Open Access
Add to BibSonomy
Abstract
Surface modification of ceramic Ce-doped Gd3Al2Ga3O12 (Ce:GAGG) was performed by exposing small samples to anhydrous phosphoric acid (H3PO4) under different conditions (temperature and duration) to investigate the effects of chemical polishing treatment. When coupled to a photomultiplier tube (PMT) and used as a radiation detector, chemical treatment for 3 min at 190 °C improved the light (signal) output by 24.8% and energy resolution by 2.5% (percentage point), respectively. This can be attributed to a reduction in surface roughness that enhanced optical properties. Thus, chemical polishing could be a low-cost alternative to mechanical polishing especially for small or complex shaped ceramic scintillators.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The Ce-doped Gd3Al3Ga3O12 (Ce:GAGG) has been widely used in X-ray and γ-ray measurements, owing to its attractive optical and spectral properties [1–4], such as high light output, good energy resolution, high stopping power, and timing performance. It can be also utilized in various fields, such as positron emission tomography, γ-camera, gamma spectroscopy and nondestructive inspection testing (NDT) such as baggage scanning [5,6]. Generally, in high-energy radiation detection, which require larger crystals, bulk single-crystal scintillators are preferred owing to their optical quality that is superior to those of their ceramic counterparts [1,7]. Nonetheless, due to the need for cost effective scintillators, numerous studies have been conducted on the alternatives to single-crystal scintillators [8–13]. However, the inhomogeneity and crystallographic defects of ceramic scintillators, such as lattice distortions and point defects, often degrade their optical performance. Moreover, the intrinsic (e.g., density and hygroscopicity) and extrinsic parameters (e.g., crystallographic defects and irregular stoichiometric compositions) affect the performance of the scintillators by altering their light-collection efficiency.
The translucency of the ceramic scintillator is an undesirable property. One of the efficient ways to fix the microstructures to reduce scattering by crystallographic imperfections is the application of high-temperature annealing. In a previously reported study, we investigated the effect of annealing on ceramic Ce:GAGG [14]. However, annealing alone cannot significantly increase the light output of the ceramic scintillator. Moreover, the transport of light photon through the ceramic structure can be further affected by the surface structure, specifically by its roughness, which often leads to the degradation of the optical performance [15]. The existence of surface heterogeneity in the ceramic scintillators causes unwanted light propagation. Consequently, a refined surface is required to enable the efficient use of these scintillators. Light loss due to surface scattering or bulk scattering of the inorganic scintillator reduces the light-collection efficiency of the photodetector. The application of total integrated scattering (TIS), proposed by Bennett and Proteus, indicates that the incident radiation can be reflected according to the condition of the scintillator’s surface. Therefore, smoothing the scintillator’s surface is necessary to achieve efficient signal transportation, thereby enabling effective radiation detection.
Research on the effect of surface finish is being conducted in various fields [16–18]. In our case, a smooth surface can be induced by mechanical polishing, which often exhibits the best optical (light collection) performance [19]. However, this process is time-consuming, expensive, and possibly challenging to perform on small-sized scintillators. To overcome these drawbacks, several researchers attempted to use an alternative method of chemical etching/polishing to smoothen the surface of rough single-crystal scintillators [20–24]. In all the above studies, the performance (light output) of the chemically polished scintillators showed improvement as compared to that of the as-cut (rough) ones. However, these studies focus on the effects of chemical polishing treatment on highly transparent bulk single-crystal scintillators. Furthermore, studies focusing on chemical polishing treatment on ceramic scintillators are scarce.
To the best of our knowledge, this study is the first to compare the light output enhancement induced by chemical polishing, mechanical polishing, and unpolished treatments on the surface of the ceramic Ce:GAGG samples. Although chemical polishing treatment is not suitable for creating a smooth surface owing to its tendency to leave pores on the surface, it can be used for polishing scintillators with complex shapes. Moreover, the use of the chemical polishing method enables the simultaneous treatment of several scintillators, which saves time and cost. Compared to the single-crystal scintillators, the changes in surface morphology by chemical polishing in ceramic scintillators could influence the luminescent and optical properties. Therefore, we examined the surface morphologies and analyzed the spectroscopic performances to evaluate the feasibility of chemical polishing in improving the light transmission from ceramic scintillators to optical sensors.
2. Experimental setup
2.1 Sample preparation
All surface treatments were performed on ceramic Ce:GAGG scintillators (Chosun Refractories Co. Ltd., Republic of Korea) manufactured using the conventional ceramic processing method. The samples were cut from a single 50 × 50 × 5 mm3 block into 5 × 5 × 2 mm3-sized pieces using a diamond wire saw (South Bay Technology, Model-850), and the as-cut samples were divided into three groups for the following surface treatments: (i) mechanical polishing (three samples), (ii) chemical polishing (nine samples: three pieces exposed to acid for 1 min, three pieces for 3 min, and three pieces for 10 min), and (iii) annealing after chemical polishing (three samples). Each sample was measured before (as-cut, baseline) and after the treatment.
Specifically, mechanical polishing treatment, on all sides of the samples, was performed with 0.1 and 1 µm alumina suspensions. Chemical polishing treatment on the scintillator samples was performed by exposing them to anhydrous or ortho-H3PO4 (85% in volume). In this process, H3PO4 was heated to 190 °C in a silicone oil (Shin-Etsu Chemical Co. Ltd., KF-54) bath [21]. The oil temperature was continuously monitored using a thermocouple, and the samples were exposed to acid for 1, 3, and 10 min. These chemically polished samples were further rinsed with deionized water and dried at room temperature. Figure 1 demonstrates the chemical polishing treatment performed on the samples. In addition, three samples were annealed at 1000 °C in air after chemical polishing for the purpose of comparison [14].
Fig. 1. 5 × 5 × 2 mm3 ceramic Ce:GAGG sample (left); the experimental setup for chemical polishing (etching) using phosphoric acid (right).
2.2 Gamma-ray measurement of ceramic Ce:GAGG and transmittance
Each of the samples was optically coupled to a photomultiplier tube (PMT) assembly (Hamamatsu Photonics, H11934-100), using silicone optical grease (BC-630, Saint-Gobain), in the form of a scintillator radiation detector to evaluate their spectral performance. To minimize signal fluctuations caused by reflectors, a reflector cap was folded out of the enhanced specular reflector (3M Enhanced Specular Reflector, 3M ESR) placed over the samples, instead of individually wrapping the scintillator with reflectors, such as Teflon tape, to reduce measurement variations. This setup was then irradiated with a 137Cs γ-ray source in a dark box to minimize stray light photons. The PMT was subsequently supplied with a voltage of −900 V using a high voltage power supply (AMETEK, ORTEC 556), and the output signal of the PMT was sent to a digitizer (CAEN, DT5730) for data acquisition. The pulse processing was performed on a personal computer. The energy resolution and relative light output (signal amplitude) of the samples were calculated for the 662 keV photopeak of 137Cs by fitting a Gaussian function to the energy spectra.
Finally, light transmission measurements were performed using a UV–Vis–NIR spectrophotometer (Agilent Technologies, Cary 5000) across a photometric range of 200–800 nm (scan rate, 600 nm/min; data interval, 1 nm; and averaging time, 0.1 s).
2.3 Surface morphology and chemical composition of ceramic Ce:GAGG
A quantitative analysis of the treated (polished) sample surfaces was also conducted to evaluate their mechanical- and chemical-induced changes. Moreover, the surface morphologies of the treated samples were observed using a scanning electron microscope (SEM) (SEC’s SNE-4500M Tabletop SEM with 5-Axis Stage Control) to examine the surface microstructure before and after treatments. Figure 5 shows the SEM images of the samples.
The light-collection efficiency of the photodetector was found to be influenced by the surface roughness of the crystal. Consequently, the deviations in the surface roughness of the samples after the treatments were analyzed using a surface profiler (Bruker, Dektak XT stylus profiler). To obtain data points across the entire face of the ceramic Ce:GAGG scintillators, several different positions were selected and scanned.
Furthermore, the changes after chemical polishing of the samples were assessed via X-ray photoelectron spectroscopy (XPS) (Kratos Axis Supra). In addition, the constituents of the sample surfaces were investigated via energy-dispersive X-ray spectrometry (EDS) (JEOL JSM-7610FPlus).
3. Results and discussion
3.1 Gamma-ray measurement of ceramic Ce:GAGG
In Fig. 2, the measured changes in the light output and energy resolution are summarized. After mechanical and chemical polishing, all the ceramic Ce:GAGG samples exhibited higher light output and better energy resolution compared with those at the baseline (as-cut, no treatment), as presented in Table 1. To evaluate the effects of the different surface treatments, three samples of each treatment method were analyzed to obtain statistically meaningful results.
Fig. 2. Relative light output (compared with that at baseline) of 5 × 5 × 2 mm3 ceramic Ce:GAGG samples after the different surface treatments: samples 1–3, chemical polishing treatment for 1 min; samples 4–6, chemical polishing treatment for 3 min; samples 7–9, chemical polishing treatment for 10 min; and samples 10–12, mechanical polishing treatment. The error bars denote the two sigma Gaussian fitting errors.
Table 1. Relative light output and changes in energy resolution (ER) after the different surface treatments.
We observed that the relative light output of the mechanically polished samples increased by 28.4% compared with those at the baseline (as-cut).
In the case of the chemically polished samples, the samples that were exposed to acid for 3 min exhibited a higher relative light output than those exposed for 1 and 10 min. This happened probably because while the surface was improved by the etching process, the surface microstructures of the samples, exposed to acid for more than 10 min, underwent significant changes due to crack or fracture formation, which altered their mechanical properties. Due to the presence of grain boundaries, ceramic scintillators are often more susceptible to chemical polishing than their single-crystal counterparts. Therefore, controlling the treatment duration is crucial to preserve the structural integrity while improving optical properties. Figure 3 presents the 137Cs source energy (pulse height) spectrum of one of the three chemically polished (3-min exposure to acid) ceramic Ce:GAGG samples. We observed that the light output of the chemically polished samples increased by an average of 24.8% compared with those at the baseline (average of three samples), and their energy resolution (full width at half maximum (FWHM)) improved by 2.5% (percentage point) from 16.5% to 14.0% (relative improvement of 15.2%).
Fig. 3. Pulse height spectra of 5 × 5 × 2 mm3 ceramic Ce:GAGG samples coupled to a PMT and irradiated with 137Cs source for one of the samples (red) chemically polished for 3 min and its as-cut baseline (black).
Finally, the relative light output of the annealed ceramic Ce:GAGG sample at 1000 °C after chemical polishing treatment for 3 min increased by 31.1% on average compared with that at the baseline (as-cut). The oxygen deficiencies were compensated for, thus improving light transportation and detection, as reported in a previous study [14]. Therefore, it can be inferred that fixing the microstructures of polycrystalline materials and reducing the surface roughness require the simultaneous use of surface polishing and annealing techniques.
In this study, several different treatment methods were employed, and the corresponding energy resolutions and relative light output results are presented in Table 1. We observed that mechanical polishing yielded the highest light output and the best energy resolution of the samples; however, in the fabrication of thin and small samples or samples with complex shapes, chemical polishing is easier to perform than mechanical polishing. However, there is a tendency for the surface and structure of the chemically polished samples to deteriorate. Therefore, when employing the chemical polishing method, the acid exposure time should be considered according to the crystallographic structure and thickness of the scintillation crystals.
Surface modification via the different treatments explained above changed the light output of ceramic Ce:GAGG. From Fig. 4, it can be observed that mechanical polishing treatment yielded the highest transmittance and the chemically polished samples exhibited better results than the unpolished ones.
Fig. 4. Relative transmittance spectra of random samples. These spectra were obtained using UV–Vis–NIR spectrophotometer (Agilent Technologies, Cary 5000).
Moreover, the modified surface roughness significantly influenced the light transmittance. The intensity of the scintillation photons traveling through the ceramic Ce:GAGG sample is expressed by [25]:
where I0, n, ν, µ, and l denote the input intensity, the number of reflections, the coefficient of the reflection loss for each time, the absorption coefficient, and the light path length, respectively. Thus, it can be inferred that the light output from a ceramic scintillator is dependent on the number of reflections within the sample and at the scintillator–photodetector boundary.3.2 Surface morphology and chemical composition of ceramic Ce:GAGG
To elucidate the improvement in the optical and spectral performances, the crystal surface properties and surface chemistry were investigated. Prior to the chemistry analysis, surface metrology was performed via SEM and profilometry analysis.
The SEM images of the ceramic Ce:GAGG samples, captured before and after the different surface treatments, are presented in Fig. 5. In the as-cut sample (Fig. 5(a)), distinct grain boundaries are visible. The mechanically polished samples (Fig. 5(b)) exhibit the smoothest surface texture among all the samples. In the chemically polished samples (Fig. 5(c)), grain boundaries are also widely visible but are smoothed locally (also see Fig. 6). These grain boundaries are significantly reduced via annealing (Fig. 5(d)).
Fig. 5. Typical SEM images of (a) as-cut, (b) mechanically polished, (c) chemically polished for 3 min, and (d) annealed and chemically polished ceramic 5 × 5 × 2 mm3 Ce:GAGG samples. The scale bar is 5 µm.
Fig. 6. Typical surface profiles at the 5 × 5 mm2 face obtained using Dektak XT Stylus Profiler. (a) (blue) Mechanically polished, (b) (red) chemically polished for 3 min at 190 °C, and (c) (black) as-cut 5 × 5 × 2 mm3 ceramic Ce:GAGG samples.
In Fig. 6 and Table 2, the results of the detailed assessment of the surface roughness are presented. The relevant surface roughness parameters need to be measured to enable the characterization of the surface after the polishing treatment. The results indicate that the process of surface smoothing of ceramic Ce:GAGG reduces the variations in surface roughness (Ra) and waviness (Wa). Compared with the surface roughness of the unpolished as-cut samples, the surface roughness of the mechanically polished samples decreased by 89% and that of the chemically polished samples decreased by 73%.
Table 2. Comparison of the typical profile roughness (Ra) and waviness (Wa) from the surface metrology data after the different surface treatments.a
The surface roughness-induced light scattering can be defined using the TIS as:
Comparing the mechanical and chemical surface treatments as in our case, factors (1) and (2) remain constant; therefore, factor (3), i.e., surface roughness, can be said to be the main factor affecting the light output.
Further, EDS and XPS analysis were conducted on the samples, and the results are presented in Fig. 7 and Fig. 8, respectively. The EDS analysis provides information on the elemental constitution of the samples. In the present case, similar spectra of chemically polished and unpolished samples were observed. However, a minute quantity of phosphorus was also observed on the samples, and the presence and contributions of PO43-, GaPO4, and GdPO4 could be observed after the anhydrous H3PO4 treatment. This indicates that metal ions diffuse during the etching process, although this does not affect the light output.
Fig. 7. SEM–EDS spectra of the samples: (a) as-cut and (b) chemically polished Ce:GAGG samples. The scale bar of the SEM images is 5 µm.
Fig. 8. XPS spectra of the chemically polished and unpolished ceramic Ce:GAGG samples: (a) Gd 3d and (b) P 2p. The y-axis indicates the arbitrary counts.
Furthermore, a significant difference was observed for Gd 3d (1186 eV) and P 2p (134 eV) peak positions between the chemically polished and unpolished samples, as presented in Fig. 8. This is due to the chemical reaction between ceramic Ce:GAGG and H3PO4 acid. Consequently, the chemically polished samples exhibited a small shift in the XPS peaks.
4. Conclusion
In this study, we attempted to perform chemical polishing (etching) treatment on the ceramic Ce:GAGG samples. The mechanical polishing method is commonly employed to smooth the surface of scintillators; however, chemical polishing may also be employed as an alternative owing to its fast treatment time, cost-effectiveness, and suitability for application to scintillators with complex shapes. Although chemical polishing is not new to scintillation radiation detectors, our study presents the outcomes of, possibly, the first effort to remove surface irregularities in ceramic Ce:GAGG via chemical polishing treatment. We found that the treatment time should be shortened to prevent the ceramic Ce:GAGG samples from developing structural inconsistencies. In our 2-mm-thick ceramic Ce:GAGG samples, those chemically polished for 3 min exhibited high light output and good energy resolution. Chemical polishing for more than 3 min (e.g., 10 min or longer) led to the formation of cracks on the surface, which declined the optical and spectral performance of the crystal samples. Finally, we also demonstrated that the chemical treatments followed by high-temperature annealing further improved the light output. Further studies could include optimizing etching temperature/time and varying chemical etchant.
Funding
National Research Foundation of Korea (NRF-2020R1I1A1A01070761, NRF-2020M2A8A4023713, NRF-2020R1A2C2007376, NRF-2020R1C1C1007296).
Disclosures
The authors declare no conflicts of interest.
References
1. K. Kamada, T. Yanagida, T. Endo, K. Tsutumi, Y. Usuki, M. Nikl, Y. Fujimoto, A. Fukabori, and A. Yoshikawa, “2 inch diameter single crystal growth and scintillation properties of Ce:Gd3Al2Ga3O12,” J. Cryst. Growth 352(1), 88–90 (2012). [CrossRef]
2. J. Iwanowska, L. Swiderski, T. Szczesniak, P. Sibczynski, M. Moszynski, M. Grodzicka, K. Kamada, K. Tsutsumi, Y. Usuki, T. Yanagida, and A. Yoshikawa, “Performance of cerium-doped Gd3Al2Ga3O12 (GAGG:Ce) scintillator in gamma-ray spectrometry,” Nucl. Instrum. Methods Phys. Res., Sect. A 712, 34–40 (2013). [CrossRef]
3. K. Kamada, Y. Shoji, V. V. Kochurikhin, S. Okumura, S. Yamamoto, A. Nagura, J. Y. Yeom, S. Kurosawa, Y. Yokota, Y. Ohashi, M. Nikl, and A. Yoshikawa, “Growth and scintillation properties of 3 in. diameter Ce doped Gd3Ga3Al2O12 scintillation single crystal,” J. Cryst. Growth 452, 81–84 (2016). [CrossRef]
4. B. Seitz, N. C. Rivera, and A. G. Stewart, “Energy Resolution and Temperature Dependence of Ce:GAGG Coupled to 3 mm × 3 mm Silicon Photomultipliers,” IEEE Trans. Nucl. Sci. 63(2), 503–508 (2016). [CrossRef]
5. J. Y. Yeom, S. Yamamoto, S. E. Derenzo, V. C. Spanoudaki, K. Kamada, T. Endo, and C. S. Levin, “First performance results of Ce:GAGG scintillation crystals with silicon photomultipliers,” IEEE Trans. Nucl. Sci. 60(2), 988–992 (2013). [CrossRef]
6. K. Kamada, M. Nikl, S. Kurosawa, A. Beitlerova, A. Nagura, Y. Shoji, J. Pejchal, Y. Ohashi, Y. Yokota, and A. Yoshikawa, “Alkali earth co-doping effects on luminescence and scintillation properties of Ce doped Gd3Al2Ga3O12 scintillator,” Opt. Mater. 41, 63–66 (2015). [CrossRef]
7. K. Kamada, S. Kurosawa, P. Prusa, M. Nikl, V. V. Kochurikhin, T. Endo, K. Tsutumi, H. Sato, Y. Yokota, K. Sugiyama, and A. Yoshikawa, “Cz grown 2-in. size Ce:Gd3(Al,Ga)5O12 single crystal; Relationship between Al, Ga site occupancy and scintillation properties,” Opt. Mater. 36(12), 1942–1945 (2014). [CrossRef]
8. S. R. Podowitz, R. M. Gaumé, W. T. Hong, A. Laouar, and R. S. Feigelson, “Fabrication and properties of translucent SrI2 and Eu:SrI2 scintillator ceramics,” IEEE Trans. Nucl. Sci. 57(6), 3827–3835 (2010). [CrossRef]
9. Y. Wang, G. Baldoni, W. H. Rhodes, C. Brecher, A. Shah, U. Shirwadkar, J. Glodo, N. Cherepy, and S. Payne, “Transparent garnet ceramic scintillators for gamma-ray detection,” Proc. SPIE 8507, 850717 (2012). [CrossRef]
10. T. Yanagida, K. Kamada, Y. Fujimoto, H. Yagi, and T. Yanagitani, “Comparative study of ceramic and single crystal Ce:GAGG scintillator,” Opt. Mater. 35(12), 2480–2485 (2013). [CrossRef]
11. L. B. Kong, Y. Huang, W. Que, T. Zhang, S. Li, J. Zhang, Z. Dong, and D. Tang, “Transparent Ceramic Materials,” in Transparent Ceramics. Topics in Mining, Metallurgy and Materials Engineering (Springer, Cham, 2015) pp. 29–91.
12. C. Greskovich and S. Duclos, “Ceramic scintillators,” Annu. Rev. Mater. Sci. 27(1), 69–88 (1997). [CrossRef]
13. N. J. Cherepy, J. D. Kuntz, J. J. Roberts, T. A. Hurst, O. B. Drury, R. D. Sanner, T. M. Tillotson, and S. A. Payne, “Transparent ceramic scintillator fabrication, properties, and applications,” Proc. SPIE 7079, 70790X (2008). [CrossRef]
14. C. Park, M. Ullah, C. Kim, S. Cho, and J. Y. Yeom, “Investigation of Optical Properties of Ceramic Ce:GAGG by High Temperature Annealing,” J. Korean Phys. Soc. 75(12), 962–967 (2019). [CrossRef]
15. J. E. Harvey, N. Choi, S. Schroeder, and A. Duparré, “Total integrated scatter from surfaces with arbitrary roughness, correlation widths, and incident angles,” Opt. Eng. 51(1), 013402 (2012). [CrossRef]
16. S. Schröder, A. Duparré, L. Coriand, A. Tünnermann, D. H. Penalver, and J. E. Harvey, “Modeling of light scattering in different regimes of surface roughness,” Opt. Express 19(10), 9820–9835 (2011). [CrossRef]
17. Z. Zhenrong, Z. Jing, and G. Peifu, “Roughness characterization of well-polished surfaces by measurements of light scattering distribution,” Opt. Appl. XL 4, 811–818 (2010).
18. A. A. Maradudin, Structured surfaces as optical metamaterials, 1st ed. (Cambridge University Press, 2011).
19. I. V. Kilimchuk, V. A. Tarasov, and I. D. Vlasova, “Study of light collection as a function of scintillator surface roughness,” Radiat. Meas. 45(3-6), 383–385 (2010). [CrossRef]
20. K. Kurashige, Y. Kurata, H. Ishibashi, and K. Susa, “Surface polishing of GSO scintillator using chemical process,” IEEE Trans. Nucl. Sci. 45(3), 522–524 (1998). [CrossRef]
21. R. Slates, A. Chatziioannou, B. Fehlberg, T. Lee, and S. Cherry, “Chemical polishing of LSO crystals to increase light output,” Proc. IEEE 2, 939–942 (1999). [CrossRef]
22. J. S. Huber, W. W. Moses, M. S. Andreaco, and O. Petterson, “An LSO scintillator array for a PET detector module with depth of interaction measurement,” Proc. IEEE 48(3), 684–688 (2001). [CrossRef]
23. Y. Shao, K. Meadors, R. W. Silverman, R. Farrell, L. Cirignano, R. Grazioso, K. S. Shah, and S. R. Cherry, “Dual APD Array Readout of LSO Crystals: Optimization of Crystal Surface Treatment,” IEEE Trans. Nucl. Sci. 49(3), 649–654 (2002). [CrossRef]
24. J. Martin and W. W. Moses, “Measuring light reflectance of BGO crystal surfaces,” IEEE Trans. Nucl. Sci. 55(5), 2443–2449 (2008). [CrossRef]
25. H. Ishibashi, S. Akiyama, and M. Ishii, “Effect of surface roughness and crystal shape on performance of bismuth germanate scintillators,” Jpn. J. Appl. Phys. 25(9), 1435–1438 (1986). [CrossRef]
