- Email: [email protected]
Spatial resolution and cathodoluminescence intensity dependence on acceleration voltage in electron beam excitation assisted optical microscopy using Y2O3:Eu3+ film
Ultramicroscopy 182 (2017) 212–215
Contents lists available at ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
Spatial resolution and cathodoluminescence intensity dependence on acceleration voltage in electron beam excitation assisted optical microscopy using Y2 O3 :Eu3+ film Yu Masuda a, Masashi Kamiya b, Atsushi Sugita b, Wataru Inami c,d,∗, Yoshimasa Kawata c,d, Hiroko Kominami d, Yoichiro Nakanishi d a
Graduate School of Science and Technology, Shizuoka University, 3-5-1, Johoku, Naka, Hamamatsu 4328561, Japan Department of Materials Science, Shizuoka University, 3-5-1, Johoku, Naka, Hamamatsu 4328561, Japan c CREST, Japan Science and Technology Agency, Japan d Research Institute of Electrics, Shizuoka University, 3-5-1, Johoku, Naka, Hamamatsu 4328561, Japan b
a r t i c l e
i n f o
Article history: Received 28 April 2017 Revised 5 July 2017 Accepted 9 July 2017 Available online 13 July 2017
a b s t r a c t This study presents relationship between acceleration voltage and spatial resolution of electron-beam assisted (EXA) optical microscope. The nanometric illumination light sources of the present EXA microscope was red-emitting cathodoluminescence (CL) in the Y2 O3 :Eu3+ thin film excited by focused electron beam. Our experimental results demonstrated that the spatial resolutions of the EXA microscope were higher as the acceleration voltage was higher. We managed to make images of the scattered gold particles with approximately 90 nm-resolutions at the voltages higher than 20 kV. The dependence of the spatial resolution on the acceleration voltage was explained by the distribution of simulated electron scattering trajectories in the luminescent thin film. © 2017 Published by Elsevier B.V.
1. Introduction The high resolution optical microscope contributes the development in the field of medical science, pharmacy, materials science, and nanotechnology. Recently, high resolution optical imaging techniques such as stimulated emission depletion (STED) [1], photoactivated localization microscope (PALM) [2], stochastic optical reconstruction microscope (STORM) [3], structured illumination microscopy (SIM) [4,5], nanoscale light source [6,7], and saturated excitation (SAX) microscope [8] have demonstrated to exceed the diffraction limit of light, and they have been used to understand the functions of biological cells at nanometric spatial resolution. These advanced microscopic technologies relied on fluorophores which were specially designed for the observation, and their applications are limited for the specific materials. We proposed the new class of high resolution optical microscope (the EXA microscope) in which the focused electron beam was used for exciting nanometric light spot of cathodoluminescence. The microscopes enabled us to observe the unstained biological specimens with higher spatial resolution than the diffrac-
∗
Corresponding author. E-mail address: [email protected] (W. Inami).
http://dx.doi.org/10.1016/j.ultramic.2017.07.010 0304-3991/© 2017 Published by Elsevier B.V.
tion limit by detecting the transmitted or scattered light from nanometric light spot. We could observe the latex spheres with 50 nm spatial resolution using the EXA microscope [9]. The performance of the EXA microscope is mainly depending on the properties of the luminescent thin film. The initial model of the EXA microscope was built up with amorphous Si3 N4 membrane which was separating the air and vacuum atmosphere. The Si3 N4 membrane emitted nanometric CL by electron beam irradiation. However, the ultraviolet CL lights from Si3 N4 membrane was toxic to most of the living specimen, including biological cells and tissues. Therefore, it is more desirable to operate the EXA microscope by using visible CL friendly to biological samples. In our previous study, Y2 O3 :Eu3+ thin film was prepared for meeting the requirements[10]. The Y2 O3 :Eu3+ have been studied for a long decades for their applications to cathode ray tubes or flat panel displays because of the high quantum efficiency in red-color CL [11–13]. The strongest CL spectra of Y2 O3 :Eu3+ corresponds to the 5 D0 →7 F2 transition of Eu3+ . Additionally, the internal 4f shell is shielded by the outer 5s and 5d shell in Eu3+ , so the optical absorption and emission spectra are quite insensitive to external perturbations [14]. In the previous paper [10], the preparations of the Y2 O3 :Eu3+ films on Si3 N4 membrane by using electron beam evaporation techniques is explained and we concentrated on the discussions about the optimal conditions for fabricating
Y. Masuda et al. / Ultramicroscopy 182 (2017) 212–215
Fig. 1. The principle of electron beam excitation. The focused electron beam is irradiated to the Y2 O3 :Eu3+ film to generate a nanometer sized CL spot. The Y2 O3 :Eu3+ film plays also an role of separating the air and vacuum environments.
luminescent films operated in the EXA microscopes. However, the operation performances such as the spatial resolutions of the EXA microscopes were not discussed. In this paper, we discuss the performance of the EXA microscope equipped with Y2 O3 :Eu3+ thin film from the viewpoints of the spatial resolution against the acceleration voltage. The spatial resolution of the microscope was characterized by the EXA microscopic images of the 100 nm-diameter gold particles at the different acceleration voltages. The experimental results were supported by the numerical simulation using Monte Carlo method [15]. We found out an optimal acceleration voltage to achieve both of the greater fluorescence quantum yield and higher spatial resolution. 2. Principle of the electron beam excitation assisted optical microscope Fig. 1 shows the schematics of the structures and the operational principle of the EXA microscope. The microscope consists of a scanning electron microscope and a conventional optical microscope. The Y2 O3 :Eu3+ thin film is deposited on the Si3 N4 membrane and this membrane separates the air and vacuum environments. The surface of the Si3 N4 film is faced to the scanning electron microscope, whereas the Y2 O3 :Eu3+ side is faced to the optical microscope. The focused electron beam is irradiated onto the luminescent film at a few nm spot sizes. The emission is attributed to the electron transitions in 4f orbital in Eu3+ ion which is excited by the electrons in the luminescent thin film. We reported that the distribution of the luminescent field on the film surface was much narrower than the diffraction limit, if the film thickness was much thinner than a few hundred nm and the kinetic energy of the electron was higher than a few keV [15]. The CL light illuminates the specimen in the area smaller than the diffraction limit on the luminescent thin film. As a result, we can obtain the optical microscopic image with the resolution greater than the diffraction limit by detecting the transmitted or scattered light from the specimen. 3. Experimental setup The Y2 O3 :Eu3+ thin film was grown on the Si3 N4 membrane (Silson Ltd.) by using an electron beam evaporation method. The Si3 N4 membrane with 50 nm-thickness was supported by the Si substrate with 200 μm-thickness, in which 50 μm-square window was etched.
213
Fig. 2. The CL intensity from Y2 O3 :Eu3+ thin film against the acceleration voltage. The acceleration voltages were tuned between 1 and 30 kV. The CL intensities increased at the voltage higher than 2 kV threshold and they had a peak at 7 kV, above which they were rather reduced. The peak CL intensities was 1.4 × 109 counts. CL intensity decrease with the acceleration voltage higher than 9 kV.
The fabrication procedure of the Y2 O3 :Eu3+ thin film is explained as follows. The pellets consisted the Y2 O3 and Eu2 O3 powders were sintered at 10 0 0 °C. The concentration of Eu3+ ions in the mixture was 2 mol%. The deposition was carried out under the condition with 4 kV, 5 mA and 10−3 Pa. The Y2 O3 :Eu3+ with 200 nm-thickness was deposited on 50 nm-thick S3 N4 membrane. The EXA microscope was fabricated using a thermal-type scanning electron microscope (SEM) (JSM-6390, JOEL Ltd.). The electron beam was irradiated to the Y2 O3 :Eu3+ film from the Si3 N4 membrane side for CL excitation. The CL was detected by a photomultiplier tube (R7400-U20, Hamamatsu Photonics K. K.). The image was reconstructed from the signals using a computer. We observed the 100 ± 8 nm gold particles (EMGC100, Funakoshi Co. Ltd.) to evaluate the spatial resolution of the EXA microscope with Y2 O3 :Eu3+ film. The electron beam trajectory in the Si3 N4 and Y2 O3 :Eu3+ film was simulated by using Monte Carlo methods. The details of the simulation methods were reported in our previous work [9,10,16]. 4. Result and discussion Fig. 2 shows the dependence of CL intensity from 200 nm Y2 O3 :Eu3+ film on the acceleration voltage. The acceleration voltage was changed between 1 and 30 kV. The CL intensities are represented in the detected photon numbers per second. It can be seen that the CL intensity increases with increasing voltage and has maximum at around 7 kV. It meant that the most of the kinetic energies of the electrons were consumed for the CL excitations. At more than 9 kV, the CL intensities are lower than that at 7 kV. It seems that the number of electrons penetrated through the Y2 O3 :Eu3+ layer increases due to the higher acceleration voltages, thus the energy interaction for CL excitation decreases. The required CL intensity was calculated to estimate the enough emission light for the dynamic image in the EXA microscope. The signal to noise ratio (S/N) of the dynamic image is decreased because the exposure time of dynamic image is short due to the high flame rate. The S/N of the image is normally required more than 5 to identify the object [17,18]. We calculated the S/N by considering the dark current and shot noise of PMT. The S/N is indicated by the following formula [19].
S/N ≈ 1.75 × 103
Ik /B
(1)
It is clear that the S/N is proportional to the square root of the cathode current Ik and is inversely proportional to the square root of the bandwidth B. By using the eq. (1), for example, 8.19 × 108 photons/s are required to obtain the S/N of 5 with
214
Y. Masuda et al. / Ultramicroscopy 182 (2017) 212–215
Fig. 3. The EXA microscope (left) and the secondary scattered electron images (right) of 100 nm gold nano particles dispersed on Y2 O3 :Eu3+ film. The acceleration voltages were (a) 15, (b) 20 and (c) 25 kV, respectively. The line profiles of the solid lines in (a), (b) and (c) were shown in (d), (e) and (f), respectively.
256 pixels × 256 pixels, and 15 fps. Since the 200 nm Y2 O3 :Eu3+ on SiN film can emits about 1.4 × 109 photons/s from Fig. 2, it is possible to get the image of S/N > 6.5 with 256 pixels × 256 pixels and 15 fps. Therefore, the 200 nm Y2 O3 :Eu3+ film can emit CL light which intensity is strong enough to achieve the high temporal resolution. Fig. 3 shows the EXA microscope images and reflection electron images of 100 nm gold particles observed with (a) 15 kV, (b) 20 kV and (c) 25 kV of acceleration voltages. The contrasts are reversed between the EXA and the reflection electron images. The particles in the EXA image are darker than the background because the CL lights were blocked by the particles. On the other hands, the particles in the reflection electron image are brighter because the electrons scatter much more at the gold particles than those of the surrounding environments. The distributions of each particle are corresponding in the EXA and the reflection electron images. The line profiles on the solid line in the EXA images are also shown. The FWHMs of the distribution were 210 nm for 15 kV, 96 nm for 20 kV, and 93 nm for 25 kV. Fig. 4 shows the averaged FWHM of the line profiles of the gold particles against the acceleration voltage. The data were not shown at the voltages lower than 10 kV, because the images were not obtained with the sufficiently high S/N at these regions. As shown in the Fig. 4, the averaged FWHM was 210 nm at 15 kV acceleration voltage and it was much wider than the particle size. At the higher than 20 kV, the FWHM is in 93 to 97 nm. This value is in good agreement with the average particle size of the gold particles. The EXA microscope was operated with the CL with 610 nm. The diffraction limit of light at this wavelength is approximately
Fig. 4. The averaged FWHM of line profile of 100 nm-diameter gold particles against the acceleration voltage.
300 nm. Hence, the EXA microscope with the Y2 O3 :Eu3+ film had the spatial resolutions better than the diffraction limits. Fig. 5 presents the numerically simulated trajectories of the inelastically scattered electrons. The concentration of Eu in the Y2 O3 :Eu3+ was 2 mol%. Because of such a low concentration, the calculations were conducted by assuming the pure Y2 O3 :Eu3+ crystalline structures. The acceleration voltages are set at 10 kV and 20 kV and the trajectories of 200 electrons were calculated.
Y. Masuda et al. / Ultramicroscopy 182 (2017) 212–215
215
In future work, we need to discuss about biocompatibility of the Y2 O3 :Eu3+ film and the damage to the specimen by transmitted electrons for bioimaging. The EXA microscope could be a promising tool for high-resolution imaging of living cells. Acknowledgments
Fig. 5. Monte Carlo simulation result of electron trajectories in the 50 nm-Si3 N4 and 200 nm-Y2 O3 :Eu3+ layers. The calculations were performed at 10 and 20 kV. The trajectories of 200 electrons were calculated. The scale bar indicates 50 nm.
The electrons started the scattering on the bottom surface of the Si3 N4 layer and then traveled into the 50 nm-Si3 N4 and 200 nmY2 O3 :Eu3+ layers, successively. At 10 kV, the electrons are scattered and spread widely in the Y2 O3 :Eu3+ layer, thus the CL spot might be generated inside of the Y2 O3 :Eu3+ layer. On the other hand, at 20 kV, electrons traveled straightly in the Y2 O3 :Eu3+ layer and the almost electrons arrived on the top surface of the Y2 O3 :Eu3+ layer. In this case, the CL spot could be smaller than that at 10 kV and be generated near the surface of the Y2 O3 :Eu3+ layer. Hence, the specimen can be illuminated by the smaller CL spot and the high spatial resolution can be achieved. The higher spatial resolution at higher acceleration voltage was in a good agreement with the observation result in Figs. 3 and 4. 6. Conclusions In this study, we presented the relationship between acceleration voltages and spatial resolutions in EXA microscope. We fabricated the 200 nm Y2 O3 :Eu3+ film for the EXA microscope and measured the CL intensity of the film depends on the acceleration voltage of the electron beam. The CL intensity indicates a peak at 7 kV because almost all of the electrons scatter in the Y2 O3 :Eu3+ film and the kinetic energies of the electrons are consumed for the CL excitations. At the voltages higher than 9 kV, the CL intensities decrease because the electrons penetrates through the Y2 O3 :Eu3+ film and the energy interaction for CL excitation decreases. We also observed 100 nm gold particles with the EXA microscope using the Y2 O3 :Eu3+ thin film at several acceleration voltages. We found that the FWHM of the 100 nm gold particle decreases with higher acceleration voltage because the electron scattering decreases. The FWHM of the 100 nm gold particles were 96.5 nm at 20 kV. We achieved the high spatial resolution than the diffraction limit at the voltages higher than 20 kV. According the simulation result, at more than 20 kV, almost all of the electrons traveled straightly to the top surface of the Y2 O3 :Eu3+ layer and the spot size at the Y2 O3 :Eu3+ surface was smaller. It could be confirmed successfully by the simulation that the spatial resolution of the EXA microscope depends on the acceleration voltage. The spatial resolution increase with higher acceleration voltages from simulation result. As another method, it is possible to obtain a higher resolution by constructing a structure on a fluorescent thin film [20].
This work has been supported by JST-CREST (Japan Science and Technology Agency, Core Research for Evolutionary Science and Technology). References [1] S.W. Hell, J. Wichmann, Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy, Opt. Lett. 19 (780) (1994) 780–782. [2] E. Betzig, G.H. Patterson, R. Sougrat, O.W. Lindwasser, S. Olenych, J.S. Bonifacino, M.W. Davidson, J. Lippincott-Schwartz, H.F. Hess, Imaging intracellular fluorescent proteins at nanometer resolution, Science 313 (5793) (2006) 1642–1645. [3] M.J. Rust, M. Bates, X. Zhuang, Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM), Nat. Methods 3 (10) (2006) 793–796. [4] R. Heintzmann, T.M. Jovin, C. Cremer, Saturated patterned excitation microscopy—a concept for optical resolution improvement, J. Opt. Soc. Am. A 19 (8) (2002) 1599–1609. [5] M.G.L. Gustafsson, Nonlinear structured-illumination microscopy: wide-field fluorescence imaging with theoretically unlimited resolution, J. Proc. Natl. Acad. Sci. 102 (37) (2005) 13081–13086. [6] G. Adamo, K.F. MacDonald, Y.H. Fu, C-M. Wang, D.P. Tsai, F.J. García de Abajo, N.I. Zheludev, Light well: a tunable free-electron light source on a chip, Phys. Rev. Lett. 103 (2009) 113901. [7] G. Adamo, K.F. MacDonald, Y.H. Fu, D.P. Tsai, F.J. García de Abajo, N.I. Zheludev, Tuneable electron-beam-driven nanoscale light source, J. Opt. 12 (2) (2010) 024012. [8] K. Fujita, M. Kobayashi, S. Kawano, M. Yamanaka, S. Kawata, High-resolution confocal microscopy by saturated excitation of fluorescence, Phys. Rev. Lett. 99 (2007) 228105. [9] W. Inami, K. Nakajima, A. Miyakawa, Y. Kawata, Electron beam excitation assisted optical microscope with ultra-high resolution, Opt. Express 18 (12) (2010) 12897–12902. [10] A. Sugita, M. Kamiya, C. Morita, A. Miyake, Y. Nawa, Y. Masuda, W. Inami, H. Kominami, Y. Nakanishi, Y. Kawata, Nanometric light spots of cathodoluminescence in Y2 O3 :Eu3+ phosphor thin films excited by focused electron beams as ultra-small illumination source for high-resolution optical microscope, Opt. Mater. Express 4 (1) (2013) 155–161. [11] S.H. Cho, S.H. Kwon, J.S. Yoo, C.W. Oh, J.D. Lee, K.J. Hong, S.J. Kwon, Cathodoluminescent characteristics of a spherical Y2 O3 :Eu phosphor screen for field emission display application, J. Electrochem. Soc. 147 (8) (20 0 0) 3143–3147. [12] S.H. Shin, J.H. Kang, D.Y. Jeon, D.S. Zang, Enhancement of cathodoluminescence intensities of Y2 O3 :Eu and Gd2 O3 :Eu phosphors by incorporation of Li ions, J. Lumin. 114 (3–4) (2005) 275–280. [13] G. Jia, M. Yang, Y. Song, H. You, H. Zhang, General and facile method to prepare uniform Y2O3:Eu hollow microspheres, Cryst. Growth Des. 9 (1) (2009) 301–307. [14] B.G. Yacobi, D.B. Holt, Cathodoluminescence Microscopy of Inorganic Solids, Plenum, 1990. [15] D.C. Joy, Monte Carlo Modeling For Electron Microscopy and Microanalysis, OXFORD, 1995. [16] M. Fukuta, W. Inami, A. Ono, Y. Kawata, Intensity distribution analysis of cathodoluminescent using the energy loss distribution of electrons, Ultramicroscopy 160 (2016) 225–229. [17] K.K. Hamamatsu Photonics, Photomultiplier Tubes Basic and Applications, third ed., 2007. [18] A. Rose, in: Advances in Electronics, 1, 1948, pp. 131–166. [19] N. de Jonge, N. Poirier-Demers, H. Demers, D. Peckys, D. Drouin, Nanometer-resolution electron microscopy through micrometers-thick water layers, Ultramicroscopy 110 (9) (2010) 1114–1119. [20] Y.-H. Lin, D.P. Tsai, Near-field scanning optical microscopy using a super-resolution cover glass slip, Opt. Express 20 (15) (2010) 16205–16211.
Contents lists available at ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
Spatial resolution and cathodoluminescence intensity dependence on acceleration voltage in electron beam excitation assisted optical microscopy using Y2 O3 :Eu3+ film Yu Masuda a, Masashi Kamiya b, Atsushi Sugita b, Wataru Inami c,d,∗, Yoshimasa Kawata c,d, Hiroko Kominami d, Yoichiro Nakanishi d a
Graduate School of Science and Technology, Shizuoka University, 3-5-1, Johoku, Naka, Hamamatsu 4328561, Japan Department of Materials Science, Shizuoka University, 3-5-1, Johoku, Naka, Hamamatsu 4328561, Japan c CREST, Japan Science and Technology Agency, Japan d Research Institute of Electrics, Shizuoka University, 3-5-1, Johoku, Naka, Hamamatsu 4328561, Japan b
a r t i c l e
i n f o
Article history: Received 28 April 2017 Revised 5 July 2017 Accepted 9 July 2017 Available online 13 July 2017
a b s t r a c t This study presents relationship between acceleration voltage and spatial resolution of electron-beam assisted (EXA) optical microscope. The nanometric illumination light sources of the present EXA microscope was red-emitting cathodoluminescence (CL) in the Y2 O3 :Eu3+ thin film excited by focused electron beam. Our experimental results demonstrated that the spatial resolutions of the EXA microscope were higher as the acceleration voltage was higher. We managed to make images of the scattered gold particles with approximately 90 nm-resolutions at the voltages higher than 20 kV. The dependence of the spatial resolution on the acceleration voltage was explained by the distribution of simulated electron scattering trajectories in the luminescent thin film. © 2017 Published by Elsevier B.V.
1. Introduction The high resolution optical microscope contributes the development in the field of medical science, pharmacy, materials science, and nanotechnology. Recently, high resolution optical imaging techniques such as stimulated emission depletion (STED) [1], photoactivated localization microscope (PALM) [2], stochastic optical reconstruction microscope (STORM) [3], structured illumination microscopy (SIM) [4,5], nanoscale light source [6,7], and saturated excitation (SAX) microscope [8] have demonstrated to exceed the diffraction limit of light, and they have been used to understand the functions of biological cells at nanometric spatial resolution. These advanced microscopic technologies relied on fluorophores which were specially designed for the observation, and their applications are limited for the specific materials. We proposed the new class of high resolution optical microscope (the EXA microscope) in which the focused electron beam was used for exciting nanometric light spot of cathodoluminescence. The microscopes enabled us to observe the unstained biological specimens with higher spatial resolution than the diffrac-
∗
Corresponding author. E-mail address: [email protected] (W. Inami).
http://dx.doi.org/10.1016/j.ultramic.2017.07.010 0304-3991/© 2017 Published by Elsevier B.V.
tion limit by detecting the transmitted or scattered light from nanometric light spot. We could observe the latex spheres with 50 nm spatial resolution using the EXA microscope [9]. The performance of the EXA microscope is mainly depending on the properties of the luminescent thin film. The initial model of the EXA microscope was built up with amorphous Si3 N4 membrane which was separating the air and vacuum atmosphere. The Si3 N4 membrane emitted nanometric CL by electron beam irradiation. However, the ultraviolet CL lights from Si3 N4 membrane was toxic to most of the living specimen, including biological cells and tissues. Therefore, it is more desirable to operate the EXA microscope by using visible CL friendly to biological samples. In our previous study, Y2 O3 :Eu3+ thin film was prepared for meeting the requirements[10]. The Y2 O3 :Eu3+ have been studied for a long decades for their applications to cathode ray tubes or flat panel displays because of the high quantum efficiency in red-color CL [11–13]. The strongest CL spectra of Y2 O3 :Eu3+ corresponds to the 5 D0 →7 F2 transition of Eu3+ . Additionally, the internal 4f shell is shielded by the outer 5s and 5d shell in Eu3+ , so the optical absorption and emission spectra are quite insensitive to external perturbations [14]. In the previous paper [10], the preparations of the Y2 O3 :Eu3+ films on Si3 N4 membrane by using electron beam evaporation techniques is explained and we concentrated on the discussions about the optimal conditions for fabricating
Y. Masuda et al. / Ultramicroscopy 182 (2017) 212–215
Fig. 1. The principle of electron beam excitation. The focused electron beam is irradiated to the Y2 O3 :Eu3+ film to generate a nanometer sized CL spot. The Y2 O3 :Eu3+ film plays also an role of separating the air and vacuum environments.
luminescent films operated in the EXA microscopes. However, the operation performances such as the spatial resolutions of the EXA microscopes were not discussed. In this paper, we discuss the performance of the EXA microscope equipped with Y2 O3 :Eu3+ thin film from the viewpoints of the spatial resolution against the acceleration voltage. The spatial resolution of the microscope was characterized by the EXA microscopic images of the 100 nm-diameter gold particles at the different acceleration voltages. The experimental results were supported by the numerical simulation using Monte Carlo method [15]. We found out an optimal acceleration voltage to achieve both of the greater fluorescence quantum yield and higher spatial resolution. 2. Principle of the electron beam excitation assisted optical microscope Fig. 1 shows the schematics of the structures and the operational principle of the EXA microscope. The microscope consists of a scanning electron microscope and a conventional optical microscope. The Y2 O3 :Eu3+ thin film is deposited on the Si3 N4 membrane and this membrane separates the air and vacuum environments. The surface of the Si3 N4 film is faced to the scanning electron microscope, whereas the Y2 O3 :Eu3+ side is faced to the optical microscope. The focused electron beam is irradiated onto the luminescent film at a few nm spot sizes. The emission is attributed to the electron transitions in 4f orbital in Eu3+ ion which is excited by the electrons in the luminescent thin film. We reported that the distribution of the luminescent field on the film surface was much narrower than the diffraction limit, if the film thickness was much thinner than a few hundred nm and the kinetic energy of the electron was higher than a few keV [15]. The CL light illuminates the specimen in the area smaller than the diffraction limit on the luminescent thin film. As a result, we can obtain the optical microscopic image with the resolution greater than the diffraction limit by detecting the transmitted or scattered light from the specimen. 3. Experimental setup The Y2 O3 :Eu3+ thin film was grown on the Si3 N4 membrane (Silson Ltd.) by using an electron beam evaporation method. The Si3 N4 membrane with 50 nm-thickness was supported by the Si substrate with 200 μm-thickness, in which 50 μm-square window was etched.
213
Fig. 2. The CL intensity from Y2 O3 :Eu3+ thin film against the acceleration voltage. The acceleration voltages were tuned between 1 and 30 kV. The CL intensities increased at the voltage higher than 2 kV threshold and they had a peak at 7 kV, above which they were rather reduced. The peak CL intensities was 1.4 × 109 counts. CL intensity decrease with the acceleration voltage higher than 9 kV.
The fabrication procedure of the Y2 O3 :Eu3+ thin film is explained as follows. The pellets consisted the Y2 O3 and Eu2 O3 powders were sintered at 10 0 0 °C. The concentration of Eu3+ ions in the mixture was 2 mol%. The deposition was carried out under the condition with 4 kV, 5 mA and 10−3 Pa. The Y2 O3 :Eu3+ with 200 nm-thickness was deposited on 50 nm-thick S3 N4 membrane. The EXA microscope was fabricated using a thermal-type scanning electron microscope (SEM) (JSM-6390, JOEL Ltd.). The electron beam was irradiated to the Y2 O3 :Eu3+ film from the Si3 N4 membrane side for CL excitation. The CL was detected by a photomultiplier tube (R7400-U20, Hamamatsu Photonics K. K.). The image was reconstructed from the signals using a computer. We observed the 100 ± 8 nm gold particles (EMGC100, Funakoshi Co. Ltd.) to evaluate the spatial resolution of the EXA microscope with Y2 O3 :Eu3+ film. The electron beam trajectory in the Si3 N4 and Y2 O3 :Eu3+ film was simulated by using Monte Carlo methods. The details of the simulation methods were reported in our previous work [9,10,16]. 4. Result and discussion Fig. 2 shows the dependence of CL intensity from 200 nm Y2 O3 :Eu3+ film on the acceleration voltage. The acceleration voltage was changed between 1 and 30 kV. The CL intensities are represented in the detected photon numbers per second. It can be seen that the CL intensity increases with increasing voltage and has maximum at around 7 kV. It meant that the most of the kinetic energies of the electrons were consumed for the CL excitations. At more than 9 kV, the CL intensities are lower than that at 7 kV. It seems that the number of electrons penetrated through the Y2 O3 :Eu3+ layer increases due to the higher acceleration voltages, thus the energy interaction for CL excitation decreases. The required CL intensity was calculated to estimate the enough emission light for the dynamic image in the EXA microscope. The signal to noise ratio (S/N) of the dynamic image is decreased because the exposure time of dynamic image is short due to the high flame rate. The S/N of the image is normally required more than 5 to identify the object [17,18]. We calculated the S/N by considering the dark current and shot noise of PMT. The S/N is indicated by the following formula [19].
S/N ≈ 1.75 × 103
Ik /B
(1)
It is clear that the S/N is proportional to the square root of the cathode current Ik and is inversely proportional to the square root of the bandwidth B. By using the eq. (1), for example, 8.19 × 108 photons/s are required to obtain the S/N of 5 with
214
Y. Masuda et al. / Ultramicroscopy 182 (2017) 212–215
Fig. 3. The EXA microscope (left) and the secondary scattered electron images (right) of 100 nm gold nano particles dispersed on Y2 O3 :Eu3+ film. The acceleration voltages were (a) 15, (b) 20 and (c) 25 kV, respectively. The line profiles of the solid lines in (a), (b) and (c) were shown in (d), (e) and (f), respectively.
256 pixels × 256 pixels, and 15 fps. Since the 200 nm Y2 O3 :Eu3+ on SiN film can emits about 1.4 × 109 photons/s from Fig. 2, it is possible to get the image of S/N > 6.5 with 256 pixels × 256 pixels and 15 fps. Therefore, the 200 nm Y2 O3 :Eu3+ film can emit CL light which intensity is strong enough to achieve the high temporal resolution. Fig. 3 shows the EXA microscope images and reflection electron images of 100 nm gold particles observed with (a) 15 kV, (b) 20 kV and (c) 25 kV of acceleration voltages. The contrasts are reversed between the EXA and the reflection electron images. The particles in the EXA image are darker than the background because the CL lights were blocked by the particles. On the other hands, the particles in the reflection electron image are brighter because the electrons scatter much more at the gold particles than those of the surrounding environments. The distributions of each particle are corresponding in the EXA and the reflection electron images. The line profiles on the solid line in the EXA images are also shown. The FWHMs of the distribution were 210 nm for 15 kV, 96 nm for 20 kV, and 93 nm for 25 kV. Fig. 4 shows the averaged FWHM of the line profiles of the gold particles against the acceleration voltage. The data were not shown at the voltages lower than 10 kV, because the images were not obtained with the sufficiently high S/N at these regions. As shown in the Fig. 4, the averaged FWHM was 210 nm at 15 kV acceleration voltage and it was much wider than the particle size. At the higher than 20 kV, the FWHM is in 93 to 97 nm. This value is in good agreement with the average particle size of the gold particles. The EXA microscope was operated with the CL with 610 nm. The diffraction limit of light at this wavelength is approximately
Fig. 4. The averaged FWHM of line profile of 100 nm-diameter gold particles against the acceleration voltage.
300 nm. Hence, the EXA microscope with the Y2 O3 :Eu3+ film had the spatial resolutions better than the diffraction limits. Fig. 5 presents the numerically simulated trajectories of the inelastically scattered electrons. The concentration of Eu in the Y2 O3 :Eu3+ was 2 mol%. Because of such a low concentration, the calculations were conducted by assuming the pure Y2 O3 :Eu3+ crystalline structures. The acceleration voltages are set at 10 kV and 20 kV and the trajectories of 200 electrons were calculated.
Y. Masuda et al. / Ultramicroscopy 182 (2017) 212–215
215
In future work, we need to discuss about biocompatibility of the Y2 O3 :Eu3+ film and the damage to the specimen by transmitted electrons for bioimaging. The EXA microscope could be a promising tool for high-resolution imaging of living cells. Acknowledgments
Fig. 5. Monte Carlo simulation result of electron trajectories in the 50 nm-Si3 N4 and 200 nm-Y2 O3 :Eu3+ layers. The calculations were performed at 10 and 20 kV. The trajectories of 200 electrons were calculated. The scale bar indicates 50 nm.
The electrons started the scattering on the bottom surface of the Si3 N4 layer and then traveled into the 50 nm-Si3 N4 and 200 nmY2 O3 :Eu3+ layers, successively. At 10 kV, the electrons are scattered and spread widely in the Y2 O3 :Eu3+ layer, thus the CL spot might be generated inside of the Y2 O3 :Eu3+ layer. On the other hand, at 20 kV, electrons traveled straightly in the Y2 O3 :Eu3+ layer and the almost electrons arrived on the top surface of the Y2 O3 :Eu3+ layer. In this case, the CL spot could be smaller than that at 10 kV and be generated near the surface of the Y2 O3 :Eu3+ layer. Hence, the specimen can be illuminated by the smaller CL spot and the high spatial resolution can be achieved. The higher spatial resolution at higher acceleration voltage was in a good agreement with the observation result in Figs. 3 and 4. 6. Conclusions In this study, we presented the relationship between acceleration voltages and spatial resolutions in EXA microscope. We fabricated the 200 nm Y2 O3 :Eu3+ film for the EXA microscope and measured the CL intensity of the film depends on the acceleration voltage of the electron beam. The CL intensity indicates a peak at 7 kV because almost all of the electrons scatter in the Y2 O3 :Eu3+ film and the kinetic energies of the electrons are consumed for the CL excitations. At the voltages higher than 9 kV, the CL intensities decrease because the electrons penetrates through the Y2 O3 :Eu3+ film and the energy interaction for CL excitation decreases. We also observed 100 nm gold particles with the EXA microscope using the Y2 O3 :Eu3+ thin film at several acceleration voltages. We found that the FWHM of the 100 nm gold particle decreases with higher acceleration voltage because the electron scattering decreases. The FWHM of the 100 nm gold particles were 96.5 nm at 20 kV. We achieved the high spatial resolution than the diffraction limit at the voltages higher than 20 kV. According the simulation result, at more than 20 kV, almost all of the electrons traveled straightly to the top surface of the Y2 O3 :Eu3+ layer and the spot size at the Y2 O3 :Eu3+ surface was smaller. It could be confirmed successfully by the simulation that the spatial resolution of the EXA microscope depends on the acceleration voltage. The spatial resolution increase with higher acceleration voltages from simulation result. As another method, it is possible to obtain a higher resolution by constructing a structure on a fluorescent thin film [20].
This work has been supported by JST-CREST (Japan Science and Technology Agency, Core Research for Evolutionary Science and Technology). References [1] S.W. Hell, J. Wichmann, Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy, Opt. Lett. 19 (780) (1994) 780–782. [2] E. Betzig, G.H. Patterson, R. Sougrat, O.W. Lindwasser, S. Olenych, J.S. Bonifacino, M.W. Davidson, J. Lippincott-Schwartz, H.F. Hess, Imaging intracellular fluorescent proteins at nanometer resolution, Science 313 (5793) (2006) 1642–1645. [3] M.J. Rust, M. Bates, X. Zhuang, Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM), Nat. Methods 3 (10) (2006) 793–796. [4] R. Heintzmann, T.M. Jovin, C. Cremer, Saturated patterned excitation microscopy—a concept for optical resolution improvement, J. Opt. Soc. Am. A 19 (8) (2002) 1599–1609. [5] M.G.L. Gustafsson, Nonlinear structured-illumination microscopy: wide-field fluorescence imaging with theoretically unlimited resolution, J. Proc. Natl. Acad. Sci. 102 (37) (2005) 13081–13086. [6] G. Adamo, K.F. MacDonald, Y.H. Fu, C-M. Wang, D.P. Tsai, F.J. García de Abajo, N.I. Zheludev, Light well: a tunable free-electron light source on a chip, Phys. Rev. Lett. 103 (2009) 113901. [7] G. Adamo, K.F. MacDonald, Y.H. Fu, D.P. Tsai, F.J. García de Abajo, N.I. Zheludev, Tuneable electron-beam-driven nanoscale light source, J. Opt. 12 (2) (2010) 024012. [8] K. Fujita, M. Kobayashi, S. Kawano, M. Yamanaka, S. Kawata, High-resolution confocal microscopy by saturated excitation of fluorescence, Phys. Rev. Lett. 99 (2007) 228105. [9] W. Inami, K. Nakajima, A. Miyakawa, Y. Kawata, Electron beam excitation assisted optical microscope with ultra-high resolution, Opt. Express 18 (12) (2010) 12897–12902. [10] A. Sugita, M. Kamiya, C. Morita, A. Miyake, Y. Nawa, Y. Masuda, W. Inami, H. Kominami, Y. Nakanishi, Y. Kawata, Nanometric light spots of cathodoluminescence in Y2 O3 :Eu3+ phosphor thin films excited by focused electron beams as ultra-small illumination source for high-resolution optical microscope, Opt. Mater. Express 4 (1) (2013) 155–161. [11] S.H. Cho, S.H. Kwon, J.S. Yoo, C.W. Oh, J.D. Lee, K.J. Hong, S.J. Kwon, Cathodoluminescent characteristics of a spherical Y2 O3 :Eu phosphor screen for field emission display application, J. Electrochem. Soc. 147 (8) (20 0 0) 3143–3147. [12] S.H. Shin, J.H. Kang, D.Y. Jeon, D.S. Zang, Enhancement of cathodoluminescence intensities of Y2 O3 :Eu and Gd2 O3 :Eu phosphors by incorporation of Li ions, J. Lumin. 114 (3–4) (2005) 275–280. [13] G. Jia, M. Yang, Y. Song, H. You, H. Zhang, General and facile method to prepare uniform Y2O3:Eu hollow microspheres, Cryst. Growth Des. 9 (1) (2009) 301–307. [14] B.G. Yacobi, D.B. Holt, Cathodoluminescence Microscopy of Inorganic Solids, Plenum, 1990. [15] D.C. Joy, Monte Carlo Modeling For Electron Microscopy and Microanalysis, OXFORD, 1995. [16] M. Fukuta, W. Inami, A. Ono, Y. Kawata, Intensity distribution analysis of cathodoluminescent using the energy loss distribution of electrons, Ultramicroscopy 160 (2016) 225–229. [17] K.K. Hamamatsu Photonics, Photomultiplier Tubes Basic and Applications, third ed., 2007. [18] A. Rose, in: Advances in Electronics, 1, 1948, pp. 131–166. [19] N. de Jonge, N. Poirier-Demers, H. Demers, D. Peckys, D. Drouin, Nanometer-resolution electron microscopy through micrometers-thick water layers, Ultramicroscopy 110 (9) (2010) 1114–1119. [20] Y.-H. Lin, D.P. Tsai, Near-field scanning optical microscopy using a super-resolution cover glass slip, Opt. Express 20 (15) (2010) 16205–16211.