced X-ray optics and operation in polarimetric or interferometric configurations broaden the frontiers of astrophysical research.

In cultural heritage, X-ray detectors are playing an increasing role thanks to their ability to investigate internal structures invisible to the eye and determine elemental composition.

Techniques such as X-ray fluorescence (XRF), digital radiography, and micro-CT tomography are employed to analyze paintings, archaeological artifacts, and historic objects without causing damage.

Detectors used in these cases must combine high sensitivity, good spatial and energy resolution, and compatibility with portable systems for onsite operation in museums or archaeological sites.

The growing interest in X-ray imaging for cultural heritage has led to the development of hybrid instruments that integrate X-ray detection with optical imaging, infrared, and Raman spectroscopy, offering a multimodal view of the artifact.

Finally, the electronics and semiconductor industry uses X-ray detectors for quality control, assembly verification, and material characterization in miniaturized devices.

Micro-CT combined with high-resolution detectors and 3D analysis software allows detection of defects like voids, cracks, delaminations, and shorts in chips and integrated circuits.

This type of analysis is fundamental in producing electronic components for automotive, aerospace, and medical sectors, where reliability is indispensable.

New trends include the use of flat-panel detectors (FPD), advanced CMOS sensors, and fast acquisition technologies to minimize scanning times and maximize productivity.

In summary, X-ray detectors today are present in an extraordinarily wide range of advanced applications, spanning fundamental research to industry, medicine to security, space to cultural heritage.

Their continuous technological evolution, driven by ever more stringent performance and reliability requirements, is significantly contributing to innovation and safety across numerous strategic sectors.

The next step will be systematic integration with artificial intelligence platforms, neural networks, and edge communication technologies, which will make X-ray detectors intelligent, adaptive, and predictive tools ready to face future scientific and industrial challenges.

9. Future Horizons in X-ray Detection:

   Nanotechnologies and Artificial Intelligence

The evolution of X-ray detectors, driven by demands for resolution, speed, sensitivity, and spectral capabilities, is entering a phase of radical transformation thanks to emerging nanotechnologies and integration with artificial intelligence (AI).

While conventional detectors are based on established principles such as ionization and scintillation, next-generation devices exploit quantum phenomena and extreme properties of matter to achieve levels of precision and versatility previously unimaginable.

Among the most promising technologies are Superconducting Nanowire Single-Photon Detectors (SNSPDs), superconducting nanowire detectors capable of detecting individual X-ray photons with near 100% efficiency, virtually zero noise, and temporal resolution on the order of picoseconds.

9-1 Overview of the X-ray detector.

 Schematic structure of the device.

b Comparison of the performance of current solid-state X-ray detectors.

Operating voltage is indicated next to each data point.

Values of the total attenuation coefficient for carbon, selenium, lead iodide and methylammonium (MAPbI₃), and Bi₂O₃ are indicated as shaded areas showing prior limits of detector technology based solely on mass attenuation processes.

c X-ray imager based on a hybrid X-ray detector and a 70 kV X-ray image of a bolt acquired using the X-ray imager.

 

These devices consist of extremely thin superconducting wires arranged in a serpentine pattern, maintained at cryogenic temperatures (generally below 2 K) to ensure superconducting operation.

When a photon strikes the nanowire, it induces a local transition from the superconducting state to the normal state, causing a measurable current drop as a signal.

The fast response, combined with spectral sensitivity and high noise rejection, makes SNSPDs ideal tools for applications requiring ultrafast timing, such as dynamic process imaging, time-resolved X-ray spectroscopy, and quantum detection.

However, the technological challenges associated with these devices are substantial.

In addition to the need for extremely low cryogenic temperatures, they require shielded environments, dedicated readout electronics, and nanostructured high-purity materials.

Large-scale production remains an obstacle to commercial diffusion, although developments are underway to create larger arrays compatible with conventional imaging systems.

9.2 Performance of a soft X-ray detector based on SnS:

a) Schematic representation of a two-electrode device exposed to soft X-rays. The drain current as a function of soft X-ray photon energies can be obtained using the Keysight Technologies source unit.

b) Device response to different photon energies.

c) Comparison of Ids response in the water window region and for energies above 1 keV; indicating excellent response around 600 eV.

d) IV curves at photon energies between 100 eV and 1 keV. Maximum photocurrent observed near 600 eV, followed by rapid decrease with increasing photon energy.

e) Drain current Ids as a function of dose rate at various bias voltages. Sensitivity values are extracted from the slope of these photocurrents.

f) Sensitivity values at different bias voltages show a linear increase with increasing bias voltage. Dose rates for photon energies above 600 eV are not included.

g) Time-dependent Ids responses at a fixed bias voltage of 1 V and 600 eV for various slit sizes. These slit sizes are used to control photon flux values at the Australian Synchrotron.

h) Temporal response of the device at 600 eV and Vb = 1 V; indicating rise (90% of saturated X-ray signal) and fall times (10% of saturated X-ray signal) of 7 and 2 ms, respectively. This is far superior to other reported soft ferromagnetic and halide perovskite X-ray detectors.

 

10. Conclusions and Perspectives:

      The Crucial Role of X-ray Detectors in NDT Innovation

X-ray detection represents one of the key technologies for the evolution of non-destructive testing (NDT), offering indispensable tools for internal material characterization, early defect diagnosis, and real-time monitoring of complex structures.

In this article, we have traced a path from the simplest and most traditional detectors, such as radiographic films, gas chambers, and indirect scintillators, to more advanced solutions based on semiconductors, photon counting, spectral imaging, and multi-energy detectors.

 

10.1 VUV photodetectors covering a wide range of applications:

(A–D) Main applications of VUV photodetectors, including VUV solar radiation monitoring (space science) (A), dark matter detection (high-energy physics) (B), VUV FEL radiation diagnostics (large-scale scientific facilities) (C), and high-resolution lithography (electronics industry) (D).

(E) Traditional VUV photodetectors currently available are shown under the electromagnetic spectrum. Device structures-including scintillator, photomultiplier tube, silicon diode, silicon photomultiplier, ultra-wide band semiconductor photodetector, and gas detector-are illustrated at the bottom from left to right. Note that the scintillator shape is somewhat arbitrary.

 

Each technology has brought with it an evolution in analytical capabilities, greater efficiency in defect detection, reduced doses and acquisition times, and improved material discrimination.

The introduction of solid-state sensors such as CdTe and CZT paved the way for direct detection with high energy resolution, reducing system complexity and enhancing stability.

Perovskites, as emerging materials, offer a promising route toward more economical and scalable devices. Photon counting detectors have radically changed the detection paradigm, enabling quantitative photon information acquisition and real-time multi-energy analysis.

 

10.2 Schematic diagram of the synergy between deep learning and metafotonics,

holography, and quantum photonics, driving the development of intelligent photonics.

The integration between AI technology and photonics represents the intersection of the digital and physical worlds.

 

The use of discrete neural networks (DNNs) in direct modeling and inverse design processes can significantly enhance efficiency and accuracy in addressing photonic problems.

Conversely, photonics can be exploited for the physical implementation of AI computing, for example through integrated photonic circuits and optical neural networks (ONNs) in free space.

Spectral and multi-energy technologies now enable obtaining elemental and structural information that previously required invasive or destructive techniques.

This has revolutionized fields such as medicine, security, and electronics industry, and continues to push the frontier of scientific applications in synchrotrons and high-energy physics laboratories.

Future prospects, as we have seen, point toward increasingly tighter integration of advanced hardware and intelligent algorithms.

Nanotechnologies allow miniaturization and customization of sensors, while artificial intelligence enables automated analysis, improved diagnostic precision, and faster response times.

In this scenario, X-ray detectors will become intelligent nodes in monitoring and control networks, capable of adapting to operating conditions and interacting with complex systems predictively and autonomously.

For non-destructive testing professionals, this implies a necessary evolution of skills, which must embrace not only physical knowledge of detectors, but also mastery of data processing tools, artificial intelligence, and numerical simulation.

Future challenges will include managing large data volumes, securing intelligent systems, standardizing techniques, and ensuring interoperability among different devices.

 

10.3 Artificial intelligence for X-ray photon counting technology

 

However, the expected benefits are considerable: greater inspection accuracy, reduced operational costs, improved safety of critical components, and opening toward new fields of application.

Ultimately, X-ray detectors are destined to remain at the core of technological innovation in non-destructive testing.

Their evolution is not only a technical matter but represents a cultural transformation in how we interpret matter, monitor processes, and make decisions.

10.4 Intelligent photonics applications across various fields, including the metaverse,

biomedicine, autonomous driving, advanced manufacturing,

optical communications, and astronomical observation.

 

Investing in research, training, and adoption of these technologies means ensuring greater safety, efficiency, and competitiveness in the most strategic sectors of our industrial and scientific society.