Open Access
Add to BibSonomy
Abstract
A structured double-period CsI scintillation screen was successfully developed to improve its detection efficiency based on an oxidized silicon micropore array template with a period value on the order of micro-scale. The structure comprises a main structure along with a sub-structure. The main structure with a period of 8 µm was arranged in a square array consisting of square columnar scintillator units. The micropore walls between the main structure units were purposely fabricated from a SiO2-Si-SiO2 layered structure. The pore walls in commonly used single-structure with a period of 4 µm use the same layered structure composition to obtain a fair comparison. The thickness of both Si and the SiO2 layers was around 0.4 µm. The unique feature of the double structure lies in the even separation of each unit within the main structure into four square columnar scintillator sub-units. These four sub-units within each sub-structure were isolated solely by SiO2 layers with a thickness of approximately 0.8 µm. As a result, the X-ray-induced optical luminescence intensity of the double-structure screen exhibited a 31% increase compared to the corresponding single-structure scintillation screen. In X-ray imaging, a spatial resolution of 109 lp/mm was achieved, which closely matched the results obtained with the single-structure CsI screen. Furthermore, the detective quantum efficiency also displayed a notable improvement.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The scintillation screen, as a key component of indirect X-ray imaging system, is utilized for converting X-ray images into UV-Vis images. It has widely been used in basic scientific research, medical imaging, nondestructive testing, and astronomical observations [1,2]. Typically, a scintillation screen needs to have a specific thickness to achieve optimal detective efficiency. Nevertheless, increasing the screen thickness results in a decrease in spatial resolution in X-ray imaging due to the lateral scattering of scintillation photons [3,4]. To address the issue, a scintillation screen featuring a micro-columnar structure was proposed. The columns in this design were arranged vertically to the surface of the screen to direct the propagation of the scintillation light along the microcolumn channels, thereby limiting its lateral spread. Among the various kinds of structured scintillation screens, using an oxidized silicon micropore array template as the foundation for the screen offers distinct advantages [5–7]. Typically, this kind of scintillation screen is created by injecting the scintillation melt into the template pores. The silicon pore walls serve to restrict the lateral transmission of scintillation light through light absorption. Additionally, the silicon dioxide (SiO2) layers covering the silicon walls can be harnessed to guide the scintillation light at an angle exceeding the critical angle, effectively directing it along the columnar channels [6,8]. It can, therefore, be used to achieve both good optical isolation as well as good light guidance.
Cesium iodide (CsI) is a conventional material for fabricating scintillation screens due to its significant light yield, suitable density, and fast decay time [9,10].In addition, its low melting point and high refractive index (higher than that of SiO2) render it highly suitable for the development of such structured scintillation screens [7,8,11]. Due to rapid technological advancements, structured scintillation screens are increasingly adopting smaller periods to achieve enhanced X-ray imaging resolutions [5,12]. The structured CsI and CsI (Tl) scintillation screen with a period of 4 µm have already been developed successfully [6,13,14]. The X-ray imaging spatial resolution that has been attained using these screens has surpassed 100 line pairs per millimeter (lp/mm). However, it is essential to restrict the thickness of the pore walls when dealing with scintillation screen structures at the micron scale. Failure to do so can result in an insufficient ratio of the CsI filler area to the CsI screen area, consequently decreasing detection efficiency [15]. Nonetheless, the pore wall is required to have a certain thickness, typically not less than 1 µm, to ensure proper optical isolation between adjacent CsI columns and light transmission along the CsI columns. Considering the optical tunneling effect and the optical absorption coefficient of Si [16], it was speculated that a structure consisting of a 0.4 µm-thick Si layer and a 0.3-0.4 µm thick SiO2-covered layer for the pore wall on either side would provide a fairly reasonable arrangement. In this scenario, it is crucial to focus on methods for improving the detection efficiency of the structured scintillation screen while preserving the X-ray imaging spatial resolution.
A scintillation screen with a double-period structure was proposed in an attempt to achieve both good X-ray detection efficiency, as well as good spatial resolution [17]. This new structure featured a primary framework and a sub-structure. The main framework was made up of columnar scintillation units arranged in a square array. Each primary unit was separated by micropore walls consisting of a SiO2-Si-SiO2 layered structure. This composition was deliberately selected for a fair comparison with the single-structure, which also utilizes the same composition for its micropore walls. The distinctive aspect of the double-structure is that each primary structural unit is evenly divided into multiple columnar scintillator sub-units, and these sub-units are separated solely by the SiO2 layer to create a sub-structure. The simulation results showed that the X-ray imaging efficiency was higher when using a scintillation screen with a double structure compared to a screen with a single structure. However, the scintillation screen with a double-period structure is still only a theoretical concept and requires experimental development. Simultaneously, the practical implementation of the double-structure scintillation screen raises questions about whether its X-ray imaging performance, encompassing spatial resolution and detective efficiency, can indeed attain the theoretically projected values. Consequently, further investigations are required to explore these aspects, both theoretically and experimentally.
This study describes the preparation of CsI scintillation screens with a double-period structure based on oxidized silicon micropore array templates. The X-ray excited optical luminescence (XEOL) properties, crystallinity, morphology, and X-ray imaging performances of the structured CsI scintillation screens were investigated in detail. Furthermore, simulations were conducted to assess the light output (LO) and modulation transfer function (MTF), which are indicative of the light-guiding and optical isolation capabilities. The findings in this study conclusively established that the successful development of the CsI scintillation screen with a double-period structure was not only possible, but its performance also aligned with the theoretical expectations. Consequently, this research offers a viable strategy for enhancing the detection efficiency of structured scintillation screens designed for high X-ray imaging resolution.
2. Structural design and simulation methods
2.1 Structural design of scintillation screens
The schematic diagrams for the CsI scintillation screens with single-period and double-period structures, as well as the propagation process of scintillation light within the respective structured screens have been presented in Fig. 1. As shown in Fig. 1(a), a single-structured scintillation screen was formed by arranging CsI square-columns in the form of a square array with a period of 4 µm. The micropore walls comprised a SiO2-Si-SiO2 layered structure. A double-structured scintillation screen consisting of a main structure and a sub-structure, was designed in this study to obtain both a high X-ray detective efficiency and a good spatial resolution, as depicted in Fig. 1(b). The units in the primary structure were formed in a square array with a period of 8 µm. The micropore walls within the main structure units featured a layered composition of SiO2-Si-SiO2, which closely matched the structure of the pore walls in the single-structure screen. The four sub-structure units created by equally dividing each of the main units were also arranged in a square array with an average period of 4 µm. In this case, the micropore walls between the sub-structure units were exclusively composed of SiO2. The scintillator area ratio within the double-structure screen was larger in this case in comparison to the single-structure screen, owing to thinner pore walls between the sub-structures, which can lead to growth in the number of produced scintillation photons. The schematic diagrams illustrating the light propagation process in single-structure and double-structure scintillation screens have been presented in Fig. 1(c) and Fig. 1(d), respectively. Following the interaction of X-rays with the CsI scintillator, the emitted scintillation light exhibited asymmetric dispersion. Scintillation photons with incident angles equal to or greater than the critical angle experienced complete reflection at the CsI-SiO2 interface due to total internal reflection. However, only a fraction of the scintillation light beams, entering at angles below the critical angle, were reflected, while the remaining photons refracted into the SiO2 layer. Within the SiO2-Si interface, a portion of the refracted light that entered the SiO2 layer was reflected, while other photons were absorbed into the Si layer. Intriguingly, no absorption occurred for refracted light entering the SiO2 layer in the pore walls between the sub-structural units. A portion of the refracted light was reflected at the SiO2-CsI interface, whereas the remaining portion was transmitted through the SiO2 layer into the adjacent CsI columns within its main structure. The transmitted photons continued to undergo reflection and refraction. A portion of these scintillation photons continued to travel along the CsI columns, resulting in a further enhancement of the light emission from the double-structured scintillation screen. The rest of the scintillating light was absorbed by the Si layers between the main structure units of the double-structure screen.
Fig. 1. Schematic diagrams for the CsI scintillation screens and illustration of the light propagation process within them. (a) and (c) represent single-period structure. (b) and (d) depict double-period structure.
2.2 Simulation methods
The simulation also considered the array structure, size, and shape of the CsI columns used in the development of the scintillation screen to calculate the light output (LO) and modulation transfer function (MTF) that can respectively reflect the effects of light guiding and optical isolation. In the X-ray imaging process, the scintillator screen interacts with X-rays to generate scintillation light, and these scintillation photons propagate within the screen. The simulation of the first part of this process was conducted using the Geant4 Monte Carlo simulation toolkit (version 10.0). It accounted for various physical phenomena, including de-excitation effects, ionization, multiple scattering, bremsstrahlung radiation, Compton scattering, the photoelectric effect, and coherent scattering [18,19]. The second process which was simulated utilizing the Monte Carlo ray-tracing technique [20] and the finite-difference time-domain (FDTD) method included light reflection, refraction and absorption before scintillating light reached the bottom of the screen [21]. The FDTD method was primarily employed for simulating the scintillation light at the CsI-SiO2 and SiO2-Si interfaces since proximity between the scintillation light wavelength and the thicknesses of the Si and SiO2 layers required consideration of the wave nature of the scintillation light.
The values for scintillation light yield, refractive index, and optical attenuation coefficient of the CsI scintillator were obtained from previous literature [10,22]. The refractive index and optical attenuation coefficient values for Si and SiO2 were also taken from prior studies [23–26].The procedural details of the simulation were also in accordance with a previously reported work [27]. The energy and dose of the incident X-rays were adjusted following the specifications of the imaging experiment. The dimensions of the scintillation screens used in the study were set to 0.5 mm × 0.5 mm. To ensure high simulation accuracy, the relative ray power threshold in the Monte Carlo ray-tracing technique and the mesh size in the FDTD method were configured to 1% and 1/34 of the wavelength, respectively.
3. Experimental and characterization
3.1 Sample preparation
The preparation process for the structured scintillation screen is depicted in Fig. 2. Initially, guide pits, arranged in a double-period square array, were patterned onto p-type (100) silicon substrates with a resistivity of approximately 30–50 Ω·cm using photolithography. These guide pits were subsequently etched using the inductively coupled plasma (ICP) method, employing a suitable gas mixture of SF6 and CHF3 [28]. The shape and size of the pits were then modified by varying the alkaline etching time of the samples in a solution of 25wt% tetramethylammonium hydroxide (TMAH) at a temperature of 87°C [29]. The arrays of vacant micropores in the silicon wafer were fabricated using the electrochemical etching method in hydrofluoric acid solution based on the guide pits [30]. The silicon micropore arrays were thermally oxidized in a muffle furnace at a temperature of 1000°C for a dwell time of 8 h for the growth of SiO2 total reflective layers on the surface of the micropore walls. The CsI scintillation screens were produced by introducing a highly pure CsI melt (99.999%) into the oxidized Si micropore array templates. This was accomplished through a process involving vacuum melting and gas pressure injection using a custom-made device. Following this, any excess CsI exsiting on the template surfaces was abradeed through polishing. Furthermore, a scintillation screen with a single-period structure was prepared for comparison. The CsI microcolumns in the single-period structure were also configured in a square array, with a period that aligned with the average period of the sub-structure in the double-period scintillation screen. The procedure for creating CsI-based scintillation screens with a single-period structure closely mirrored that of the double-period structure screen, with the only distinction being the utilization of different masks during the photolithography process.
Fig. 2. Flow diagram depicting the sequence of fabricating processes for developing the structured scintillation screens.
3.2 Characterization
Morphologies of the templates based on oxidized Si micropore arrays and the scintillation screens were measured using a Gemini 300 field emission scanning electron microscope (FE-SEM, ZEISS). The distribution of elements including oxygen and silicon in the templates was obtained using an Xplore 30 energy dispersive X-ray spectrometer (EDS, OXFORD). The crystalline structure of the samples was analyzed using a DX-2700 X-ray diffractometer (XRD, Haoyuan Instrument Co., Ltd.), which utilized Cu Kα radiation (λ = 0.15405 nm). The XRD instrument was operated at a voltage of 40 kV and a current of 30 mA. For recording the XEOL spectra, a custom-made X-ray excited spectrometer was employed. The excitation source was an F30-III X-ray tube from Shanghai Nucl. Med. Instrum. Co., Ltd., which had a tungsten anode and was operated at 80 kV with a current of 4 mA. XEOL spectra were recorded using an SBP300 monochromator (Zolix Instruments Co., Ltd.) and a PMTH-S1-CR 131 photomultiplier (Hamamatsu Photonics K.K.).
The MTF and the detective quantum efficiency (DQE) for the prepared scintillation screens were then obtained to evaluate the X-ray imaging performance. The MTF cahracterizes the spatial resolution for a given imaging system, which was determined utilizing the standard edge measurement [31,32]. The MTF value is generally normalized to unity at zero frequency. The spatial resolution of the detection system was calculated at the frequency corresponding to the point where the MTF value decreased to 0.1 [17]. The DQE is represented in the frequency-domain spectrum of signal-to-noise ratio chataceteristic for a particular detective system. The measurement method employed in this investigation was the International Standard IEC 62220-1 [33]. The measurements were conducted at the X-ray Imaging and Biomedical Application Beamline (BL13HB) at the Shanghai Synchrotron Radiation Facility (SSRF) by using an X-ray beam with energy of 20 keV. The experimental setup and method were consistent with those described in our previous research [13,34]. The exposure time and incident X-ray photon flux density were kept at 10 s and 1.48 × 106 phs/s/mm2, respectively. The relevant parameters of calculation for the incident X-ray photon flux have been given in Table S1 of Supplement 1. The Supplement 1 provides a more comprehensive explanation of the experiment and its derivation process.
4. Results and discussions
4.1 Crystallinity and morphology
The morphologies of the micropores arranged in single-period and double-period arrays in Si templates, along with the high-magnification SEM images of the pore walls in the respective templates are shown in Fig. 3. The top and side views of the single-period micropore array templates have been depicted in Fig. 3(a), (c) and (e), whereas the top and side views of the double-period array template are depicted in Fig. 3(b), (d) and (f). All the micropores formed perfect square columns with a depth of approximately 40 µm. They were arranged in the same square array structures, with a period of 4 µm for the single-period array and 8 µm for the double-period array. High-magnification SEM images of the pore walls and the spatial distribution of O (pink) and Si (green) elements in the single-period array, as well as the main structure and sub-structure of the double-period array, are shown in Fig. 3(c-1), (d-1), and (d-2), respectively. The pore wall of the main structure in the double-period array was identical to that of the single-period array, and both were prepared using a sandwich structure consisting of SiO2-Si-SiO2. The total thickness of the pore wall including the sandwich structure was found to be around 1.20 µm and the thicknesses of the oxide and the silicon layers were each measured to be about 0.40 µm.
Fig. 3. Morphologies of the micropores arranged in single-period and double-period array templates, high magnification SEM images of the pore walls, and the spatial distribution of O (pink) and Si (green) elements. (a), (c) and (e) depict the top and side views of the micropores arranged in a single-period array, whereas (b), (d), and (f) depict the top and side views of the micropores arranged in a double-period array. (c-1) The pore wall of a single-period template. (d-1) and (d-2) show images of the pore wall of the main structure and the sub-structure in the double-period template.
Interestingly, it was found that the pore wall of the sub-structure in the double-period array was completely oxidized, and its thickness was about 0.80 µm. The dark grey oxide layers shown in the high-magnification SEM images were identified as SiO2 layers, based on the findings of one of our previous studies [34]. As a note, the main reason for a higher concentration of O towards the left of the elemental distribution can be attributed to the fact that the secondary electron detector of EDS was located diagonally above the samples. Significantly, a thinner pore wall thickness can increase the filling ratio of the CsI scintillator in the screen, and the pore wall made of SiO2 can enhance the efficiency of light guidance.
The morphologies of the scintillation screens with single-period and double-period structures have been depicted in Fig. 4. The images in the upper row represent the single-structured CsI screen, whereas the images in the lower row correspond to the double-structured CsI screen. It can be seen from the side views shown in Fig. 4(a) and (d) that CsI was filled compactly and homogeneously into pores of the templates without the formation of any microbubbles or fractures. The disappearing CsI microcolumns from the pores adhered to the opposite side of the cut samples. The thickness of both the single-period and double-period scintillation screens was approximately 40 µm. Top views of the single-structure CsI scintillation screen are shown in Fig. 4(b) and (c), while top views of the double-structured screen are presented in Fig. 4(e) and (f). It is visually evident that the arrangement and spacing of the CsI scintillation square columns were identical to those that were observed for micropores in the corresponding templates.
Fig. 4. Morphologies of the CsI scintillation screens with (a) - (c) single-period and (d) - (f) double-period structures.
The X-ray diffraction patterns of the CsI scintillation screens with single-period and double-period structures are presented in Fig. 5. For comparison, the diffraction pattern of CsI powder is included, along with the XRD pattern of the Si template and the standard PDF card pattern of CsI (ICPDS 06-311). All diffraction peaks for both structured scintillation screens matched the standard PDF card of CsI (ICPDS 06-311), except for the peak at a 2θ value of 69.20°, which corresponds to X-ray diffraction from the template. A comparison between the diffraction peak intensities for the CsI scintillation screens and the CsI powder revealed that the intensity of the (110) diffraction peak corresponding to the scintillation screen was more prominent in comparison to the others. This indicated that the CsI microcolumns of the structured scintillation screens had a preferred orientation in the (110) planar direction.
Fig. 5. The X-ray diffraction patterns of the obtained CsI scintillation screens with single-period and double-period arrays. The observed diffraction patterns for the CsI powder, the Si template, and the standard PDF card pattern of CsI (ICPDS 06-311) are shown for comparison.
4.2 X-ray excited optical luminescence
The XEOL spectra of the CsI scintillation screens with single-period and double-period structures are shown in Fig. 6(a). Two emission peaks were found at 315 nm and 515 nm, and a small shoulder peak was found at 410 nm. The emission peak at 315 nm is suggested to be associated with the self-trapped excitons of the Vk + e and H + F types [35,36]. Opinions on the origin of the emission peak at 410 nm are relatively diverse. Some believe that it originates from the lattice defects caused by iodine vacancies [37–39], and some believe that it originates from trace impurities [35,36]. The emission peak at 515 nm is mostly thought to be caused by extrinsic defects associated with impurities [35,36]. The bars on the left side of Fig. 6(b) represent the relative LO for the single-structure and double-structure CsI screens, with the ordinate shown in unit of LO for the screen with a single-period structure. It was observed that the LO value for the double-structured CsI screen was approximately 1.31 ± 0.03 times greater than that for the single-period structured CsI screen. This demonstrates that the X-ray detection efficiency of the scintillation screen can be significantly improved by utilizing the double-period structure. The increase in the light output can be attributed to two factors. First, the filling CsI area in the double-structure screen increased by about 14.8% in comparison to that in the single-structured screen. This resulted in the emission of more scintillation light under the same amount of X-ray exposure. Second, the pore walls of the sub-structure in the double-period template were completely oxidized to SiO2. The absence of the silicon absorption layer in the pore wall improved the light-guiding performance of the structured scintillation screen. The two bars on the right side of Fig. 6(b) represent the relative LO, with the left one corresponding to the single-structured CsI screen and the right one to the double-structured CsI screen, as obtained through simulations. Structural parameters of the scintillation screens used in simulations were measured experimentally. The LO of the double-structure CsI screen was about 1.28 times greater than that obtained for the single-structured CsI screen, which is in agreement with the experimental results within a reasonable error range. About 18% of the observed increment in LO was due to the increase in light yield resulted from the higher of CsI filling rate, indicated by the purple segment on the rightmost side of Fig. 6(b). Additionally, about 10% of the increment (depicted by the orange portion of the bar on the rightmost side in Fig. 6(b)) was attributed to improvements in light guide efficiency and light yield. The increase rate of light yield is larger than that of CsI filled area because the pore wall forming the sub-structure of the double-structure screen is relatively thin. The electrons generated by X-ray excitation are more likely to enter the CsI microcolumns to produce scintillation light. This investigated results thus confirm that the efficiency of the developed double-structure CsI scintillation screen can indeed reach the theoretical value in terms of LO.
Fig. 6. (a) The obtained XEOL spectra of the CsI scintillation screens with single-period and double-period structures. (b) The variation in light output between the single-structured and the double-structured scintillation screens. The bars on the left represent experimental results, whereas the bars on the right represent the simulated results.
4.3 Performance of X-ray imaging
The images captured using the single-structure and double-structure scintillating screens under uniform X-ray exposure have been presented in Fig. 7(a) and (b), respectively. The luminescence of both screens was generally observed to be fairly uniform, with the double-structured screen being brighter than the single-structured screen. Additionally, the pore walls between the sub-structures of the double-structured screen displayed a good light isolation effect, albeit somewhat weaker compared to the pore walls of the main structure and the single-structured screen. Histograms of scintillation light intensity for the CsI microcolumns on the single and double structured screens are provided in Fig. 7(c). The results then quantitatively confirmed the relative intensity and uniformity of the scintillation light for the two screens. The light intensity of the double-structure screen was found to be 30% higher than the light intensity measured for the single-structure screen. These results regarding relative light intensity were consistent with the findings obtained from their XEOL spectra. The variation in scintillation light intensity among individual CsI microcolumns within the screens was approximately 7% or less, suggesting a fairly uniform distribution of luminescence for both screens.
Fig. 7. Imaging performance under uniform X-rays exposure utilizing the single-structure and double-structured screens. Images for the (a) single-structure and (b) double-structure scintillation screens when exposed to uniform X-rays. (c) Histograms of the relative scintillation light intensity for CsI microcolumns on the structured screens. (d) Experimental and (e) simulated MTF curves plotted against spatial frequency. (f) Scintillation light intensities for the double-structure screen at different X-ray exposure times. (g) The DQE plots for the X-ray imaging system as a function of the spatial frequency. X-ray images of the micro resolution plate (JIMA RT RC-02) using the (h) single-structure and (i) double-structure CsI scintillation screens. Gray values along (j) red line and (k) blue line of resolution plate imaged using single-structure and double-structure scintillation screens.
The obtained experimental and simulated MTF plots of the X-ray imaging system for the single and double-structured screens have been shown in Fig. 7(d) and (e). Both sets of results showed that both types of structured scintillation screens provided excellent X-ray imaging resolution. Additionally, the resolution achieved with the double-structure scintillation screen was only slightly lower than that achieved with the single-structure scintillation screen. The spatial resolutions measured using the double-structure and the single-structure scintillation screens were 109 lp/mm and 111 lp/mm, respectively, which are very close to their corresponding values obtained from the simulation results (112 lp/mm and 115 lp/mm). This suggests that the development of CsI scintillation screens in this study, using both double-structures and single-structures, was successful. The slight reduction in spatial resolution of the double-structure screen may be due to the lower ability of the fully oxidized pore walls between the sub-structures to confine lateral propagation of the scintillation photons, in comparison to the partially oxidized pore walls between the main structures.
The average scintillation light intensities for the double-structure screen at different X-ray exposure times were subsequently measured, and have been presented in Fig. 7(f). The average intensity of the scintillation screen grew linearly with an increase in the X-ray exposure time, which was beneficial for achieving a good imaging contrast. The obtained DQEs for the X-ray imaging system utilizing CsI screens with the single-structure and the double-structure are provided in Fig. 7(g). The DQEs increasingly reduce as spatial frequency enhances. The DQE for the double-structure screen was higher than that of the single-structure CsI screen at low spatial frequencies. This was because the double-structure screen had a higher X-ray detection efficiency and a better light guide efficiency due to the larger CsI filling rate and the completely oxidized pore walls in the sub-structures. At higher spatial frequencies, the DQE obtained from the double-structure screen was slightly lower in comparison to that from the single-structure screen, since at this moment, the influence of the MTF becomes prominent. However, the value of MTF for the double-structure screen was only slightly lower than what was determined for the single-structure screen. The slight reduction in imaging resolution can be considered negligible in light of the improvements in LO and DQE values for the double-structure scintillation screen. These findings confirm that the X-ray imaging efficiency using CsI screens with a double-structure surpasses the performance of the single-structure screen. X-ray images of the micro resolution plate (JIMA RT RC-02, Japan Inspection Instruments Manufacturers’ Association) were captured using the prepared CsI scintillation screens based on the single-structure and double-structure shown in Fig. 7(h) and (i) respectively. On a large scale, the image captured using the double-structure screen is brighter and clearer than that captured using the single-structure screen. Combined with the results of DQE in Fig. 7(g), one can conclude that the double-structure screen has a higher signal-to-noise ratio in almost all frequency regions below resolution limit. In order to better understand the imaging quality of the two screens at a small scale, the images of the line pairs in dashed wireframes in Fig. 7(h) and (i) are enlarged and displayed in Fig. 7(j) and (k). It can be seen that the images obtained by both single-structure and double-structure scintillation screens can basically distinguish the line pairs of 4 µm. The difference between peaks and valleys of gray scale detected by the single-structure screen is slightly larger than that detected by the double-structure screen, which means that the image captured by the single-structure screen is slightly clearer at this scale. Although the spatial resolution of the image captured by the double-structure screen seems to be slightly worse than that captured by the single-structure screen, the signal-to-noise ratio of the image captured by the double-structure screen is higher than those captured by the single-structure screen in almost all frequency regions below resolution limit. Therefore, in general, the imaging quality using the double-structure screen is better than that using the single-structure screen.
5. Conclusion
The development of a CsI scintillation screen with a double-period structure based on an oxidized Si micropore array template was successfully achieved to simultaneously attain both a high detective efficiency and excellent spatial resolution for X-ray imaging. A range of processes that comprise electrochemical etching, ICP, photolithography, thermal oxidation, and vacuum melting gas pressure injection were employed to prepare these CsI scintillation screens. The new scintillation screen structure featured a combination of a main structure and a sub-structure. The main structure had a period of 8 µm and was composed of square columnar scintillator units arranged in a square array configuration. Each of these units had a length of about 40 µm. The main units were separated by micropore walls with a layered structure composed of SiO2-Si-SiO2, which was the same composition used in the widely adopted single-structured scintillation screens. The each thickness of both the Si and SiO2 layers in the micropore walls was about 0.4 µm. The unique aspect of the double structure was that each main structural unit was equally divided into four square columnar scintillator sub-units. To produce the sub-structure, the sub-units were separated just by SiO2 layers with a thickness of about 0.8 µm. The CsI melt was packed into the template in an even and compact manner ensuring that no microbubbles or fractures were generated. The CsI microcolumns of the structured scintillation screen exhibited a preferential orientation in the (110) planar direction. A single-period structured CsI scintillation screen was prepared for comparison using the same process. The thickness of the screen was approximately 40 µm. In the single-structured CsI scintillation screen, the CsI scintillation microcolumns were square and arranged in a square array. The period of this single-structured screen was matched to the average period of the sub-structure in the CsI scintillation screen with a double-period structure. The obtained XEOL spectra of the screens revealed that the luminescent intensity of the double-structure CsI screen was about 31% greater than that of the single-structure screen, which corroborated quite well with the values obtained from the simulation results. These findings suggest that the developed double-structure was successful in enhancing X-ray absorption and the propagation of scintillation light within the CsI scintillation screen. Additionally, the results indicate that the double-structured CsI scintillation screen maintained excellent spatial resolution for X-ray imaging, achieving 109 lp/mm, which is very close to the resolution provided by the single-structured CsI screen. Owing to the higher X-ray detection efficiency and excellent imaging spatial resolution, the double-structure CsI scintillation screen also exhibited a higher DQE in comparison to the CsI scintillation screen based on the single-structure, throughout almost the entire frequency range below the resolution limit. The superiority of the double-structured CsI scintillation screen was also confirmed by a comparison of the X-ray images captured using both the single-structured and double-structured CsI scintillation screens.
In summary, the double-structured CsI scintillation screen was successfully prepared and demonstrated superior performance in terms of detection efficiency and X-ray imaging quality compared to the CsI scintillation screen based on the single-structure. It maintained nearly the same high spatial resolution as the single-structured screen. Although promising progress has been made in this research, the development of scintillation screens with both excellent imaging resolution and detection efficiency is still on the way. The imaging performance of the structured scintillation screen is expected to be further improved by changing the arrangement of the double-structure, period size, micropore shape, screen thickness and so on to meet the specific X-ray imaging requirements.
Funding
National Natural Science Foundation of China (11675121, 11475128, 11775160); X-ray Imaging and Biomedical Application Beamline (BL13HB) at Shanghai Synchrotron Radiation Facility.
Disclosures
The authors declare no conflict of interest.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. J. P. Moy, “Recent developments in X-ray imaging detectors,” Nucl. Instrum. Methods Phys. Res., Sect. A 442(1-3), 26–37 (2000). [CrossRef]
2. V. V. Nagarkar, T. K. Gupta, S. R. Miller, et al., “Structured CsI(Tl) scintillators for X-ray imaging applications,” IEEE Trans. Nucl. Sci. 45(3), 492–496 (1998). [CrossRef]
3. V. V. Nagarkar, S. V. Tipnis, V. B. Gaysinskiy, et al., “New design of a structured CsI(Tl) screen for digital mammography,” Medical Imaging 2003 Conference, San Diego, CA, Feb 17-20, 5030, 541–546 (2003).
4. B. J. Kim, B. K. Cha, H. Jeon, et al., “A study on spatial resolution of pixelated CsI(Tl) scintillator,” Nucl. Instrum. Methods Phys. Res., Sect. A 579(1), 205–207 (2007). [CrossRef]
5. P. Kleimann, J. Linnros, C. Fröjdh, et al., “An X-ray imaging pixel detector based on scintillator filled pores in a silicon matrix,” Nucl. Instrum. Methods Phys. Res., Sect. A 460(1), 15–19 (2001). [CrossRef]
6. Y. Hormozan, I. Sychugov, and J. Linnros, “High-resolution X-ray imaging using a structured scintillator,” Med. Phys. 43(2), 696–701 (2016). [CrossRef]
7. O. Svenonius, A. Sahlholm, P. Wiklund, et al., “Performance of an X-ray imaging detector based on a structured scintillator,” Nucl. Instrum. Methods Phys. Res., Sect. A 607(1), 138–140 (2009). [CrossRef]
8. X. Badel, J. Linnros, P. Kleimann, et al., “Metallized and oxidized silicon macropore arrays filled with a scintillator for CCD-based X-ray imaging detectors,” IEEE Trans. Nucl. Sci. 51(3), 1001–1005 (2004). [CrossRef]
9. C. Amsler, D. Grögler, W. Joffrain, et al., “Temperature dependence of pure CsI: scintillation light yield and decay time,” Nucl. Instrum. Methods Phys. Res., Sect. A 480(2-3), 494–500 (2002). [CrossRef]
10. A. Boyarintsev, Y. Boyarintseva, A. Gektin, et al., “Study of radiation hardness of pure CsI crystals for Belle-II calorimeter,” J. Instrum. 11(03), P03013 (2016). [CrossRef]
11. X. Badel, B. Norlin, P. Kleimann, et al., “Performance of scintillating waveguides for CCD-based X-ray detectors,” IEEE Trans. Nucl. Sci. 53(1), 3–8 (2006). [CrossRef]
12. M. Simon, K. J. Engel, B. Menser, et al., “X-ray imaging performance of scintillator-filled silicon pore arrays,” Med. Phys. 35(3), 968–981 (2008). [CrossRef]
13. S. Liu, M. Gu, H. Chen, et al., “Performance of pixelated CsI scintillation screen with hexagonal array arrangement prepared by vacuum melting injection method,” Nucl. Instrum. Methods Phys. Res., Sect. A 903, 18–24 (2018). [CrossRef]
14. K. Chen, M. Gu, Z. X. Sun, et al., “Influence of preparation process on the transparency of CsI microcolumns in the structured CsI scintillation screen based on oxidized silicon micropore array template,” Nucl. Instrum. Methods Phys. Res., Sect. A 991, 164999 (2021). [CrossRef]
15. H. Chen, M. Gu, Z. Sun, et al., “Influence of silicon wall thickness on the performance of structured CsI(Tl) scintillation screen based on oxidized silicon micropore array template in X-ray imaging,” Opt. Express 27(10), 14871–14880 (2019). [CrossRef]
16. Z. Sun, M. Gu, Y. Teng, et al., “Influence of the thickness of a SiO2 reflective layer on the performance of a structured CsI(Tl) scintillation screen based on an oxidized Si micropore array template in X-ray imaging,” Opt. Express 30(12), 21324–21337 (2022). [CrossRef]
17. Z. Sun, M. Gu, Q. Li, et al., “Performance of a CsI(Tl) scintillation screen with a dual-periodic structure based on an oxidized silicon micropore array template in X-ray imaging,” Opt. Express 29(15), 23752–23763 (2021). [CrossRef]
18. S. Agostinelli, J. Allison, K. Amako, et al., “GEANT4-a simulation toolkit,” Nucl. Instrum. Methods Phys. Res., Sect. A 506(3), 250–303 (2003). [CrossRef]
19. J. Allison, K. Amako, J. Apostolakis, et al., “Geant4 developments and applications,” IEEE Trans. Nucl. Sci. 53(1), 270–278 (2006). [CrossRef]
20. D. Stankevich and Y. Shkuratov, “Monte Carlo ray-tracing simulation of light scattering in particulate media with optically contrast structure,” J. Quant. Spectrosc. Radiat. Transfer 87(3-4), 289–296 (2004). [CrossRef]
21. A. Taflove, S. C. Hagness, and M. Piket-May, “Computational Electromagnetics: The Finite-Difference Time-Domain Method,” in Electrical Engineering Handbook, 3th, ed. (Academic, 2005). [CrossRef]
22. J. E. Eldridge, “Cesium Iodide (CsI),” in Handbook of Optical Constants of Solids, E. D. Palik, ed. (Academic, 1997), pp. 853–874.
23. D. F. Edwards, “Silicon (Si)*,” in Handbook of Optical Constants of Solids, E. D. Palik, ed. (Academic, 1997), pp. 547–569.
24. L. Gao, F. Lemarchand, and M. Lequime, “Refractive index determination of SiO2 layer in the UV/Vis/NIR range: spectrophotometric reverse engineering on single and bi-layer designs,” J. Eur. Opt. Soc.-Rapid Publ. 8, 13010 (2013). [CrossRef]
25. D. E. Aspnes and A. A. Studna, “Dielectric functions and optical parameters of Si, Ge, GaP, GaAs, GaSb, InP, InAs, and InSb from 1.5 to 6.0 eV,” Phys. Rev. B 27(2), 985–1009 (1983). [CrossRef]
26. H. R. Philipp, “Silicon Dioxide (SiO2) (Glass),” in Handbook of Optical Constants of Solids, E. D. Palik, ed. (Academic, 1997), pp. 749–763.
27. Z. Sun, M. Gu, X. Liu, et al., “Influence of Si wall thickness of CsI(Tl) micro-square-frustums on the performance of the structured CsI(Tl) scintillation screen in X-ray imaging,” Sci. Rep. 12(1), 8748 (2022). [CrossRef]
28. Y. J. Oh, C. A. Ross, Y. S. Jung, et al., “Cobalt nanoparticle arrays made by templated solid-state dewetting,” Small 5(7), 860–865 (2009). [CrossRef]
29. H. Okayama, K. Fukami, R. Plugaru, et al., “Ordering and disordering of macropores formed in prepatterned p-type silicon,” J. Electrochem. Soc. 157(1), D54–D59 (2010). [CrossRef]
30. J. H. Kim, K. P. Kim, H. K. Lyu, et al., “Three-dimensional macropore mrrays in p-type silicon fabricated by electrochemical etching,” J. Korean Phy. Soc. 55(1), 5–9 (2009). [CrossRef]
31. E. Samei, M. J. Flynn, and D. A. Reimann, “A method for measuring the presampled MTF of digital radiographic systems using an edge test device,” Med. Phys. 25(1), 102–113 (1998). [CrossRef]
32. H. Fujita, D. Y. Tsai, T. Itoh, et al., “A simple method for determining the modulation transfer function in digital radiography,” IEEE Trans. Med. Imaging 11(1), 34–39 (1992). [CrossRef]
33. C. Michail, I. Valais, N. Martini, et al., “Determination of the detective quantum efficiency (DQE) of CMOS/CsI imaging detectors following the novel IEC 62220-1-1:2015 International Standard,” Radiat. Meas. 94, 8–17 (2016). [CrossRef]
34. Y. X. Teng, M. Gu, Z. X. Sun, et al., “Optimization of SiO2 reflective layer thickness for improving the performance of structured CsI scintillation screen based on oxidized Si micropore array template in X-ray imaging,” Opt. Express 31(15), 24097–24109 (2023). [CrossRef]
35. Z. Wu, B. Yang, and P. D. Townsend, “Radioluminescence and thermoluminescence properties of X-ray-irradiated pure CsI,” J. Lumin. 128(7), 1191–1196 (2008). [CrossRef]
36. D. Totsuka, T. Yanagida, Y. Fujimoto, et al., “Growth and scintillation properties of pure Csl crystals grown by micro-pulling-down method,” Opt. Mater. 34(7), 1087–1091 (2012). [CrossRef]
37. M. M. Hamada, Y. Nunoya, S. Kubota, et al., “Suppression of the slow emission component in pure CsI by heat treatment,” Nucl. Instrum. Methods Phys. Res., Sect. A 365(1), 98–103 (1995). [CrossRef]
38. D. Z. Shen, Q. Deng, X. L. Yuan, et al., “Luminescent performances of pure CsI crystals,” J. Inorg. Mater. 12(3), 273–278 (1997).
39. G. H. Ren, Z. H. Song, Z. C. Zhang, et al., “Luminescence and decay time properties of pure CsI crystals,” J. Inorg. Mater. 32(2), 169–174 (2017). [CrossRef]
