Electrochemically Fabricated SrGa₂S₄:Eu/ZnO-Vertical Nanowire Bilayer Scintillator for X-ray Imaging

Soda-lime glass substrates with Ga-doped ZnO (GZO) layers (GZO/SLG, GEOMATEC Co., Ltd.) were used as substrates, and after ultrasonic cleaning in acetone for 3 min as a pretreatment, degreasing was performed in a solution of 0.1 mol l⁻¹ NaOH at an applied anodic current of 0.1 mA cm⁻² for 1 min. ZnO-vnws were prepared by electrodeposition in aqueous solutions containing zinc nitrate hydrate (NACALAI TESQUE, INC.) and zinc chloride hydrate (NACALAI TESQUE, INC.) with the total Zn concentration set at 8 mmol L⁻¹ with a zinc chloride molar ratio of 0.1 mol%. The aqueous solution was prepared using deionized water (15 MΩ·cm) prepared by Millipore Ellix-UV-Advantage and special grade reagents. The electrodeposition was carried out using a potentiostat/galvanostat (Princeton Applied Research, VersaSTAT 4–400) at −0.8 V referenced to an Ag/AgCl electrode, under a temperature of 335 K, deposition time of 20 h, and a deposition area of 1 × 1 cm².

The SrGa₂S₄:Eu layer was stacked on the ZnO-vnws/GZO/SLG and bare GZO/SLG substrate by electrophoresis deposition in an ethanol solution containing SrGa₂S₄:Eu particles. The solution contained 50 ml of ethanol, 0.5 g of SrGa₂S₄:Eu (Mitsui Mining & Smelting Co., Ltd.), and 20 μg of iodine. The electrophoresis depositions were carried out by applying a voltage of 60 V for 10, 15, 20, and 30 s using a DC power supply (PAN 160–3.5 A, Kikusui Electronics Corporation). Subsequently, an Ag layer with a thickness of 100 nm was deposited using thermal evaporation in vacuum (VCP-260F, ULVAC, Inc.) at a deposition rate of 0.1 nm s⁻¹ for the fixation of SrGa₂S₄:Eu particles and the formation of the reflective layer.

X-ray diffraction patterns were measured by X-ray diffractometer (RINT-2500, Rigaku Corporation) with CuKα radiation, and the surface and cross-sectional morphologies were observed using an FE-SEM (SU8000, Hitachi High-Tech Corporation). The optical properties were evaluated by absorption and transmission spectra measured using an ultraviolet–visible-near-infrared spectrophotometer (U4100, Hitachi High-Technologies), and the photoluminescence and excitation spectra were measured using a spectrophotometer (F7000, Hitachi High-Technologies). The emission lifetime was obtained by excitation at 337.1 nm using pulsed nitrogen laser and photomultiplier as a detector. The decay lifetime of the emission was estimated by the single exponential fitting of the emission decay curve using the non-linear least square method. X-ray images were recorded using an X-ray inspection system (SMX-1000Plus, Shimadzu Corporation) with installing a customized optical system consisting of a CMOS camera (ORCA-Flush4.0 V3, Hamamatsu Photonics K. K.) and an optical lens with fine focus adjustment. The chart (RT RC-05B, Japan Inspection Instruments Manufactures' Association (JIMA)) with a Line and Space (L/S) of 9, 10, 15, 20, 45, and 50 μm was used for estimating the spatial resolution, and a 50-μm-thick LuAG scintillator (Hamamatsu Photonics K.K.) was used for comparison. An exposure time of 10 s was used for capturing dark (no X-ray irradiation), background (no chart in X-ray irradiation), and chart images. The intensity of the dark image was subtracted from that of the background and chart images, which the chart image was then divided by the background image. By taking the natural logarithm of the image-processed chart image, it was converted into X-ray absorption. X-ray absorption images of the chart were provided for evaluation of the spatial resolution by using the Sigmoid curve-fitting on an absorption profile of a chart bar edge. The fitted function used in the present study is as follows:

$$\begin{eqnarray}&&I\left(x\right)=\displaystyle \frac{a}{1+b{e}^{-cx}}+d\end{eqnarray} \tag{ 1 }$$

where I(x) is the X-ray absorption at position x. The constant values of a, b, c, and d were detected by a fitting function of MATLAB. After finding the best-fit constants, the full width at half maximum (FWHM) value was determined from the differential curve of Eq. 1, and this value was used as a parameter for the spatial resolution.

Figure 1 shows the schematic illustration of the Ag/SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator and the appearance of the SrGa₂S₄:Eu/ZnO-vnws/GZO bilayer scintillator prepared at the condition of 60 V-20 s for the SrGa₂S₄:Eu electrophoresis deposition. The appearance observed from the side of the SLG substrate was yellow-green in color originated from the SrGa₂S₄:Eu particle, while it appeared gray from the side of the Ag layer. From the constant appearances and colors covering the entire surfaces for both sides, the SrGa₂S₄:Eu and Ag layers were formed uniformly. The SrGa₂S₄:Eu particles were properly fixed to the ZnO-vnws, and no detachment of particles could be observed during the fabrication process including the vacuum evaporation of Ag. The SrGa₂S₄:Eu particles were deposited only on the ZnO-vnws/GZO cathode surface by electrophoresis deposition, suggesting that the SrGa₂S₄:Eu particles were positively charged. Since the GZO and ZnO substrates are n-type semiconductors whose majority carriers are electrons, the electrostatic attraction between the positively charged SrGa₂S₄:Eu particles and the cathode substrate may have also been beneficial to the fixation. However, no deposits could be obtained even under applied voltage in the SrGa₂S₄:Eu particle dispersed ethanol solution without iodine addition. An ethanol-iodine solution was generally used for electrophoresis deposition, besides acetone-iodine solution. It was reported the reaction of the acetone and iodine as shown in Fig. S1(available online at stacks.iop.org/JES/169/042507/mmedia), ^17,18 and the particles dispersed in the solution charged positive by adsorbing the H⁺ ion on the particles. In this study, the SrGa₂S₄:Eu particles were only deposited on the ZnO-vnws/GZO cathode, and the appearance of the SUS anode turned brown after the electrophoresis deposition. This indicated that the positive charge of the SrGa₂S₄:Eu particle was generated by a reaction similar to the acetone-iodine reactions in the solution.

Zoom In Zoom Out Reset image size

Figure 2 shows the X-ray diffraction patterns of the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillators prepared at 60 V-30 s for the SrGa₂S₄:Eu electrophoresis deposition. The X-ray diffraction patterns were almost the same in profile and peak angle, irrespective of the condition of the SrGa₂S₄:Eu electrophoresis deposition, except for the intensity of the SrGa₂S₄:Eu peaks. All the diffracted X-ray peaks were identified as those for ZnO, ¹⁹ SrGa₂S₄, ²⁰ and Ag. ²¹ The ZnO (002) peak with very high intensity at around 400 kcps was composed of a peak and a shoulder originating from the (002) plane by CuKα, and CuKα₂ radiation, indicating a very small FWHM value. High intensity and low FWHM value indicated an excellent (001) out-of-plane orientation, as already reported. ¹⁵ An X-ray diffraction peak originating from the GZO substrate was observed for the X-ray diffraction pattern for the bare substrate, but the (002) GZO peak was included in the ZnO (002) peak originated from the ZnO-vnws, due to its low intensity and expanded FWHM value. The lattice constant c for the ZnO lattice was estimated to be approximately 0.524 nm, which was close to the standard value. ¹⁹

Zoom In Zoom Out Reset image size

Figure 3 shows the cross-sectional and surface FE-SEM images of the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillators prepared at 60 V for 10, 15, 20, and 30 s. The SrGa₂S₄:Eu particles with grain sizes ranging from 0.13 ∼7 μm were stacked by the electrophoresis deposition on the substrate irrespective of the existence of the ZnO-vnws and electrophoresis conditions. The top surface of SrGa₂S₄:Eu particles was covered entirely with Ag granular grains (not shown). The thicknesses of the SrGa₂S₄:Eu layers were estimated from 15 location points of the cross-sectional images, and the average values calculated were represented by orange circles and gray squares, with corresponding error bars showing the standard deviations in Fig. 3e. The average SrGa₂S₄:Eu layer thickness increased with the increase in the electrophoresis deposition time, irrespective of the existence of ZnO-vnws, although slight fluctuations remained depending on the location. Although there were gaps between the SrGa₂S₄:Eu particles, overall uniformly dense layers were formed. The SrGa₂S₄:Eu particles stacked adhesively on the underlayer, irrespective of the existence of the ZnO-vnws, and no defects such as space between the SrGa₂S₄:Eu layer and ZnO-vnws could be located. The ZnO-vnws layer with the width of 0.16 ∼0.2 μm and length of 5.4 μm was composed of columnar grains grown normal to the GZO substrate surface. The ZnO-vnws possessed an excellent (0001) out-of-plane orientation, as reported from a pole-figure measurement, which orientation was developed by a lattice relationship of $\left(1\times 1\right)\left(0001\right)\left[11\underline{2}0\right]{\rm{ZnO}}$//$\left(1\times 1\right)\left(0001\right)\left[11\underline{2}0\right]\,$ GZO with a lattice mismatch of 1.1%. ¹⁵ The width and length of the ZnO-vnws were controlled by adjusting the total Zn and Cl concentration. ¹⁶

Zoom In Zoom Out Reset image size

Figure 4 shows the optical absorption spectra of the SrGa₂S₄:Eu /ZnO-vnws bilayer and the SrGa₂S₄:Eu layer, the emission and excitation spectra, and the decay curve for the bilayer. The SrGa₂S₄:Eu layer did not show a clear absorption edge at wavelengths ranging from 400 to 800 nm, although a slight change was observed around 500 nm. The bare SLG substrate possessed a strong absorption edge around 350 nm with no absorption edge observed in the range of 400 ∼800 nm, and this observation was consistent with the bandgap energy of 4.2 eV as already reported for the SrGa₂S₄ phosphor, ²² which corresponding absorption edge is at about 295 nm. The absorption spectrum was almost the same in profile at wavelengths ranging from 450 to 800 nm for both, although the absorbance slightly increased with the presence of ZnO-vnws. An additional absorption edge was observed at approximately 400 nm for the bilayer, and its bandgap energy was estimated to be 3.2 eV with the assumption of a direct transition as shown in Fig. S2, indicating that the absorption edge around 400 nm was attributed to the ZnO-vnws. ¹³

Zoom In Zoom Out Reset image size

Only a visible emission at about 533 nm (excitation wavelength of 320 nm) was observed from the SrGa₂S₄:Eu/ZnO-vnws bilayer and the SrGa₂S₄:Eu layer irrespective of the presence of ZnO-vnws and the condition of the electrophoresis deposition. The emission at the wavelength of 533 nm was previously reported to be attributed to the 4f⁶5d→4f⁷ transition of Eu²⁺ ions in SrGa₂S₄:Eu. ^22,23 The excitation spectrum (emission wavelength 533 nm) showed a broad form with a peak at around 390 nm. The excitation spectrum with a peak at 278 nm and shoulders at 313 and 391 nm (emission wavelength of 532 nm) was reported for SrGa₂S₄:Eu containing 2 mol% Eu, ²² and a broad excitation spectrum with a peak at 464 nm (emission wavelength of 538 nm) was reported for Sr_1–0.03Ga₂S₄:Eu_0.03. ⁹ The excitation band at 350 ∼500 nm was assigned to the 4f⁷ → 4f⁶5d transition of Eu²⁺ ions in the SrGa₂S₄:Eu phosphor. ²³ While the peak wavelength and the shape of the emission spectrum are almost the same, the peak wavelengths and the shape of the excitation spectra differ greatly among reports. These differences are probably due to differences in measurement conditions, and the associated band structures derived from the composition and coordination environment of the Eu atom. The emission spectrum recorded for a single ZnO-vnws layer under the same conditions showed very low intensities with small emission peaks, as shown in Fig. S2.

The decay and fitting curves were represented by black and red lines in Fig. 4d, and the lifetime was estimated for the decay curve by the following process. The relationship between the intensity and lifetime can be expressed by the following equation, ²⁴

$$\begin{eqnarray*}&&Y=A1+A2\exp \left(\displaystyle \frac{x}{\tau }\right)\,\end{eqnarray*}$$

where Y, x, τ are intensity, time, lifetime, and A1 and A2 are constants. The lifetime τ was estimated from the slope of the linear part of the natural logarithm plot between the intensity and time, which was estimated to be 0.389 μs.

Figure 5 shows the measured X-ray chart image of line profile with the Line and Space (L/S) of 9, 10 15, 20, 45, and 50 μm for the SrGa₂S₄:Eu/ZnO-vnws bilayer and SrGa₂S₄:Eu layer by WKα radiation. The line profiles with L/S of 9 ∼50 μm were clearly observed for all the SrGa₂S₄:Eu/ZnO-vnws bilayer and the SrGa₂S₄:Eu layer, evidencing that the SrGa₂S₄:Eu/ZnO-vnws bilayer and SrGa₂S₄:Eu layer were functioning as scintillators for the X-ray radiation. A line profile of L/S = 6 μm was observed with a ZnO-vnws-free-SrGa₂S₄:Eu scintillator, as already reported. ¹⁰ Since it was difficult to estimate quantitatively the spatial resolution by the detection of the L/S line profile, FWHM values were estimated using the Sigmoid curve-fitting technique for the spatial resolution estimation as shown in Fig. 6a. The blue circles and the orange line respectively represent the raw absorbance value at a given X position and the fitting curve by using the Sigmoid curve. The green line and red dashed line represent the differential curve for the Sigmoid fitting curve and the FWHM, which was used as the spatial resolution.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 6b shows the relation of the FWHM value on the SrGa₂S₄:Eu layer thickness for SrGa₂S₄:Eu/ZnO-vnws bilayer and SrGa₂S₄:Eu scintillators. The FWHM value decreased with the decrease in SrGa₂S₄:Eu layer thickness for the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator, but the SrGa₂S₄:Eu scintillator without the ZnO-vnws showed the opposite tendency and showed decreased FWHM value at increased thickness. The reason for the improved FWHM value at increased SrGa₂S₄:Eu layer thicknesses in the absence of the ZnO-vnws is not clear at present, but it was speculated that the irregularity of the top and bottom interfaces of the SrGa₂S₄:Eu layer to the Ag layer and GZO/SLG substrate affected the FWHM value at the thin thickness of 6.9 μm, while such detrimental effects were suppressed with the increase in thickness of the SrGa₂S₄:Eu layer. The SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator showed the best resolution of 6.8 μm at the SrGa₂S₄:Eu layer thickness of 6.1 μm, and the best resolution for SrGa₂S₄:Eu scintillator without the ZnO-vnws was 8.1 μm at the SrGa₂S₄:Eu layer thickness of 12 μm, indicating the improvement in spatial resolution by introducing the ZnO-vnws. The best spatial resolution for the SrGa₂S₄:Eu (60 V-10 s)/ZnO-vnws bilayer scintillator was slightly larger than 5.4 μm estimated for the ready-made LuAG scintillator in the same manner. This is thought to be due to the light-guiding effect caused by the ZnO-vnws morphology functioning as an optical waveguide ¹² which guided the radial light emitted by SrGa₂S₄:Eu into unidirectional light. However, the spatial resolution of the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator became larger as the thickness of the SrGa₂S₄:Eu layer increased which then remained almost constant at a corresponding thickness ratio of 1.9 to the ZnO-vnws layer with the SrGa₂S₄:Eu layer thickness over 10.3 μm. Therefore, the optical guiding effect of the ZnO-vnws layer was more effective when the SrGa₂S₄:Eu layer was thin. The thickness of the SrGa₂S₄:Eu layer and the thickness ratio of the SrGa₂S₄:Eu layer to the length of the ZnO-vnws were speculated to affect the spatial resolution of the bilayer scintillator.

Figure 7 shows the dependence of luminescence intensity and the unit intensity per SrGa₂S₄:Eu thickness on the thickness of SrGa₂S₄:Eu for the SrGa₂S₄:Eu/ZnO-vnws bilayer and the SrGa₂S₄:Eu layer. The dashed lines in Figs. 7a and 7b represent the linear approximations obtained for the values of SrGa₂S₄:Eu scintillators. Since the luminescence intensity of the SrGa₂S₄:Eu/ZnO-vnws bilayer and SrGa₂S₄:Eu scintillators increased as the thickness of the SrGa₂S₄:Eu layer increased, the luminescence intensity strongly depended on the thickness of the SrGa₂S₄:Eu layer. The luminescence intensities of the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillators were located on or below the approximate lines when the SrGa₂S₄:Eu thicknesses are over 10.2 μm, but the intensities were located at the upper side of the approximate line below the SrGa₂S₄:Eu layer thicknesses of ∼7.1 μm. To investigate the effect of the ZnO-vnws introduction on the luminescence intensity, the luminescence intensity per SrGa₂S₄:Eu thickness, which is abbreviated as unit intensity, was calculated and is shown in Fig. 7b. Since the unit intensity of the SrGa₂S₄:Eu/ZnO-vnws scintillator was located on or below the approximate line obtained for the ZnO-vnws-free scintillator, the effect of the ZnO-vnws introduction could not be observed for the SrGa₂S₄:Eu layer thicknesses over 10.3 μm. As the length of the ZnO-vnws was 5.4 μm, the thickness ratio of the SrGa₂S₄:Eu layer to the ZnO-vnws length was estimated to be 1.9 and 2.9 for 10.3 μm and 15.4 μm in SrGa₂S₄:Eu thicknesses. The unit intensities were located above the approximate line at layer thicknesses of the SrGa₂S₄:Eu layer below 7.1 μm, and the effect of the enhancement of the luminescence intensity could be clearly observed with the introduction of the ZnO-vnws. The luminescence intensity was evidently stronger with the introduction of the ZnO-vnws while at the same SrGa₂S₄:Eu layer thicknesses below 7.1 μm. The thickness ratios were estimated to be 1.1 and 1.3 for the respective thicknesses of the SrGa₂S₄:Eu layers of 6.1 and 7.1 μm. The effect of the intensity enhancement by the ZnO-vnws introduction appeared at thickness ratios around 1.1 ∼1.3 but disappeared over approximately 2 for the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator. The SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator with the thickness ratios of 1.1 and 1.3 revealed improved resolutions and enhanced intensities, compared to those of ZnO-vnws-free scintillators.

Zoom In Zoom Out Reset image size

Figure 8 shows the schematic illustrations of the light-guiding effects of the ZnO-vnws in the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator compared to that of a ZnO-vnws-free SrGa₂S₄:Eu scintillator, which are described as follows. The absorption coefficient for the 23μm-Eu:CaS/6μm-ZnO-vnws/GZO/SLG was estimated to be 0.87 cm⁻¹ at 59 KeV corresponding to the photon energy of WKα radiation when experimented in a synchrotron radiation facility (SPring-8), as shown in Fig. S3. The absorption coefficient for the SrGa₂S₄:Eu/ZnO-vnws/GZO/SLG fabricated in this study was not estimated in the same manner, but was speculated to be close to that of the Eu:CaS/ZnO-vnws/GZO/SLG. The WKα radiation at a photon energy of 59 keV was transmitted through the SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator due to the low absorption coefficient, which caused the SrGa₂S₄:Eu phosphor particles to emit 533 nm visible light. The emitted lights then propagated radially from the emission point and reached the interface to the ZnO-vnws. The lights penetrating the ZnO-vnws were propagated inside the nanowires by the light-guiding effects. The light-guiding effect of the ZnO-vnws ²⁵ improving the spatial resolution was demonstrated in an X-ray imaging application by comparing to a randomly dispersed ZnO-vnws and a bare substrate without ZnO-vnws. ¹¹ Additionally, P. Yang et al., reported that the ZnO nanowires can generate stimulated emission by light excitation and enhance the excitation light by acting as optical resonators. ²⁶ Since the ZnO-vnws possessed top and bottom interfaces of ZnO/air and ZnO/SLG and the refractive indices are 2.45 for ZnO, 1.45 for SLG, and 1 for air, the interfaces may have acted as partial mirrors which contributed to the function of ZnO-vnws as optical resonators. On the other hand, there was almost no effect on the improvement of the resolution and intensity, when the thickness ratios of the SrGa₂S₄:Eu layer to the ZnO-vnws were over approximately 2. Further investigations of the effects of the thickness of each layer and the thickness ratios for the bilayer scintillator are needed, but it is evident that the enhancement of the spatial resolution and luminescence intensity by the ZnO-vnws layer is effective when the SrGa₂S₄:Eu layer is relatively thin.

Zoom In Zoom Out Reset image size

SrGa₂S₄:Eu phosphor/ZnO vertical nanowires (ZnO-vnws) bilayer scintillators were successfully fabricated by electrodeposition of ZnO-vnws followed by electrophoresis deposition of SrGa₂S₄:Eu particles. The ZnO-vnws were composed of vertically aligned ZnO nanowires with excellent (0001) out-of-plane orientation, and the SrGa₂S₄:Eu layer was composed of the particles adhered to the ZnO-vnws. The SrGa₂S₄:Eu/ZnO-vnws bilayer scintillators emitted visible light at a wavelength of 533 nm with a lifetime of 0.389 μs and acted as a scintillator for X-ray irradiation. The resolution and luminescence intensity for SrGa₂S₄:Eu/ZnO-vnws bilayer scintillators were improved by introducing the ZnO-vnws compared to those of the ZnO-vnws-free SrGa₂S₄:Eu layer when the thickness ratio of SrGa₂S₄:Eu layer to ZnO-nws length was constructed at around 1.1 ∼1.3. The SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator revealed a spatial resolution of 6.8 μm, which was slightly larger than 5.4 μm of a ready-made LuAG single-crystal scintillator. The SrGa₂S₄:Eu/ZnO-vnws bilayer scintillator with an optimum thickness and thickness ratio is a realistic candidate for a new-generation scintillator with high performance and extended versatility.

This work was partly performed under the Strategic Core Technology Advancement Program (Supporting Industry Program, METI (Ministry of Economy, Trade, and Industry)). The synchrotron radiation experiment was performed with the approval of the Japan Synchrotron Radiation Center (JASR), SPring-8 (approval number, 2019A1629).