A 50nm spatial resolution EUV imaging–resolution dependence on object thickness and illumination...

A 50nm spatial resolution EUV imaging–resolution dependence on object thickness and

illumination bandwidth

Przemyslaw W. Wachulak,1,*

Andrzej Bartnik,1 Henryk Fiedorowicz,

1 and Jerzy

Kostecki1

1Institute of Optoelectronics, Military University of Technology, ul. gen. S. Kaliskiego 2, 00-908 Warsaw, Poland *[email protected]

Abstract: In this paper we report a desk-top microscopy reaching 50nm spatial resolution in very compact setup using a gas-puff laser plasma EUV source. The thickness of an object and the bandwidth of illuminating radiation were studied in order to estimate their quantitative influence on the EUV microscope spatial resolution. EUV images of various thickness objects obtained by illumination with variable bandwidth EUV radiation were compared in terms of knife-edge spatial resolution to study the bandwidth/object thickness parasitic influence on spatial resolution of the EUV microscope.

©2011 Optical Society of America

OCIS codes: (050.1965) Diffractive lenses; (110.7440) X-ray imaging; (180.7460) X-ray microscopy; (340.7480) X-rays, soft x-rays, extreme ultraviolet (EUV).

References and links

1. D. Attwood, Soft X-Rays and Extreme Ultraviolet Radiation (Cambridge University, 1999). 2. W. Chao, J. Kim, S. Rekawa, P. Fischer, and E. H. Anderson, “Demonstration of 12 nm resolution Fresnel zone

plate lens based soft x-ray microscopy,” Opt. Express 17(20), 17669–17677 (2009). 3. P. A. C. Takman, H. Stollberg, G. A. Johansson, A. Holmberg, M. Lindblom, and H. M. Hertz, “High-resolution

compact X-ray microscopy,” J. Microsc. 226(2), 175–181 (2007). 4. G. Vaschenko, C. Brewer, F. Brizuela, Y. Wang, M. A. Larotonda, B. M. Luther, M. C. Marconi, J. J. Rocca, C.

S. Menoni, E. H. Anderson, W. Chao, B. D. Harteneck, J. A. Liddle, Y. Liu, and D. T. Attwood, “Sub-38 nm resolution tabletop microscopy with 13 nm wavelength laser light,” Opt. Lett. 31(9), 1214–1216 (2006).

5. P. W. Wachulak, A. Bartnik, and H. Fiedorowicz, “Sub-70 nm resolution tabletop microscopy at 13.8 nm using a compact laser-plasma EUV source,” Opt. Lett. 35(14), 2337–2339 (2010).

6. D. S. DiCicco, D. Kim, R. Rosser, and S. Suckewer, “First stage in the development of a soft-x-ray reflection imaging microscope in the Schwarzschild configuration using a soft-x-ray laser at 18.2 nm,” Opt. Lett. 17(2), 157–159 (1992).

7. L. B. Da Silva, J. E. Trebes, S. Mrowka, T. W. Barbee, Jr., J. Brase, J. A. Koch, R. A. London, B. J. Macgowan, D. L. Matthews, D. Minyard, G. Stone, T. Yorkey, E. Anderson, D. T. Attwood, and D. Kern, “Demonstration of x-ray microscopy with an x-ray laser operating near the carbon K edge,” Opt. Lett. 17(10), 754–756 (1992).

8. M. Wieland, Ch. Spielmann, U. Kleineberg, T. Westerwalbesloh, U. Heinzmann, and T. Wilhein, “Toward time-resolved soft X-ray microscopy using pulsed fs-high-harmonic radiation,” Ultramicroscopy 102(2), 93–100 (2005).

9. M. Kishimoto, M. Tanaka, R. Tai, K. Sukegawa, M. Kado, N. Hasegawa, H. Tang, T. Kawachi, P. Lu, K. Nagashima, H. Daido, Y. Kato, K. Nagai, and H. Takenaka, “Development of soft X-ray microscopy system using X-ray laser in JAERI Kansai,” J. Phys. IV 104, 141–143 (2003).

10. I. A. Artioukov, A. V. Vinogradov, V. E. Asadchikov, Y. S. Kas’yanov, R. V. Serov, A. I. Fedorenko, V. V. Kondratenko, and S. A. Yulin, “Schwarzschild soft-x-ray microscope for imaging of nonradiating objects,” Opt. Lett. 20(24), 2451 (1995).

11. G. Vaschenko, F. Brizuela, C. Brewer, M. Grisham, H. Mancini, C. S. Menoni, M. C. Marconi, J. J. Rocca, W. Chao, J. A. Liddle, E. H. Anderson, D. T. Attwood, A. V. Vinogradov, I. A. Artioukov, Y. P. Pershyn, and V. V. Kondratenko, “Nanoimaging with a compact extreme-ultraviolet laser,” Opt. Lett. 30(16), 2095–2097 (2005).

12. F. Brizuela, Y. Wang, C. A. Brewer, F. Pedaci, W. Chao, E. H. Anderson, Y. Liu, K. A. Goldberg, P. Naulleau, P. Wachulak, M. C. Marconi, D. T. Attwood, J. J. Rocca, and C. S. Menoni, “Microscopy of extreme ultraviolet lithography masks with 13.2 nm tabletop laser illumination,” Opt. Lett. 34(3), 271–273 (2009).

13. K. 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, “Compact soft x-ray transmission microscopy with sub-50nm spatial resolution,” Phys. Med. Biol. 51(99–N), 107 (2006).

#143602 - $15.00 USD Received 4 Mar 2011; revised 8 Apr 2011; accepted 11 Apr 2011; published 2 May 2011(C) 2011 OSA 9 May 2011 / Vol. 19, No. 10 / OPTICS EXPRESS 9541

14. L. Juschkin, R. Freiberger, and K. Bergmann, “EUV microscopy for defect inspection by dark-field mapping and zone plate zooming,” J. Phys.: Conf. Ser. 186, 012030 (2009).

15. H. Fiedorowicz, A. Bartnik, R. Jarocki, J. Kostecki, J. Krzywinski, J. Mikołajczyk, R. Rakowski, A. Szczurek, and M. Szczurek, “Compact laser plasma EUV source based on a gas puff target for metrology applications,” J. Alloy. Comp. 401(1-2), 99–103 (2005).

16. P. W. Wachulak, A. Bartnik, H. Fiedorowicz, T. Feigl, R. Jarocki, J. Kostecki, R. Rakowski, P. Rudawski, M. Sawicka, M. Szczurek, A. Szczurek, and Z. Zawadzki, “A compact, quasi-monochromatic laser-plasma EUV source based on a double-stream gas-puff target at 13.8 nm wavelength,” Appl. Phys. B 100(3), 461–469 (2010).

17. R. Rakowski, A. Bartnik, H. Fiedorowicz, F. Gaufridy de Dortan, R. Jarocki, J. Kostecki, J. Mikołajczyk, L. Ryć, M. Szczurek, and P. Wachulak, “Characterization and optimization of the laser-produced plasma EUV source at 13.5nm based on a double-stream Xe/He gas puff target,” Appl. Phys. B 101(4), 773–789 (2010).

18. R. L. Kelly, “Atomic and ionic spectrum lines below 2000 angstroms: hydrogen through krypton,” J. Phys. Chem. Ref. Data 16(1), (1987).

19. S. S. Churilov, Y. N. Joshi, and J. Reader, “High-resolution spectrum of xenon ions at 13.4 nm,” Opt. Lett. 28(16), 1478–1480 (2003).

20. G. O’Sullivan, “Charge-dependent wavefunction collapse in ionised xenon,” J. Phys. B 15(21), L765–L771 (1982).

21. N. I. S. T. Atomic Spectra Database Lines Data, http://physics.nist.gov/PhysRefData/ASD/lines_form.html

22. J. M. Heck, D. T. Attwood, W. MeyerIlse, and E. H. Anderson, “Resolution determination in Xray microscopy: an analysis of the effects of partial coherence and illumination spectrum,” J.X-Ray Sci. Technol. 8, 95 (1998).

23. P. W. Wachulak, A. Bartnik, H. Fiedorowicz, P. Rudawski, R. Jarocki, J. Kostecki, and M. Szczurek, ““Water window” compact, table-top laser plasma soft X-ray sources based on a gas puff target,” Nucl. Instrum. Methods Phys. Res. B 268(10), 1692–1700 (2010).

1. Introduction

Future developments of nanoscience and nanotechnology demand imaging tools capable of resolving features on nanometer scale. Various imaging methods and techniques are currently under development, one of them being an extreme ultraviolet (EUV) and soft X-ray (SXR) microscopy, based on Fresnel zone plates [1], that is under active pursuit worldwide. The short wavelength laboratory sources provide a path to improve the spatial resolution thus photon based SXR microscopy is capable of reaching resolutions down to 12nm using synchrotron radiation [2]. The demonstration of bright EUV and SXR sources paved a way for the development of table top microscopes that can render images of nanoscale objects with exposures as short as a few seconds and a spatial resolution approaching that of synchrotron based microscopes [3–5]. The development of bright, compact, short wavelength sources will be a significant step forward in commercialization of high resolution imaging tools in near future.

Many imaging experiments were carried out so far, using coherent and incoherent EUV/SXR sources. A 700nm half-pitch resolution images with the use of EUV recombination laser at λ = 18.2nm has been reported in the early imaging work [6]. 75nm resolution was reported employing SXR laser at λ = 4.48nm pumped by fusion-class NOVA laser limiting image acquisition by laser repetition rate to several shots per day [7]. Recently different approaches for sub-micrometer resolution imaging have emerged due to the development of smaller-scale short-wavelength sources such as high-order harmonics [8], SXR lasers [9] and incoherent laser-plasma based sources [10]. Using radiation from a capillary discharge laser, λ = 46.9nm wavelength, EUV images were obtained with a spatial resolution of 120-150nm [11]. A 13.2nm wavelength radiation from Ni-like Cd EUV laser allowed for a 55nm in reflection mode [12] and sub-38nm resolution nano-imaging in transmission mode [4], using pulses of 1J, 8ps for pumping. A quasi-monochromatic emission from incoherent SXR source based on liquid nitrogen, at λ = 2.88nm, in the “water window” range, allowed to demonstrate SXR microscopy with sub-50nm spatial resolution (~17λ) [13]. Finally, using xenon based gas discharge EUV source, Schwarzschild objective and Fresnel zone-plate optic for second magnification step, EUV imaging was demonstrated reaching the spatial resolution of ~100nm [14].

In this paper we report a desk-top microscopy using a gas-puff laser plasma EUV source reaching 50nm spatial resolution in very compact setup. This type of EUV microscope that utilizes short wavelength radiation from a very compact source will allow imaging of objects


with high spatial resolution. The spatial resolution of the microscope, however, can be affected by many parasitic factors. The thickness of an object and the bandwidth of illuminating radiation were studied in order to estimate quantitative influence on the EUV microscope spatial resolution. Object with thickness much larger than the depth of focus of the microscope objective will degrade the resolution, moreover if the objective is highly dispersive, the illumination bandwidth plays a key role as well, since its focal length dependence on wavelength. EUV images obtained by illumination of the object with variable bandwidth EUV radiation were compared to study the bandwidth influence on spatial resolution of the EUV microscope.

2. Experimental setup

The microscope was equipped with an ellipsoidal mirror with Mo/Si coating to focus extreme ultraviolet (EUV) radiation onto an object. A Fresnel zone plate objective was used to form the magnified image onto a EUV-sensitive CCD camera in the transmission mode. The use of the gas puff target eliminates the debris production problem associated with solid targets. Quasi-monochromatic EUV radiation, required for the Fresnel optics, was produced by spectral selection of a single line emitting at 13.8 nm wavelength from argon plasma or quasi-continuum emission in band (13-14nm wavelength range) from xenon plasma. Moreover two distinct objects were imaged – Cu mesh with thickness of ~4μm and carbon foil with holes, coated additionally with a thin layer of gold having total thickness of 70nm. EUV images of the sample objects have been obtained with the half-pitch spatial resolution approaching ~50 nm (3.7λ) in a very compact set up. The scheme and experimental arrangement are shown in Fig. 1(a) and 1(b), respectively.

Fig. 1. (Color online) (a) scheme and (b) experimental arrangement of the EUV microscope (not to scale) using a laser-plasma EUV source based on gas puff target.

The laser plasma EUV source, used in the experiment, has been developed for EUV metrology applications in frame of the MEDEA + project [15] and later modified for quasi-monochromatic emission in 13-14nm wavelength range, described in details elsewhere [16]. This source has an advantage over other compact sources that is the possibility to change the working gas, thus allowing to change both the peak emission wavelength and the inverse relative bandwidth of the emission.

Firstly, to study the bandwidth influence on spatial resolution of the EUV microscope Ar and Xe plasmas were produced using pumping laser pulses from Nd:YAG laser (Eksma) with pulse duration of 4ns and energy 0.74J. The plasmas radiate in a very broad range of wavelengths, dominantly in the EUV range (5-50nm) and by using additional spectral filtering it is possible to shape the spectral emission of the source. The source can operate up to 10Hz

repetition rate. A pressure of 2x103

bar was constantly maintained in the chamber during the source operation. The experimental setup is extremely compact. The microscope is located inside the vacuum chamber, 24cm in diameter and 35cm in length and the entire system fits on top of a single 2x0.6m

2 optical table.


EUV radiation from the plasma was collected, focused and spectrally filtered by an ellipsoidal, off-axis, 80mm in diameter mirror with Mo/Si multilayers. The mirror was

corrected for the spherical aberrations. The multilayers were optimized for 13.5+/0.5nm (FWHM) wavelength range and incidence angle of 45 degrees. The theoretical reflectivity of the mirror at 13.5nm wavelength is 37.7% for an unpolarized light from the laser-plasma EUV source. The mirror was designed to image the plasma with a unity lateral magnification, having both the object and image distances equal to 254mm. It was developed in co-operation with Reflex s.r.o., at present Rigaku Inc., (mirror substrate) and Fraunhofer Institut für Angewandte Optik und Feinmechanik (coating).

The laser plasma source was optimized for efficient EUV radiation generation from Ar [16] and Xe [17] plasma. Previously measured [16], the in band (λ = 13-14nm) photon flux

was equal to 10

8.8 0.5 10 photons per pulse in a horizontally elongated spot with FWHM

widths 1.09x0.39mm2. That corresponds to 1.29µJ/pulse. For the Xe plasma, due to much

more efficient EUV photon production and the risk of ablating the zone plate objective, we did not optimize the source particularly for Xe and still obtained the photon flux approximately 5 times larger than in case of Ar plasma, with similar plasma size. For both plasmas, to eliminate longer wavelengths (λ>18nm) a 100nm thick, 10 mm diameter, free-standing Zr filter (Lebow) was used. The Zr filter was positioned ~4-5mm upstream the object. The EUV spectra, emitted from the source, have been measured using the flat-field reflection grating spectrometer equipped with a 1200-line/mm grating with varied groove spacing (Hitachi), 25.5μm entrance slit and 1300x400 pixels back illuminated CCD camera (Princeton Instruments). From geometry of the spectrometer and the grating, the inverse

relative bandwidth (IRB) of the spectrometer was estimated to be / 585 . The

spectrometer was calibrated using eight most prominent Ar lines, the fitting polynomial was a

parabolic function described by equation y = 2.15·106

·x2 + 9.23·10

3·x + 6.18, where y-value

is the wavelength expressed in [nm] while x is a horizontal CCD pixel. The spectrum for Ar and Xe plasmas, corrected for Zr filter transmission, both direct and

spectrally filtered by condenser optic are shown in Fig. 2. Figure 2(a) depicts Ar emission spectrum in wavelength range from 9 to 20nm. This quasi-monochromatic spectrum has two main spectral peaks, namely 2p

63p-2p

65d at 13.793nm and 2p

63s-2p

65p at 13.844nm

transitions in ArVIII

[18]. The IRB (FWHM) of filtered spectrum is λ/Δλ = 140 at λ = 13.84nm. Figure 2(b) shows emission spectrum from Xe plasma in the same wavelength range for comparison. The spectrum is more complex, quasi-continuous, filling in the entire reflection band of the condenser. The main spectral transitions are 4p

64d

8-4p

54d

9 and 4p

64d

8-4p

64d

74f in

XeXI

from 11 to 13nm wavelength, 4p64d

8-4p

64d

74f and 4p

64d

8-4p

64d

75p in Xe

XI between 13

and 14 nm wavelength and predominantly 4d9-4d

85p and 4d

9-4d

84f transitions in Xe

X, 4d

10-

4d95p and 3d

104s-3d

104p in Xe

IX and Xe

XXVI, respectively, at wavelengths above 14nm,

according to [19–21]. The inverse relative bandwidth of filtered Xe spectrum is λ/Δλ = 14, at peak emission around λ = 13.78nm wavelength. The values of inverse relative bandwidth might be underestimated, however, due to relatively wide spectral response of our spectrometer (0.24Å), but still the differences for Ar and Xe plasmas in terms of character of their spectral emissions and the IRB (at least 10x difference) are evident.


Fig. 2. (Color online) Spectrum of radiation emitted from the EUV source for Ar plasma (a) and Xe plasma (b).

Second interesting experiment was related to study of the object thickness influence on spatial resolution. For that two different objects, with different thicknesses in relation to depth of focus (DOF) of the microscope, were used. First one is a copper G2000HS Fine Square Mesh (SPI Supplies) with 12.5µm period, 5µm width bar and 4µm thick (~11x DOF). The second one is a Quantifoil holey 300M carbon foil supported on a steel mesh (SPI supplies), 10nm thick according to manufacturer’s specifications, having 1.5µm diameter holes spaced on a square grid with 2.5µm period. To improve the optical contrast carbon foil was additionally coated with ~60nm thick layer of gold, having a transmission of 4.26% at 13.84nm wavelength, thus in total having thickness of ~70nm (0.2x DOF). Typical SEM images of the mesh and the perforated foil are shown in Fig. 3(a) and 3(b), respectively.

Fig. 3. SEM images of two objects used during studies of object thickness influence on the EUV microscope spatial resolution: a copper mesh, 4mm thick (a) and perforated carbon foil, coated additionally by a thin layer of gold, total thickness of 70nm (b).

The objects, placed 254mm from the mirror, are imaged using a Fresnel zone plate (ZP) objective onto EUV sensitive CCD camera iKon-M (Andor) with 1024x1024 pixels and 13x13µm

2 pixel size. The ZP was fabricated by Zone Plates Ltd., using electron beam

lithography in 220nm thick PMMA (polymethyl methacrylate) layer spin coated on top of a 50nm thick silicon nitride Si3N4 membrane. ZP diameter is equal to D = 200µm, number of zones NZP = 1000 and an outer zone width Δr = 50nm. The ZP was fabricated and optimized

for λ = 13.5nm, resulting in focal length of /f D r equal to 740.7µm and numerical

aperture / 2NA r = 0.135. In case of quasi-monochromatic illumination at λ =

13.84nm from Ar plasma the focal length of the ZP was shifted to f = 722.5µm and the


numerical aperture was equal to NA = 0.138. For spectrally broader emission from Xe it is

more difficult to state the ZP parameters, because of high ZP dispersion, 1~ , ~f NA

,

however for λ = 13.78nm wavelength peak emission the focal length was equal to f =

725.7µm. The DOF, calculated as 22/DOF NA was equal to +/363.4nm. For both

types of spectra and both objects the magnifications, ranging from 520 to 840x, were used, adjusted by changing the camera-ZP distance and refocusing. Details about magnification, pixel size and field of view (FOV) are in Table 1. More information regarding the EUV microscope can be found in [5]. Geometrical numerical apertures of the collecting ellipsoidal mirror in horizontal and vertical directions are equal to NAcH = 0.11 and NAcV = 0.15, respectively, and are similar to the numerical aperture of the ZP, NAZP = 0.138, thus

providing incoherent illumination [22], since , ,

/H V cH V ZP

NA NA equal to , 0.8,1.1H V

.

The ZP was mounted on three axis translation stage driven by vacuum compatible step motor actuators (Standa). A piezoelectric, 25µm travel, single axis flexure stage, model NF15A (Thorlabs), was used to adjust precisely the object-ZP distance at a rate of 3V/µm with theoretical resolution reaching 10nm. To provide a conical illumination of the object and avoid stray light through the ZP a circular beam block, 12mm in diameter, was placed ~15cm from the ZP. In case of “binary” transmission mesh object and quasi-monochromatic radiation from Ar plasma to obtain a single EUV image 50 EUV pulses were necessary, at 2Hz repetition rate, while in case of quasi-continuum radiation spectrum from Xe, due to the increased flux, the required exposure dropped to 10 EUV pulses at 2Hz repetition rate. The source can operate at up to 10Hz repetition rate, however, the pressure buildup in the microscope chamber might cause re-absorption of EUV photons in neutral gas. The CCD

camera was cooled down to 20 °C to decrease intrinsic noise during the image acquisition.

3. Experimental results

Four sets of measurements were performed for two types of objects (mesh and foil) and for two plasmas (Ar and Xe). For each case the microscope alignment was optimized to provide a uniform illumination in the entire FOV, also, during the image acquisition, the object-ZP distance was changed using the piezoelectric stage by Δz~330nm, corresponding to 1V/step,

to obtain the sharpest possible EUV image, over the z-range of ~+/20μm from the ZP focal point. The Δz was chosen to be smaller than DOF. From the entire set of images, for each object/bandwidth combination, the “sharpest” EUV image was chosen for subsequent resolution measurements. Resolution of the microscope was assessed by a well established knife edge (KE) test. For incoherent illumination the 10-90% intensity transition across a sharp edge corresponds to a well known Rayleigh resolution and to twice the value of half-pitch grating resolution of the optical system [1].

Typical EUV images of the mesh object under illumination by Ar (a,c) and Xe (b,d) plasma radiation are depicted in Fig. 4. Small rectangular boxed regions, where subsequent KE resolution measurements were carried out, are magnified to show an obvious resolution decrease by edge blurring for the case of Xe illumination, shown in Fig. 4(d)). Similarly, for the perforated foil object, Fig. 5 depicts the EUV images obtained for different illumination bandwidths. An additional object blurring can be observed due to the 10x wider illumination bandwidth from Xe plasma.

Table 1 shows experimental details for each set of measurements, such as object thickness, exposure, magnification, image pixel size and field of view. It also shows the results of half-

pitch KE measurements - KEr , based on 6 independent KE measurements for each

object/bandwidth combination and obtained based on period of imaged structure for higher

accuracy. KE

denotes standard deviation calculated from the measurements. Lineouts across

EUV images were assessed by averaging 5 adjacent lines in the EUV image to improve a signal to noise ratio.


Fig. 4. EUV images of Cu mesh object with Ar (a,c) and Xe (b,d) plasma illumination. (c,d) are the magnified subsections of larger EUV images indicated by boxed regions showing the edge in detail.

Fig. 5. EUV images of perforated carbon/Au foil object with Ar (a,c) and Xe (b,d) plasma illumination. (c,d) are the magnified subsections of the EUV images in boxed regions showing a single hole in detail.

Table 1. EUV Imaging Experimental Details and Resolution Measurement Results for Different Objects and Illumination Bandwidth

Object mesh foil

Thickness t 11 x DOF 0.2 x DOF

Exposure for 2Hz repetition rate [s] 25 10 50 50

Illumination type Ar Xe Ar Xe

IRB λ/Δλ 140 14 140 14

Magnification [x] 840 523

Image pixel size [nm2] 15.9x15.9nm2 25.9x25.9nm2

Field of view [μm2] 16.3x16.3μm2 26.4x26.4μm2

KE half-pitch resolution rKE

KEr [nm] 72.7 151.8 51.0 139.9

KE [nm] 5.0 11.8 10.6 15.3


The best KE half-pitch resolution was measured to be 51.0+/10.6nm based on 6 independent measurements for thin, perforated foil object and quasi-monochromatic illumination from Ar plasma. Although, less prominent, the influence of the object thickness on the spatial resolution can also be found from the measurements, because a slightly worse resolution of

72.7+/5.0nm was obtained for thicker mesh object. This corresponds to previously assessed

spatial resolution of the microscope equal to 69.4+/4.0nm, reported recently in [5]. A real resolution change was noted, however, for broad band, quasi-continuous illumination from Xe

plasma. For “thin” object the measured spatial half-pitch resolution was 139.9+/15.3nm and

for the “thick” one 151.8+/11.8nm. Each measurement and statistical resolutions are presented in Fig. 6. Additionally the ZP theoretical resolution limit in terms of half-pitch resolution, under an incoherent illumination, equal to 0.61Δr [1], was indicated as a dashed line.

Fig. 6. (Color online) Graphic representation of the KE resolution measurements for various

objects and types of illumination with error bars corresponding to +/ standard deviation calculated from the measurements.

Typical KE lineouts, obtained in boxed regions, for the images in Figs. 4 and 5, for each object/bandwidth combination and the theoretical KE resolution limited by ZP outer zone width Δr, are depicted in Fig. 7. Figure shows also the full (10-90%) and half-pitch Rayleigh resolution measurements.

Fig. 7. (color online) Typical KE lineouts indicating 10-90% intensity transition in the EUV image related to Rayleigh resolution criterion for both objects and different illumination bandwidths and the theoretical KE limit.


4. Discussion of the results

The theoretical half-pitch resolution of the microscope can be expressed as

/(2 )KE ZPk NA k rr , where k is illumination dependent and resolution test specific

constant [22], NAZP is the numerical aperture of the objective ZP, for incoherent illumination (k = 0.61) this resolution is equal to 30.5nm, so it is better than the measured half-pitch resolution of 51nm in the best case. This is due to the fact, that the zone plate is a highly dispersive diffractive element. The theoretical resolution can only be achieved for

monochromatic radiation, when the monochromaticity criterion / ZPN will be fulfilled,

namely the IRB will be larger than the number of ZP zones. In our case, even for the quasi-monochromatic radiation from Ar plasma, this criterion is not satisfied, thus the relatively small IRB of our EUV source will introduce some achromatic blurring to the image. Secondly the resolution loss is related to the object thickness. If the object thickness does not fulfill the

relation t DOF , its thickness will influence the resolution of the EUV images as well,

thus perforated foil object, having thickness much smaller than the ZP depth of focus, allows to avoid any spatial resolution decrease due to the object thickness and to study only the dependence of the illumination bandwidth on resolution.

To estimate the influence of both the EUV bandwidth and the object thickness on spatial resolution a Gaussian type point spread function (PSF) was assumed. Any image, convolved with PSF of non-zero width will suffer from resolution loss. The wider the PSF the more high resolution details, or high spatial frequency components in the image will be removed. Moreover, many PSF functions, attributed to various resolution decreasing factors, can be convolved together leading to a total PSF function of an optical system. Gaussian PSF is neither the best, nor the most accurate estimate, but allows for quick de-convolution of the results to obtain a quick estimation of contribution of each factor on the spatial resolution. The results of de-convolution are shown in Table 2. For monochromatic illumination the theoretical half-pitch resolution achievable will be 30.5nm. For quasi-monochromatic Ar

plasma emission, / ZPN , the resolution is worse, equal to 51nm. Thus it can be

attributed to a Gaussian type PSFAr having FWHM width equal to ~35nm. It is even more pronounced for Xe plasma, since for spatial resolution of 139.9nm, comparing to 51nm for Ar illumination, the required FWHM PSFXe width is ~113nm. To quantize the object thickness influence on spatial resolution we can analyze the resolutions obtained for Ar illumination and for both “thick” and “thin” objects. For Ar illumination, if the object is much thicker than the DOF, 11t DOF for mesh, measured resolution is 72.7nm resulting in PSFth width ~45nm,

similarly analyzing the resolutions for both objects and Xe illumination we ended up having PSFth width ~51nm, comparable to the one obtained before for Ar illumination.

Table 2. Estimation of Gaussian PSF Approximation by De-Convolution Based on the KE Resolution Measurements

Gaussian PSF approximation by de-convolution

Experimental half-pitch resolution

[nm]

FWHM Gaussian PSF width [nm] (parasitic effect)

Theoretical ZP resolution 0.61Δr for monochromatic illumination, λ/Δλ>1000,

30.5 0

Quasi-monochromatic illumination from Ar plasma λ/Δλ = 140, “thin” object

51.0 35.3

Quasi-monochromatic illumination from Ar plasma λ/Δλ = 140, “thick” object, thickness t = 11xDOF

72.7 44.8

Quasi-continuum illumination from Xe plasma, λ/Δλ = 14, “thin” object

139.9 112.7

Quasi-continuum illumination from Xe plasma, λ/Δλ = 14, + “thick” object

151.8 50.9


5. Conclusions

In summary, we have demonstrated near 50nm spatial resolution desk-top EUV transmission microscope based on a Fresnel diffractive optics and a gas-puff, laser-plasma EUV source. This desk-top microscope, under the incoherent illumination, allows to capture images at 13.8nm wavelength with the spatial resolution of 51nm and the exposure time of 50s, comparable to larger table-top systems. We have also presented a detailed analysis of illumination bandwidth and object thickness influence on the spatial resolution. Using either quasi-monochromatic radiation from Ar plasma or quasi-continuous from Xe plasma we were able to quantify a bandwidth-related resolution decreasing factor. On the other hand using “thick” Cu mesh object and “thin” perforated foil object we have found quantitatively how the object thickness relates to the resolution decreasing factor and expressed both factors in terms of Gaussian type PSF functions.

Future improvements of the pumping system will reduce the exposure time by a factor of 5, increasing the repetition rate to 10Hz. Moreover, employing an additional zone plate to decrease the illumination area will improve significantly the throughput of the imaging system, further decreasing the acquisition time. Furthermore, using SXR radiation from Ar or N2 gas puff targets [23] it will be possible to build a compact SXR microscope in the “water-window” spectral range. These results should be useful for the realization of high resolution desk-top imaging systems for actinic inspection of EUV lithography masks and defect characterization or as a compact imaging tool that can be used in material science, biology and nanotechnology.

Acknowledgments

The research was supported by the Foundation for Polish Science under the HOMING 2009 Programme (grant number HOM2009/14B), the EC’s 7. Framework Program (LASERLAB-EUROPE - grant agreement n° 228334 and COST Action MP0601, also co-funded by Polish Ministry of Science and Education, decision number 816/N-COST/2010/0), European Economic Area (EEA) Grants. We would like to thank to Dr. R. Jarocki, Dr. R. Rakowski, and Dr. M. Szczurek, for their contribution to the development of the EUV source, to Prof. L. Pina and Dr. T. Feigl, for fabrication of the multilayer condenser optic, to Prof. M. C. Marconi and Prof. C. S. Menoni for help during the preparation of this manuscript and to Michaela Martinkova for her constructive comments and participation in the experiment during her EUV Training Programme at MUT.


A 50nm spatial resolution EUV imaging–resolution dependence on object thickness and illumination...

Documents

Transcript of A 50nm spatial resolution EUV imaging–resolution dependence on object thickness and illumination...