Sub-micrometer resolution proximity X-ray microscope with digital image registration N. I. Chkhalo, A. E. Pestov, N. N. Salashchenko, A. V. Sherbakov, E. V. Skorokhodov, and M. V. Svechnikov Citation: Review of Scientific Instruments 86, 063701 (2015); doi: 10.1063/1.4921849 View online: http://dx.doi.org/10.1063/1.4921849 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/86/6?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Comment on “A planar refractive x-ray lens made of nanocrystalline diamond” [J. Appl. Phys. 108, 123107 (2010)] J. Appl. Phys. 113, 206102 (2013); 10.1063/1.4807582 X-ray luminescence based spectrometer for investigation of scintillation properties Rev. Sci. Instrum. 83, 103112 (2012); 10.1063/1.4764772 Probing grain boundaries in ceramic scintillators using x-ray radioluminescence microscopy J. Appl. Phys. 111, 013520 (2012); 10.1063/1.3676222 High resolution x-ray microscope Appl. Phys. Lett. 90, 181111 (2007); 10.1063/1.2734895 Wolter-like high resolution x-ray imaging microscope for Rayleigh Taylor instabilities studies Rev. Sci. Instrum. 76, 063707 (2005); 10.1063/1.1902803
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
REVIEW OF SCIENTIFIC INSTRUMENTS 86, 063701 (2015)
Sub-micrometer resolution proximity X-ray microscope with digital image registration N. I. Chkhalo,1 A. E. Pestov,2 N. N. Salashchenko,1 A. V. Sherbakov,1,a) E. V. Skorokhodov,2 and M. V. Svechnikov1 1
Institute for Physics of Microstructures of the Russian Academy of Sciences, GSP-105, 603087 Nizhny Novgorod, Russia 2 Nizhny Novgorod N. I. Lobachevskii State University, Gagarina Ave. 23, 603950 Nizhny Novgorod, Russia
(Received 24 April 2014; accepted 17 May 2015; published online 4 June 2015) A compact laboratory proximity soft X-ray microscope providing submicrometer spatial resolution and digital image registration is described. The microscope consists of a laser-plasma soft X-ray radiation source, a Schwarzschild objective to illuminate the test sample, and a two-coordinate detector for image registration. Radiation, which passes through the sample under study, generates an absorption image on the front surface of the detector. Optical ceramic YAG:Ce was used to convert the X-rays into visible light. An image was transferred from the scintillator to a charge-coupled device camera with a Mitutoyo Plan Apo series lens. The detector’s design allows the use of lenses with numerical apertures of NA = 0.14, 0.28, and 0.55 without changing the dimensions and arrangement of the elements of the device. This design allows one to change the magnification, spatial resolution, and field of view of the X-ray microscope. A spatial resolution better than 0.7 µm and an energy conversion efficiency of the X-ray radiation with a wavelength of 13.5 nm into visible light collected by the detector of 7.2% were achieved with the largest aperture lens. C 2015 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4921849]
I. INTRODUCTION
Due to its short wavelength, weak scattering, relatively large penetration depth of the radiation into the substance, and resonant interaction with the atoms of light elements (O, N, and C), soft X-rays (wavelengths of 10–13 nm) are finding increasing application to X-ray microscopy, especially of biological samples.1 Usually, Fresnel zone plates (FZPs) as an imaging element, as well as a low-noise backside illuminated charge-coupled device (CCD) camera with a typical pixel size of about 13 µm as a detector, have been used in the majority of nano-diagnostics studies in recent years.2 The main advantages of this approach are its high spatial resolution and detection efficiency, as virtually all the incident X-ray quanta are recorded by the detector, and the relatively small size of the microscope. The disadvantages of the approach are as follows. Because of the relatively large pixel size of the CCD, this scheme requires working at high magnification (at least 1000×) in order to obtain a desired spatial resolution better than 30 nm. The low diffraction efficiency, large chromatic aberration, and small geometrical aperture of FZPs impose stringent requirements on the X-ray source. Small geometric apertures of 100 µm are typical for imaging FZPs and result in a small focal distance that decreases with increasing wavelength, which further complicates the use of FZPs at wavelengths above 3 nm, while the area of the “water” (wavelength 2.3–4.4 nm) and “carbon” (4.4–6 nm) transparency windows is of greatest interest for biological studies.4,5 Nevertheless, this approach achieves compact laboratory X-ray a)E-mail: [email protected]
microscopes6,7 using liquid-jet laser plasma sources of X-ray radiation. Normal incidence multilayer interference mirrors are an alternative to FZPs. In particular, it is shown in Ref. 8 that the efficiency of an X-ray microscope can be increased from a few to thousands of times, depending on the operating wavelength. Furthermore, in the long wavelength region of soft X-rays (wavelengths greater than 4 nm), there is practically no alternative to a multilayer mirror. Creating laboratory microscopes based on multilayer imaging optics has been discussed and overcomes some shortcomings of FZPs: low energy efficiency9 and a small field of view.10 However, as a result of the high requirements regarding the roughness and form precision of substrates for mirrors, the large size and complexity of the optical system, sensitivity to vibrations and temperature gradients, and high cost, only a small number of these microscopes have been produced. Therefore, it is significant to develop inexpensive and simple-to-use microscopes that have an easily tunable operating wavelength and are designed to work in the soft X-ray range. A comparison of the characteristics of various schemes of X-ray microscopes conducted in Ref. 8 showed that the proximity microscope, where the sample is installed in the immediate vicinity of the detector and a shadow image of the sample is formed on the detector’s input, satisfies those requirements best. Frequently, either an image plate or an Xray film is used as a 2D detector.11 This is a major disadvantage of microscopes of this type. Thus, methods of image recording require large doses of radiation. Long-term image treatment reduces work efficiency. In this paper, a proximity microscope with digital image registration is described. The main characteristics of the
0034-6748/2015/86(6)/063701/7/$30.00 86, 063701-1 © 2015 AIP Publishing LLC This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-2
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
II. MICROSCOPE DESIGN
FIG. 1. (a) Schematic diagram of the microscope and (b) photograph of the microscope.
microscope are presented. Potentially, the optical scheme of the detector allows one to change the magnification, spatial resolution, and field of view of the X-ray microscope during the measurement process, in analogy with classical optical microscopy, when lenses with various magnifications are mounted in the turret, and only a minor adjustment of focus is required. Schwarzschild lens (SL) is used as an X-ray collector. SL’s with multilayer coatings are widely used in soft X-ray researches (X-ray microscopy and X-ray nanolithography) as imaging and illuminating optics for it low spherical aberration, wide field of view, and large numerical aperture (NA).12–14
An optical diagram and a photograph of the microscope are shown in Fig. 1. The main elements of the microscope are a laser-plasma source (LPS) of soft X-ray radiation, Schwarzschild lens, a table for the test samples (ST), and a highresolution detector (HRD). The operation of the unit is as follows. An image of the laser-plasma source is transferred to the sample mounted on the table (ST) with a tenfold magnification using a SL. The HRD is installed just behind the test sample. An absorption shadow image of the sample is formed at the entrance of the detector and detected with a CCD camera after conversion into visible light by a scintillator. Aside from the detector, the key parameters determining the spatial resolution of the proximity microscope are the distance δx between the sample and the detector and the properties of the illuminating beam. The divergence of the incident radiation and the diffraction of light on the details of the object cause a loss of resolution. Figures 2 and 3 shows the calculated dependences of the intensity distribution of X-rays passing through the test strips with widths of 0.6 and 1 µm installed at different distances from the detector. The calculation was performed using the ZEMAX code for a radiation source of size 100 µm and a wavelength of 13.5 nm. As can be seen from the figures, according to the Rayleigh criterion, the limiting distances are 50 and 100 µm for strips of 0.6 and 1 µm width, respectively, which can be implemented in practice. A. Laser plasma source of radiation
A LPS with a solid target was used to generate X-rays. An Nd:YAG laser with the following parameters was used: a pulse energy of 0.3 J, pulse duration of 7 ns, and pulse repetition rate of 13 Hz. The target material was turned after each shot of the laser so that the subsequent shot hit the clean space to improve the stability of the emission characteristics and reduce the mass of the debris emitted from the target, similarly to Ref. 15.
FIG. 2. Calculation of image of 2 µm period strips and their cross sections at distances from the front surface of the scintillator of (a) 30 µm, (b) 60 µm, and (c) 100 µm. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-3
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
FIG. 3. Calculation of image of 1.2 µm period strips and their cross sections at distances from the front surface of the scintillator of (a) 0 µm, (b) 30 µm, and (c) 60 µm.
After making one revolution using a linear translator, the target moved forward and the process was repeated. Protection of the entrance window from the target’s erosion products was carried out using a quartz plate mounted at a distance of about 3 cm from the target. At this distance, the laser power density on the plate is about 108 W/cm2, which is sufficient for the film that occurs after the previous laser pulse to be “atomized” by the following one. A “fan” of flat plates under voltage was installed along the X-ray beam, which allowed the ions from the material to be captured in order to protect the mirrors of the SL. The target material can be replaced depending on the operating wavelength so that the highly charged ions have spectral lines in this area to improve device efficiency. Tin was used for λ = 13.5 nm. The measured spectral dependence of the number of photons in 5% of the bandwidth emitted in 6.1 × 10−4 sr of the radiation source is shown in Fig. 4. As seen from the figure, this source works effectively in the field of 13–17 nm. The absolute number of photons in the probe beam was measured with a calibrated metrology tool16 and was 1012
photons/s. Taking into account the solid angle and reflectivity of the mirrors and bandwidth ∆λSL = 0.34 nm in the vicinity of 13.5 nm of the SL, this corresponds to a conversion efficiency of laser energy to EUV (extreme ultraviolet), radiated into the half space in a 2% (0.27 nm) spectral bandpass, at the level of 2%. This value is comparable to the record levels (5% in the 0.27 nm band14). B. Schwarzschild lens
A classic SL coated with multilayer mirrors with a total energy efficiency of 36% was used for illumination of the test sample. The lens has a magnification of 10×. The reception angular aperture of the SL is ±6◦, and correspondingly, a quasiparallel beam with a divergence of ±0.6◦ is at the output. A filter that blocks long-wave radiation that passes through the mirror lens in a regular manner was installed after the second mirror. The filter’s transmission at a wavelength of 13.5 nm is 70%. The scheme and basic dimensions of the lens are shown in Fig. 5 and Table I. C. Detector design
The layout of the detector, which is similar to that described in Ref. 17, is shown in Fig. 6. The front surface of the scintillator (YAG:Ce optical ceramics) is located in the object plane of the optical lens (Mitutoyo Plan Apo series). The Xray image generated by the X-ray lens in the front layer of the scintillator is converted into an optical image. With the
FIG. 4. Spectral dependence of the number of photons in 5% of the bandwidth emitted in 6.1 × 10−4 sr for the LPS used. FIG. 5. Lens construction and linear dimensions, mm. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-4
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
TABLE I. SL parameters.
The radius of curvature, mm Diameter, mm Coating The reflection coefficient, % Bandwidth, nm
The concave mirror
The convex mirror
428.5 124 Mo/Si 60 0.48
133.05 22 Mo/Si 60 0.48
optical system consisting of the lens and an image-forming lens, this image is transferred with magnification to a CCD camera. The distances between the scintillator, lens, imageforming lens, and detector were chosen according to the manufacturer’s recommendations and based on the size of the CCD (2/3 in. camera BRM 1400HM-U made by ES Experts, St. Petersburg). To reduce the length of the detector, the imageforming lens had a focal length F = 75 mm, which differs from the manufacturer’s recommended focal length of 200 mm. An important advantage of the proposed scheme of the detector is that when replacing one lens of the Mitutoyo Plan Apo series with another, the relative locations of the detector elements are practically unchanged, and then only a minor adjustment of CCD to the image plane is required by manual shifter, in analogy to classical optical microscopy. Thus, by replacing the optical lens, it is possible to change the magnification and, accordingly, the spatial resolution and field of view without reconfiguring the X-ray part of the microscope. The energy efficiency of the detector is defined as the ratio of the radiation energy that is recorded with the CCD camera to the energy of the incident X-ray radiation and depends on the geometrical (GE) and conversion (SCE) efficiencies of the lens and scintillator, respectively. The GE of the lens represents the fraction captured by an optical lens of the total number of light quanta emitted by the scintillator. The SCE of the scintillator is defined as the ratio of the radiated photon energy to the energy expended in its birth. GE is determined by the NA of the lens. Assuming that the scintillator emits light isotropically and that its surface is coated with a visible light reflector of 100%, it is easy to show that GE would be 1.0%, 4.3%, and 21.7% for NA = 0.14, 0.28, and 0.55, respectively. Without reflective coating, it is one-
half of that. It is important to note that, ceteris paribus, the illumination pixel CCD matrix is practically independent of the NA of the lens. This is due to the fact that the quadratic GE drops with a decreasing NA compensated by the same increase in the emitting area of the object due to the reduction of the magnification. The SCE was measured at a wavelength of 13.5 nm. As a reference, we used the semiconductor diode SPD-100UV with a Zr/Si spectral-purity filter produced by the A. F. Joffe Physical-Technical Institute, St. Petersburg, and calibrated at PTB, BESSY-2, Berlin.18 The experiment was performed as follows. On the body of the diode at a distance of 12 mm, the scintillator under study (without a light-reflecting coating) was placed, spanning about one-half of the aperture of the diode. Then, the diode unit (together with the scintillator) moved across the probe X-ray beam and the dependency of the measured detector current on the coordinate was recorded. When the detector was taken out of the beam, the current was equal to zero. When the beam hit the diode directly, we observed the maximal signal, corresponding to direct registration of the probe beam by the SPD-100UV. When the beam hit only the scintillator, we registered only fluorescent light and the signal had fallen sharply. From the known geometry of the experiment and the number of photons in the probe beam, and taking into account the diode calibration data for the detector at a wavelength of 13.5 nm,16 we were able to estimate the SCE as a ratio of the power emitted by the scintillator light to the power of the incident X-ray beam, ISC K350 · Ω4πD WSC = , SCE = WB I B K13.5
(1)
where W B is the incident X-ray beam power, WSC is the power emitted by the scintillator visible light, ISC is the diode current when detecting scintillated light, K350 is the detector sensitivity at the wavelength of scintillation, I B is the diode current when detecting the incident X-ray beam, K13.5 is the detector sensitivity at a wavelength of 13.5 nm,19 and Ω D is the solid angle of the detector from the point of scintillation. SCE amounted to 0.33 ± 0.03 (quantum of visible light 3.5 eV/10.5 eV X-rays). Thus, the energy efficiency of the detector is 0.3%, 1.4%, and 7.2% for NA = 0.14, 0.28, and 0.55, respectively. Subsequently, these values were used to calculate and optimize the efficiency of the X-ray circuit of the optical microscope.
III. RESULT OF TESTING THE SPATIAL RESOLUTION AND FIELD OF VIEW OF THE MICROSCOPE
The main optical characteristics of the microscope are magnification, field of view, and spatial resolution. The first two characteristics were studied using as a test sample a system of metal strips of 10 µm width with a duty cycle of 1/2. The test object was placed in the object plane of the lens and illuminated by a red light-emitting diode. Knowing the width of the strips and their number as well as the size of the CCD, it is not difficult to determine the magnification and field of view of the detector. Corresponding values for each lens are shown in Table II.
FIG. 6. Detector design. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-5
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
TABLE II. Detector’s basic characteristics. Lens, NA Energetic efficiency, % Magnification, × Spatial resolution, µm Field of view, µm
0.14 0.3 1.96 10.6 4480 × 3138
0.28 1.4 4.23 1.01 2080 × 1456
0.55 7.2 20 0.58 440 × 308
The magnification and field of view of the lenses differ from those given in the manufacturer’s specifications due to the use of a short-focus imaging lens (F = 75 mm instead of the recommended F = 190 mm). Despite the loss in magnification, this approach can significantly reduce the overall size of the detector to about 125 mm in length and 25 mm in diameter. Another test object was used to study the spatial resolution and the point spread function. The test object was a source of a spherical wave based on a single mode optical fiber with a narrowed exit aperture, described in Ref. 20. An image of the emitting source was taken by the detector in visible light. The source size (250 nm) was smaller than the wavelength (632 nm), so received picture demonstrates the point spread function of the detecting system. Figures 7(a) and 7(b) obtained image of the test object and its cross section’s respectively. To estimate resolution on soft X-rays a system of strips with widths of 0.7, 1.0, 1.5, and 2 µm deposited directly on the scintillator was used as a test object. Figures 8(a) and 8(b) show the images of the test object taken by the lens with NA = 0.55 and illuminated by a red light-emitting diode (LED) and with a spectral-purity filter (illumination at 13.5 nm), respectively. Figure 8(c) shows an image of the test object taken by scanning electron microscope. The broken lines and the roughness of the edges, which are observed in electron microscopic images too, are linked to the quality of the lithographic pattern. Cross sections of the images obtained with visible light and a wavelength of 13.5 nm are shown in Fig. 9. As can be seen from the figure, according to the Rayleigh criterion, the detector with the lens with NA = 0.55 has a resolution better
FIG. 8. Test-object images: (a) received through the lens with NA = 0.55 in visible light; (b) received through the lens with NA = 0.55 at a wavelength of 13.5 nm; (c) produced by an electron microscope.
FIG. 7. (a) Image of the point source and (b) its cross section’s. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-6
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
quality at the wavelength of 13.5 nm was mainly due to the noise of the uncooled CCD camera operating in a vacuum. Using cooling for the camera will presumably improve the quality of the images and the spatial resolution. Another opportunity to improve the optical resolution is to increase the numerical aperture of the detector’s optical system through the use of a more powerful commercially available lens or the formation of a scintillator in the form of a truncated sphere, as was done in Ref. 21. These methods can improve the spatial resolution of such a device down to 0.4 µm.
IV. RESULTS AND DISCUSSION
The proposed and implemented soft X-ray proximity microscope with digital image registration has a spatial resolution at the level of 0.6–0.7 µm, which is of interest for many biological applications. The detector energy efficiency (up to 7.2%) is about 14 times inferior to that of a backside illuminated CCD matrix, but it is sufficient for recording Xray images obtained with the use of a low-power laser with a pulse energy below 300 mJ, pulse duration of 7 ns, and an angular aperture of the monochromator of 0.03 sr. At the same time, the detector provides a spatial resolution about 20 times better than that provided by a low-noise backside illuminated CCD matrix. The design of the microscope provides a unique opportunity to vary parameters such as the resolution, field of view, and magnification directly in the vacuum chamber through the use of a set of optical lenses, which is important in the study of samples with complicated structures. The ultimate resolution is achieved when the distance between the detector and the test sample is about 100 µm, which is sufficient for most experiments. This distance can be easily increased without changing the settings of the device, but only by increasing the magnification of the SL. This increase does not lead to a drop in the probe beam photon density on the sample, since it is accompanied by an increase in the angular aperture of the primary mirror of the SL. This device can also be used to study samples with better resolution up to tens of nanometers. Film, image plates, and, for the nanometer scale, photoresists can be used instead of the
FIG. 9. Cross section of the strip width of (a) 0.7, (b) 1.0, (c) 1.5, and (d) 2 µm with the lens NA = 0.55. Images obtained at a wavelength of 13.5 nm and in visible light.
than 0.7 µm in visible light. The difference in amplitude of the peaks, as mentioned above, is explained by the rough edges and defects in the test object. As can be seen, in the soft Xray range, the resolution is about 0.7 µm. The drop in image
FIG. 10. Photo of porous Mylar film obtained using the photoresist. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-7
Chkhalo et al.
developed digital X-ray detector. An image of porous Mylar film was made to demonstrate the compatibility with photoresists (Fig. 10). Using this detector, the described microscope demonstrates a spatial resolution better than 100 nm. The scheme allows the microscope to operate at multiple wavelengths. For this purpose, multilayer coatings with the desired resonance wavelengths are deposited on the conjugate mirror segments, and the working segment is selected using the moveable diaphragm. Thus, this paper proposed and studied a compact laboratory X-ray proximity microscope that provides digital recording images with a resolution better than 0.6 µm, making it possible to work in the mode of a traditional proximity microscope with registration on photosensitive materials at nanoscale resolution. ACKNOWLEDGMENTS
This work was supported by RFBR Grant Nos. 12-0231678, 13-02-00377, and 13-02-97045 and the Ministry of Education and Science of Russia. 1J.
Kirz and M. Howells, Q. Rev. Biophys. 28, 130 (1995). Chao, J. Kim, S. Rekawa, P. Fischer, and E. H. Anderson, Opt. Express 17(20), 17669–17677 (2009). 3W. Chao, B. Harteneck, J. Liddle et al., Nature 435(7046), 1210 (2005). 4J.-F. Adam, J.-P. Moy, and J. Susini, Rev. Sci. Instrum. 76, 091301 (2005). 5H. Legall, G. Blobel, H. Stiel, W. Sandner, C. Seim, P. Takman, D. H. Martz, M. Selin, U. Vogt, H. M. Hertz, D. Esser, H. Sipma, J. Luttmann, M. Höfer, H. D. Hoffmann, S. Yulin, T. Feigl, S. Rehbein, P. Guttmann, G. Schneider, 2W.
Rev. Sci. Instrum. 86, 063701 (2015) U. Wiesemann, M. Wirtz, and W. Diete, Opt. Express 20(16), 18362–18369 (2012). 6K. W. Kim, Y. Kwon, K.-Y. Nam, J.-H. Lim, K.-G. Kim, K. S. Chon, B. H. Kim, D. E. Kim, J. G. Kim, B. N. Ahn, H. J. Shin, S. Rah, K.-H. Kim, J. S. Chae, D. G. Gweon, D. W. Kang, S. H. Kang, J. Y. Min, K.-S. Choi, S. E. Yoon, E.-A. Kim, Y. Namba, and K.-H. Yoon, Phys. Med. Biol. 51, N99–N107 (2006). 7P. A. C. Takman, H. Stollberg, G. A. Johansson, A. Holmberg, M. Lindblom, and H. M. Hertz, J. Microsc. 226(2), 175–181 (2007). 8M. Barysheva, A. Pestov, N. Salashchenko, M. Toropov, and N. Chkhalo, Phys.-Usp. 55(7), 681 (2012). 9T. Ejima, F. Ishida, H. Murata, M. Toyoda, T. Harada, T. Tsuru, T. Hatano, M. Yanagihara, M. Yamamoto, and H. Mizutani, Opt. Express 18(7), 7203 (2010). 10L. Juschkin and R. Freiberger, Proc. SPIE 7360, 736001 (2009). 11I. A. Artioukov and A. V. Vinogradov, Opt. Lett. 20(24), 2451 (1995). 12F. Barkusky et al., Rev. Sci. Instrum. 76, 105102 (2005). 13C. Montcalm et al., Proc. SPIE 3331, 42 (1998). 14T. Haga et al., Jpn. J. Appl. Phys., Part 1 42, 3771 (2003). 15S. Yu. Zuev, A. E. Pestov, N. N. Salashchenko, M. N. Toropov, N. I. Chkhalo, and A. V. Shcherbakov, Bull. Russ. Acad. Sci.: Phys. 77(1), 6–9 (2013). 16N. I. Chkhalo, S. V. Golubev, D. Mansfeld, N. N. Salashchenko, L. A. Sjmaenok, and A. V. Vodopyanov, J. Micro/Nanolithogr., MEMS, MOEMS 11, 021123 (2012). 17S. S. Harilal, T. Sizyuk, V. Sizyuk, and A. Hassanein, Appl. Phys. Lett. 96, 111503 (2010). 18T. Ejima, Y. Neichi, F. Ishida, and M. Yanagihara, J. Phys.: Conf. Ser. 463, 012055 (2013). 19P. Aruev, M. Barysheva, B. Ber, N. Zabrodskaya, V. Zabrodskii, A. Lopatin, A. Pestov, M. Petrenko, V. Polkovnikov, N. Salashchenko, V. Sukhanov, and N. Chkhalo, Quantum Electron. 42(10), 943 (2012). 20N. I. Chkhalo, A. Yu. Klimov, V. V. Rogov, N. N. Salashchenko, and M. N. Toropov, Rev. Sci. Instrum. 79, 033107 (2008). 21B. Lagarde, M. Bordessoule, G. Cauchon, D. Dallé, K. Desjardins, S. Hustache, C. Miron, C. Nicolas, and F. Polack, J. Phys.: Conf. Ser. 425, 152002 (2013).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
REVIEW OF SCIENTIFIC INSTRUMENTS 86, 063701 (2015)
Sub-micrometer resolution proximity X-ray microscope with digital image registration N. I. Chkhalo,1 A. E. Pestov,2 N. N. Salashchenko,1 A. V. Sherbakov,1,a) E. V. Skorokhodov,2 and M. V. Svechnikov1 1
Institute for Physics of Microstructures of the Russian Academy of Sciences, GSP-105, 603087 Nizhny Novgorod, Russia 2 Nizhny Novgorod N. I. Lobachevskii State University, Gagarina Ave. 23, 603950 Nizhny Novgorod, Russia
(Received 24 April 2014; accepted 17 May 2015; published online 4 June 2015) A compact laboratory proximity soft X-ray microscope providing submicrometer spatial resolution and digital image registration is described. The microscope consists of a laser-plasma soft X-ray radiation source, a Schwarzschild objective to illuminate the test sample, and a two-coordinate detector for image registration. Radiation, which passes through the sample under study, generates an absorption image on the front surface of the detector. Optical ceramic YAG:Ce was used to convert the X-rays into visible light. An image was transferred from the scintillator to a charge-coupled device camera with a Mitutoyo Plan Apo series lens. The detector’s design allows the use of lenses with numerical apertures of NA = 0.14, 0.28, and 0.55 without changing the dimensions and arrangement of the elements of the device. This design allows one to change the magnification, spatial resolution, and field of view of the X-ray microscope. A spatial resolution better than 0.7 µm and an energy conversion efficiency of the X-ray radiation with a wavelength of 13.5 nm into visible light collected by the detector of 7.2% were achieved with the largest aperture lens. C 2015 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4921849]
I. INTRODUCTION
Due to its short wavelength, weak scattering, relatively large penetration depth of the radiation into the substance, and resonant interaction with the atoms of light elements (O, N, and C), soft X-rays (wavelengths of 10–13 nm) are finding increasing application to X-ray microscopy, especially of biological samples.1 Usually, Fresnel zone plates (FZPs) as an imaging element, as well as a low-noise backside illuminated charge-coupled device (CCD) camera with a typical pixel size of about 13 µm as a detector, have been used in the majority of nano-diagnostics studies in recent years.2 The main advantages of this approach are its high spatial resolution and detection efficiency, as virtually all the incident X-ray quanta are recorded by the detector, and the relatively small size of the microscope. The disadvantages of the approach are as follows. Because of the relatively large pixel size of the CCD, this scheme requires working at high magnification (at least 1000×) in order to obtain a desired spatial resolution better than 30 nm. The low diffraction efficiency, large chromatic aberration, and small geometrical aperture of FZPs impose stringent requirements on the X-ray source. Small geometric apertures of 100 µm are typical for imaging FZPs and result in a small focal distance that decreases with increasing wavelength, which further complicates the use of FZPs at wavelengths above 3 nm, while the area of the “water” (wavelength 2.3–4.4 nm) and “carbon” (4.4–6 nm) transparency windows is of greatest interest for biological studies.4,5 Nevertheless, this approach achieves compact laboratory X-ray a)E-mail: [email protected]
microscopes6,7 using liquid-jet laser plasma sources of X-ray radiation. Normal incidence multilayer interference mirrors are an alternative to FZPs. In particular, it is shown in Ref. 8 that the efficiency of an X-ray microscope can be increased from a few to thousands of times, depending on the operating wavelength. Furthermore, in the long wavelength region of soft X-rays (wavelengths greater than 4 nm), there is practically no alternative to a multilayer mirror. Creating laboratory microscopes based on multilayer imaging optics has been discussed and overcomes some shortcomings of FZPs: low energy efficiency9 and a small field of view.10 However, as a result of the high requirements regarding the roughness and form precision of substrates for mirrors, the large size and complexity of the optical system, sensitivity to vibrations and temperature gradients, and high cost, only a small number of these microscopes have been produced. Therefore, it is significant to develop inexpensive and simple-to-use microscopes that have an easily tunable operating wavelength and are designed to work in the soft X-ray range. A comparison of the characteristics of various schemes of X-ray microscopes conducted in Ref. 8 showed that the proximity microscope, where the sample is installed in the immediate vicinity of the detector and a shadow image of the sample is formed on the detector’s input, satisfies those requirements best. Frequently, either an image plate or an Xray film is used as a 2D detector.11 This is a major disadvantage of microscopes of this type. Thus, methods of image recording require large doses of radiation. Long-term image treatment reduces work efficiency. In this paper, a proximity microscope with digital image registration is described. The main characteristics of the
0034-6748/2015/86(6)/063701/7/$30.00 86, 063701-1 © 2015 AIP Publishing LLC This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-2
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
II. MICROSCOPE DESIGN
FIG. 1. (a) Schematic diagram of the microscope and (b) photograph of the microscope.
microscope are presented. Potentially, the optical scheme of the detector allows one to change the magnification, spatial resolution, and field of view of the X-ray microscope during the measurement process, in analogy with classical optical microscopy, when lenses with various magnifications are mounted in the turret, and only a minor adjustment of focus is required. Schwarzschild lens (SL) is used as an X-ray collector. SL’s with multilayer coatings are widely used in soft X-ray researches (X-ray microscopy and X-ray nanolithography) as imaging and illuminating optics for it low spherical aberration, wide field of view, and large numerical aperture (NA).12–14
An optical diagram and a photograph of the microscope are shown in Fig. 1. The main elements of the microscope are a laser-plasma source (LPS) of soft X-ray radiation, Schwarzschild lens, a table for the test samples (ST), and a highresolution detector (HRD). The operation of the unit is as follows. An image of the laser-plasma source is transferred to the sample mounted on the table (ST) with a tenfold magnification using a SL. The HRD is installed just behind the test sample. An absorption shadow image of the sample is formed at the entrance of the detector and detected with a CCD camera after conversion into visible light by a scintillator. Aside from the detector, the key parameters determining the spatial resolution of the proximity microscope are the distance δx between the sample and the detector and the properties of the illuminating beam. The divergence of the incident radiation and the diffraction of light on the details of the object cause a loss of resolution. Figures 2 and 3 shows the calculated dependences of the intensity distribution of X-rays passing through the test strips with widths of 0.6 and 1 µm installed at different distances from the detector. The calculation was performed using the ZEMAX code for a radiation source of size 100 µm and a wavelength of 13.5 nm. As can be seen from the figures, according to the Rayleigh criterion, the limiting distances are 50 and 100 µm for strips of 0.6 and 1 µm width, respectively, which can be implemented in practice. A. Laser plasma source of radiation
A LPS with a solid target was used to generate X-rays. An Nd:YAG laser with the following parameters was used: a pulse energy of 0.3 J, pulse duration of 7 ns, and pulse repetition rate of 13 Hz. The target material was turned after each shot of the laser so that the subsequent shot hit the clean space to improve the stability of the emission characteristics and reduce the mass of the debris emitted from the target, similarly to Ref. 15.
FIG. 2. Calculation of image of 2 µm period strips and their cross sections at distances from the front surface of the scintillator of (a) 30 µm, (b) 60 µm, and (c) 100 µm. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-3
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
FIG. 3. Calculation of image of 1.2 µm period strips and their cross sections at distances from the front surface of the scintillator of (a) 0 µm, (b) 30 µm, and (c) 60 µm.
After making one revolution using a linear translator, the target moved forward and the process was repeated. Protection of the entrance window from the target’s erosion products was carried out using a quartz plate mounted at a distance of about 3 cm from the target. At this distance, the laser power density on the plate is about 108 W/cm2, which is sufficient for the film that occurs after the previous laser pulse to be “atomized” by the following one. A “fan” of flat plates under voltage was installed along the X-ray beam, which allowed the ions from the material to be captured in order to protect the mirrors of the SL. The target material can be replaced depending on the operating wavelength so that the highly charged ions have spectral lines in this area to improve device efficiency. Tin was used for λ = 13.5 nm. The measured spectral dependence of the number of photons in 5% of the bandwidth emitted in 6.1 × 10−4 sr of the radiation source is shown in Fig. 4. As seen from the figure, this source works effectively in the field of 13–17 nm. The absolute number of photons in the probe beam was measured with a calibrated metrology tool16 and was 1012
photons/s. Taking into account the solid angle and reflectivity of the mirrors and bandwidth ∆λSL = 0.34 nm in the vicinity of 13.5 nm of the SL, this corresponds to a conversion efficiency of laser energy to EUV (extreme ultraviolet), radiated into the half space in a 2% (0.27 nm) spectral bandpass, at the level of 2%. This value is comparable to the record levels (5% in the 0.27 nm band14). B. Schwarzschild lens
A classic SL coated with multilayer mirrors with a total energy efficiency of 36% was used for illumination of the test sample. The lens has a magnification of 10×. The reception angular aperture of the SL is ±6◦, and correspondingly, a quasiparallel beam with a divergence of ±0.6◦ is at the output. A filter that blocks long-wave radiation that passes through the mirror lens in a regular manner was installed after the second mirror. The filter’s transmission at a wavelength of 13.5 nm is 70%. The scheme and basic dimensions of the lens are shown in Fig. 5 and Table I. C. Detector design
The layout of the detector, which is similar to that described in Ref. 17, is shown in Fig. 6. The front surface of the scintillator (YAG:Ce optical ceramics) is located in the object plane of the optical lens (Mitutoyo Plan Apo series). The Xray image generated by the X-ray lens in the front layer of the scintillator is converted into an optical image. With the
FIG. 4. Spectral dependence of the number of photons in 5% of the bandwidth emitted in 6.1 × 10−4 sr for the LPS used. FIG. 5. Lens construction and linear dimensions, mm. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-4
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
TABLE I. SL parameters.
The radius of curvature, mm Diameter, mm Coating The reflection coefficient, % Bandwidth, nm
The concave mirror
The convex mirror
428.5 124 Mo/Si 60 0.48
133.05 22 Mo/Si 60 0.48
optical system consisting of the lens and an image-forming lens, this image is transferred with magnification to a CCD camera. The distances between the scintillator, lens, imageforming lens, and detector were chosen according to the manufacturer’s recommendations and based on the size of the CCD (2/3 in. camera BRM 1400HM-U made by ES Experts, St. Petersburg). To reduce the length of the detector, the imageforming lens had a focal length F = 75 mm, which differs from the manufacturer’s recommended focal length of 200 mm. An important advantage of the proposed scheme of the detector is that when replacing one lens of the Mitutoyo Plan Apo series with another, the relative locations of the detector elements are practically unchanged, and then only a minor adjustment of CCD to the image plane is required by manual shifter, in analogy to classical optical microscopy. Thus, by replacing the optical lens, it is possible to change the magnification and, accordingly, the spatial resolution and field of view without reconfiguring the X-ray part of the microscope. The energy efficiency of the detector is defined as the ratio of the radiation energy that is recorded with the CCD camera to the energy of the incident X-ray radiation and depends on the geometrical (GE) and conversion (SCE) efficiencies of the lens and scintillator, respectively. The GE of the lens represents the fraction captured by an optical lens of the total number of light quanta emitted by the scintillator. The SCE of the scintillator is defined as the ratio of the radiated photon energy to the energy expended in its birth. GE is determined by the NA of the lens. Assuming that the scintillator emits light isotropically and that its surface is coated with a visible light reflector of 100%, it is easy to show that GE would be 1.0%, 4.3%, and 21.7% for NA = 0.14, 0.28, and 0.55, respectively. Without reflective coating, it is one-
half of that. It is important to note that, ceteris paribus, the illumination pixel CCD matrix is practically independent of the NA of the lens. This is due to the fact that the quadratic GE drops with a decreasing NA compensated by the same increase in the emitting area of the object due to the reduction of the magnification. The SCE was measured at a wavelength of 13.5 nm. As a reference, we used the semiconductor diode SPD-100UV with a Zr/Si spectral-purity filter produced by the A. F. Joffe Physical-Technical Institute, St. Petersburg, and calibrated at PTB, BESSY-2, Berlin.18 The experiment was performed as follows. On the body of the diode at a distance of 12 mm, the scintillator under study (without a light-reflecting coating) was placed, spanning about one-half of the aperture of the diode. Then, the diode unit (together with the scintillator) moved across the probe X-ray beam and the dependency of the measured detector current on the coordinate was recorded. When the detector was taken out of the beam, the current was equal to zero. When the beam hit the diode directly, we observed the maximal signal, corresponding to direct registration of the probe beam by the SPD-100UV. When the beam hit only the scintillator, we registered only fluorescent light and the signal had fallen sharply. From the known geometry of the experiment and the number of photons in the probe beam, and taking into account the diode calibration data for the detector at a wavelength of 13.5 nm,16 we were able to estimate the SCE as a ratio of the power emitted by the scintillator light to the power of the incident X-ray beam, ISC K350 · Ω4πD WSC = , SCE = WB I B K13.5
(1)
where W B is the incident X-ray beam power, WSC is the power emitted by the scintillator visible light, ISC is the diode current when detecting scintillated light, K350 is the detector sensitivity at the wavelength of scintillation, I B is the diode current when detecting the incident X-ray beam, K13.5 is the detector sensitivity at a wavelength of 13.5 nm,19 and Ω D is the solid angle of the detector from the point of scintillation. SCE amounted to 0.33 ± 0.03 (quantum of visible light 3.5 eV/10.5 eV X-rays). Thus, the energy efficiency of the detector is 0.3%, 1.4%, and 7.2% for NA = 0.14, 0.28, and 0.55, respectively. Subsequently, these values were used to calculate and optimize the efficiency of the X-ray circuit of the optical microscope.
III. RESULT OF TESTING THE SPATIAL RESOLUTION AND FIELD OF VIEW OF THE MICROSCOPE
The main optical characteristics of the microscope are magnification, field of view, and spatial resolution. The first two characteristics were studied using as a test sample a system of metal strips of 10 µm width with a duty cycle of 1/2. The test object was placed in the object plane of the lens and illuminated by a red light-emitting diode. Knowing the width of the strips and their number as well as the size of the CCD, it is not difficult to determine the magnification and field of view of the detector. Corresponding values for each lens are shown in Table II.
FIG. 6. Detector design. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-5
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
TABLE II. Detector’s basic characteristics. Lens, NA Energetic efficiency, % Magnification, × Spatial resolution, µm Field of view, µm
0.14 0.3 1.96 10.6 4480 × 3138
0.28 1.4 4.23 1.01 2080 × 1456
0.55 7.2 20 0.58 440 × 308
The magnification and field of view of the lenses differ from those given in the manufacturer’s specifications due to the use of a short-focus imaging lens (F = 75 mm instead of the recommended F = 190 mm). Despite the loss in magnification, this approach can significantly reduce the overall size of the detector to about 125 mm in length and 25 mm in diameter. Another test object was used to study the spatial resolution and the point spread function. The test object was a source of a spherical wave based on a single mode optical fiber with a narrowed exit aperture, described in Ref. 20. An image of the emitting source was taken by the detector in visible light. The source size (250 nm) was smaller than the wavelength (632 nm), so received picture demonstrates the point spread function of the detecting system. Figures 7(a) and 7(b) obtained image of the test object and its cross section’s respectively. To estimate resolution on soft X-rays a system of strips with widths of 0.7, 1.0, 1.5, and 2 µm deposited directly on the scintillator was used as a test object. Figures 8(a) and 8(b) show the images of the test object taken by the lens with NA = 0.55 and illuminated by a red light-emitting diode (LED) and with a spectral-purity filter (illumination at 13.5 nm), respectively. Figure 8(c) shows an image of the test object taken by scanning electron microscope. The broken lines and the roughness of the edges, which are observed in electron microscopic images too, are linked to the quality of the lithographic pattern. Cross sections of the images obtained with visible light and a wavelength of 13.5 nm are shown in Fig. 9. As can be seen from the figure, according to the Rayleigh criterion, the detector with the lens with NA = 0.55 has a resolution better
FIG. 8. Test-object images: (a) received through the lens with NA = 0.55 in visible light; (b) received through the lens with NA = 0.55 at a wavelength of 13.5 nm; (c) produced by an electron microscope.
FIG. 7. (a) Image of the point source and (b) its cross section’s. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-6
Chkhalo et al.
Rev. Sci. Instrum. 86, 063701 (2015)
quality at the wavelength of 13.5 nm was mainly due to the noise of the uncooled CCD camera operating in a vacuum. Using cooling for the camera will presumably improve the quality of the images and the spatial resolution. Another opportunity to improve the optical resolution is to increase the numerical aperture of the detector’s optical system through the use of a more powerful commercially available lens or the formation of a scintillator in the form of a truncated sphere, as was done in Ref. 21. These methods can improve the spatial resolution of such a device down to 0.4 µm.
IV. RESULTS AND DISCUSSION
The proposed and implemented soft X-ray proximity microscope with digital image registration has a spatial resolution at the level of 0.6–0.7 µm, which is of interest for many biological applications. The detector energy efficiency (up to 7.2%) is about 14 times inferior to that of a backside illuminated CCD matrix, but it is sufficient for recording Xray images obtained with the use of a low-power laser with a pulse energy below 300 mJ, pulse duration of 7 ns, and an angular aperture of the monochromator of 0.03 sr. At the same time, the detector provides a spatial resolution about 20 times better than that provided by a low-noise backside illuminated CCD matrix. The design of the microscope provides a unique opportunity to vary parameters such as the resolution, field of view, and magnification directly in the vacuum chamber through the use of a set of optical lenses, which is important in the study of samples with complicated structures. The ultimate resolution is achieved when the distance between the detector and the test sample is about 100 µm, which is sufficient for most experiments. This distance can be easily increased without changing the settings of the device, but only by increasing the magnification of the SL. This increase does not lead to a drop in the probe beam photon density on the sample, since it is accompanied by an increase in the angular aperture of the primary mirror of the SL. This device can also be used to study samples with better resolution up to tens of nanometers. Film, image plates, and, for the nanometer scale, photoresists can be used instead of the
FIG. 9. Cross section of the strip width of (a) 0.7, (b) 1.0, (c) 1.5, and (d) 2 µm with the lens NA = 0.55. Images obtained at a wavelength of 13.5 nm and in visible light.
than 0.7 µm in visible light. The difference in amplitude of the peaks, as mentioned above, is explained by the rough edges and defects in the test object. As can be seen, in the soft Xray range, the resolution is about 0.7 µm. The drop in image
FIG. 10. Photo of porous Mylar film obtained using the photoresist. This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
063701-7
Chkhalo et al.
developed digital X-ray detector. An image of porous Mylar film was made to demonstrate the compatibility with photoresists (Fig. 10). Using this detector, the described microscope demonstrates a spatial resolution better than 100 nm. The scheme allows the microscope to operate at multiple wavelengths. For this purpose, multilayer coatings with the desired resonance wavelengths are deposited on the conjugate mirror segments, and the working segment is selected using the moveable diaphragm. Thus, this paper proposed and studied a compact laboratory X-ray proximity microscope that provides digital recording images with a resolution better than 0.6 µm, making it possible to work in the mode of a traditional proximity microscope with registration on photosensitive materials at nanoscale resolution. ACKNOWLEDGMENTS
This work was supported by RFBR Grant Nos. 12-0231678, 13-02-00377, and 13-02-97045 and the Ministry of Education and Science of Russia. 1J.
Kirz and M. Howells, Q. Rev. Biophys. 28, 130 (1995). Chao, J. Kim, S. Rekawa, P. Fischer, and E. H. Anderson, Opt. Express 17(20), 17669–17677 (2009). 3W. Chao, B. Harteneck, J. Liddle et al., Nature 435(7046), 1210 (2005). 4J.-F. Adam, J.-P. Moy, and J. Susini, Rev. Sci. Instrum. 76, 091301 (2005). 5H. Legall, G. Blobel, H. Stiel, W. Sandner, C. Seim, P. Takman, D. H. Martz, M. Selin, U. Vogt, H. M. Hertz, D. Esser, H. Sipma, J. Luttmann, M. Höfer, H. D. Hoffmann, S. Yulin, T. Feigl, S. Rehbein, P. Guttmann, G. Schneider, 2W.
Rev. Sci. Instrum. 86, 063701 (2015) U. Wiesemann, M. Wirtz, and W. Diete, Opt. Express 20(16), 18362–18369 (2012). 6K. W. Kim, Y. Kwon, K.-Y. Nam, J.-H. Lim, K.-G. Kim, K. S. Chon, B. H. Kim, D. E. Kim, J. G. Kim, B. N. Ahn, H. J. Shin, S. Rah, K.-H. Kim, J. S. Chae, D. G. Gweon, D. W. Kang, S. H. Kang, J. Y. Min, K.-S. Choi, S. E. Yoon, E.-A. Kim, Y. Namba, and K.-H. Yoon, Phys. Med. Biol. 51, N99–N107 (2006). 7P. A. C. Takman, H. Stollberg, G. A. Johansson, A. Holmberg, M. Lindblom, and H. M. Hertz, J. Microsc. 226(2), 175–181 (2007). 8M. Barysheva, A. Pestov, N. Salashchenko, M. Toropov, and N. Chkhalo, Phys.-Usp. 55(7), 681 (2012). 9T. Ejima, F. Ishida, H. Murata, M. Toyoda, T. Harada, T. Tsuru, T. Hatano, M. Yanagihara, M. Yamamoto, and H. Mizutani, Opt. Express 18(7), 7203 (2010). 10L. Juschkin and R. Freiberger, Proc. SPIE 7360, 736001 (2009). 11I. A. Artioukov and A. V. Vinogradov, Opt. Lett. 20(24), 2451 (1995). 12F. Barkusky et al., Rev. Sci. Instrum. 76, 105102 (2005). 13C. Montcalm et al., Proc. SPIE 3331, 42 (1998). 14T. Haga et al., Jpn. J. Appl. Phys., Part 1 42, 3771 (2003). 15S. Yu. Zuev, A. E. Pestov, N. N. Salashchenko, M. N. Toropov, N. I. Chkhalo, and A. V. Shcherbakov, Bull. Russ. Acad. Sci.: Phys. 77(1), 6–9 (2013). 16N. I. Chkhalo, S. V. Golubev, D. Mansfeld, N. N. Salashchenko, L. A. Sjmaenok, and A. V. Vodopyanov, J. Micro/Nanolithogr., MEMS, MOEMS 11, 021123 (2012). 17S. S. Harilal, T. Sizyuk, V. Sizyuk, and A. Hassanein, Appl. Phys. Lett. 96, 111503 (2010). 18T. Ejima, Y. Neichi, F. Ishida, and M. Yanagihara, J. Phys.: Conf. Ser. 463, 012055 (2013). 19P. Aruev, M. Barysheva, B. Ber, N. Zabrodskaya, V. Zabrodskii, A. Lopatin, A. Pestov, M. Petrenko, V. Polkovnikov, N. Salashchenko, V. Sukhanov, and N. Chkhalo, Quantum Electron. 42(10), 943 (2012). 20N. I. Chkhalo, A. Yu. Klimov, V. V. Rogov, N. N. Salashchenko, and M. N. Toropov, Rev. Sci. Instrum. 79, 033107 (2008). 21B. Lagarde, M. Bordessoule, G. Cauchon, D. Dallé, K. Desjardins, S. Hustache, C. Miron, C. Nicolas, and F. Polack, J. Phys.: Conf. Ser. 425, 152002 (2013).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 165.123.34.86 On: Wed, 01 Jul 2015 11:32:17
