Infrared diffractive filtering for extreme ultraviolet multilayer Bragg reflectors

We report on the development of a hybrid mirror realized by integrating an EUV-reflecting multilayer coating with a lamellar grating substrate. This hybrid mirror acts as an efficient Bragg reflector for extreme ultraviolet (EUV) radiation at a given wavelength while simultaneously providing spectral-selective suppression of the specular reflectance for unwanted longer-wavelength radiation due to the grating phase-shift resonance. The test structures, designed to suppress infrared (IR) radiation, were fabricated by masked deposition of a Si grating substrate followed by coating of the grating with a Mo/Si multilayer. To give the proof of principle, we developed such a hybrid mirror for the specific case of reflecting 13.5 nm radiation while suppressing 10 μm light, resulting in 61% reflectance at the wavelength of 13.5 nm together with the 70 × suppression rate of the specular reflection at the wavelength of 10 μm, but the considered filtering principle can be used for a variety of applications that are based on utilization of broadband radiation sources. ©2013 Optical Society of America OCIS codes: (310.4165) Multilayer design; (340.7480) X-rays, soft x-rays, extreme ultraviolet (EUV); (130.7408) Wavelength filtering devices; (050.5080) Phase shift; (050.1950) Diffraction gratings. References and links 1. J.-M. André, R. Benbalagh, R. Barchewitz, M.-F. Ravet, A. Raynal, F. Delmotte, F. Bridou, G. Julie, A. Bosseboeuf, R. Laval, G. Soullié, C. Rémond, and M. Fialin, “Soft x-ray multilayer monochromator with improved resolution and low specular background,” XRay Spectrom. 30, 212–215 (2001). 2. R. Benbalagh, J.-M. Andre, R. Barchewitz, P. Jonnard, G. Julie, L. Mollard, G. Rolland, C. Remond, P. Troussel, R. Marmoret, and E. O. Filatova, “Lamellar multilayer amplitude grating as soft-X-ray Bragg monochromator,” Nucl. Instrum. Methods Phys. Res. A 541(3), 590–597 (2005). 3. I. V. Kozhevnikov, R. van der Meer, H. M. J. Bastiaens, K. J. Boller, and F. Bijkerk, “High-resolution, highreflectivity operation of lamellar multilayer amplitude gratings: identification of the single-order regime,” Opt. Express 18(15), 16234–16242 (2010). 4. D. L. Voronov, M. Ahn, E. H. Anderson, R. Cambie, C.-H. Chang, E. M. Gullikson, R. K. Heilmann, F. Salmassi, M. L. Schattenburg, T. Warwick, V. V. Yashchuk, L. Zipp, and H. A. Padmore, “High-efficiency 5000 lines/mm multilayer-coated blazed grating for extreme ultraviolet wavelengths,” Opt. Lett. 35(15), 2615–2617 (2010). 5. D. L. Voronov, E. H. Anderson, R. Cambie, S. Cabrini, S. D. Dhuey, L. I. Goray, E. M. Gullikson, F. Salmassi, T. Warwick, V. V. Yashchuk, and H. A. Padmore, “A 10,000 groove/mm multilayer coated grating for EUV spectroscopy,” Opt. Express 19(7), 6320–6325 (2011). 6. A. J. R. van den Boogaard, E. Louis, F. A. van Goor, and F. Bijkerk, “Optical element for full spectral purity from IR-generated EUV light sources,” Proc. SPIE 7271, 72713B, 72713B-6 (2009). 7. A. J. R. van den Boogaard, F. A. van Goor, E. Louis, and F. Bijkerk, “Wavelength separation from extreme ultraviolet mirrors using phaseshift reflection,” Opt. Lett. 37(2), 160–162 (2012). 8. J. Benschop, V. Banine, S. Lok, and E. Loopstra, “Extreme ultraviolet lithography: status and prospects,” J. Vac. Sci. Technol. B 26(6), 2204–2207 (2008). 9. I. V. Fomenkov, B. La Fontaine, D. Brown, I. Ahmad, P. Baumgart, N. R. Böwering, D. C. Brandt, A. N. Bykanov, S. De Dea, A. I. Ershov, N. R. Farrar, D. J. Golich, M. J. Lercel, D. W. Myers, C. Rajyaguru, S. N. #191179 $15.00 USD Received 24 May 2013; revised 1 Jul 2013; accepted 1 Jul 2013; published 9 Jul 2013 (C) 2013 OSA 15 July 2013 | Vol. 21, No. 14 | DOI:10.1364/OE.21.016964 | OPTICS EXPRESS 16964 Srivastava, Y. Tao, and G. O. Vaschenko, “Development of stable extreme-ultraviolet sources for use in lithography exposure systems,” J. Micro/Nanolith. MEMS MOEMS 11(2), 021110 (2012). 10. J. Fujimoto, T. Abe, S. Tanaka, T. Ohta, T. Hori, T. Yanagida, H. Nakarai, and H. Mizoguchi, “Laser-produced plasma-based extreme-ultraviolet light source technology for high-volume manufacturing extreme-ultraviolet lithography,” J. Micro/Nanolith. MEMS MOEMS 11(2), 021111 (2012). 11. J. Fujimoto, T. Hori, T. Yanagida, T. Ohta, Y. Kawasuji, Y. Shiraishi, T. Abe, T. Kodama, H. Nakarai, T. Yamazaki, and H. Mizoguchi, “Development of laser-produced plasma-based EUV light source technology for HVM EUV lithography,” Proc. SPIE 8332, 83220F, 83220F-13 (2012). 12. V. Y. Banine, K. N. Koshelev, and G. H. P. M. Swinkels, “Physical processes in EUV sources for microlithography,” J. Phys. D Appl. Phys. 44(25), 253001 (2011). 13. C. Mbanaso, A. Antohe, H. Bull, F. Goodwin, A. Hershcovitch, and G. Denbeaux, “Out-of-band radiation mitigation at 10.6 mm by molecular absorbers in laser-produced plasma extreme ultraviolet sources,” J. Micro/Nanolith. MEMS MOEMS 11(2), 021116 (2012). 14. V. P. Belik, Y. M. Zadiranov, N. D. Il’inskaya, A. V. Korlyakov, V. V. Luchinin, M. A. Markosov, R. P. Seisyan, and E. M. Sher, “Free-standing optical filters for a nanolithography source operating in the 12-15 nm wavelength range,” Tech. Phys. Lett. 33(6), 508–511 (2007). 15. N. I. Chkhalo, M. N. Drozdov, E. B. Kluenkov, A. Y. Lopatin, V. I. Luchin, N. N. Salashchenko, N. N. Tsybin, L. A. Sjmaenok, V. E. Banine, and A. M. Yakunin, “Free-standing spectral purity filters for extreme ultraviolet lithography,” J. Micro/Nanolith. MEMS MOEMS 11(2), 021115 (2012). 16. W. A. Soer, M. J. J. Jak, A. M. Yakunin, M. M. J. W. van Herpen, and V. Y. Banine, “Grid spectral purity filters for suppression of infrared radiation in laser-produced plasma EUV sources,” Proc. SPIE 7271, 72712Y, 72712Y-9 (2009). 17. W. A. Soer, P. Gawlitza, M. M. J. W. van Herpen, M. J. J. Jak, S. Braun, P. Muys, and V. Y. Banine, “Extreme ultraviolet multilayer mirror with near-zero IR reflectance,” Opt. Lett. 34(23), 3680–3682 (2009). 18. M. M. J. W. van Herpen, R. W. E. van de Kruijs, D. J. W. Klunder, E. Louis, A. E. Yakshin, S. A. van der Westen, F. Bijkerk, and V. Banine, “Spectral-purity-enhancing layer for multilayer mirrors,” Opt. Lett. 33(6), 560–562 (2008). 19. V. V. Medvedev, A. E. Yakshin, R. W. E. van de Kruijs, V. M. Krivtsun, A. M. Yakunin, K. N. Koshelev, and F. Bijkerk, “Infrared suppression by hybrid EUV multilayer IR etalon structures,” Opt. Lett. 36(17), 3344–3346 (2011). 20. V. V. Medvedev, A. E. Yakshin, R. W. E. van de Kruijs, V. M. Krivtsun, A. M. Yakunin, K. N. Koshelev, and F. Bijkerk, “Infrared antireflective filtering for extreme ultraviolet multilayer Bragg reflectors,” Opt. Lett. 37(7), 1169–1171 (2012). 21. E. Louis, A. E. Yakshin, T. Tsarfati, and F. Bijkerk, “Nanometer interface and materials control for multilayer EUV-optical applications,” Prog. Surf. Sci. 86(11-12), 255–294 (2011). 22. L. D. Landau and E. M. Lifshitz, The classical theory of fields (Butterworth-Heinemann, 1994). 23. A. Barbara, P. Quémerais, E. Bustarret, T. López-Rios, and T. Fournier, “Electromagnetic resonances of subwavelength rectangular metallic gratings,” Eur. Phys. J. D 23, 143–154 (2003). 24. M. G. Moharam and T. K. Gaylord, “Rigorous coupled-wave analysis of metallic surface-relief gratings,” J. Opt. Soc. Am. A 3(11), 1780–1787 (1986). 25. K. N. Koshelev, V. V. Ivanov, V. G. Novikov, V. Medvedev, A. S. Grushin, and V. M. Krivtsun, “RZLINE code modeling of distributed tin targets for laser-produced plasma sources of extreme ultraviolet radiation,” J. Micro/Nanolith. MEMS MOEMS 11(2), 021112 (2012). 26. N. R. Böwering, A. I. Ershov, W. F. Marx, O. V. Khodykin, B. A. M. Hansson, E. L. Vargas, J. A. Chavez, I. V. Fomenkov, D. W. Myers, and D. C. Brandt, “EUV source collector,” Proc. SPIE 6151, 61513R (2006). 27. A. R. Giehl, M. Kessler, A. Grosse, N. Herhammer, and H. Fouckhardt, “Deep reactive ion etching of GaSb in Cl2/Ar-plasma discharges using single-layer soft mask technologies,” J. Micromech. Microeng. 13(2), 238–245 (2003). 28. A. J. R. van den Boogaard, E. Louis, E. Zoethout, S. Müllender, and F. Bijkerk, “Surface morphology of Kr+polished amorphous Si layers,” J. Vac. Sci. Technol. A 28(4), 552 (2010). 29. A. J. R. van den Boogaard, E. Louis, E. Zoethout, K. A. Goldberg, and F. Bijkerk, “Characterization of Mo/Si multilayer growth on stepped topographies,” J. Vac. Sci. Technol. B 29(5), 051803 (2011). 30. D. L. Voronov, P. Gawlitza, R. Cambie, S. Dhuey, E. M. Gullikson, T. Warwick, S. Braun, V. V. Yashchuk, and H. A. Padmore, “Conformal growth of Mo/Si multilayers on grating substrates using collimated ion beam sputtering,” J. Appl. Phys. 111(9), 093521 (2012).


Introduction
The use of diffractive optical elements plays an important role in advanced optical systems that are based on the utilization of radiation emitted by broadband sources.Various spectralselective diffractive elements are applied either for spectral analysis or for spectral filtering.In the past 20 years various combinations of grating structures with multilayer Bragg-reflector coatings were intensively explored as optical elements with improved spectral characteristics specifically for the EUV and soft X-ray ranges of the electromagnetic spectrum [1][2][3][4][5][6][7].
Particularly, such multilayer gratings are attractive for optical applications that require high reflection for radiation in the EUV wavelength range and simultaneously demand suppression of radiation outside this prime range.A most obvious need for radiation with high "spectral purity" can be found in extreme ultraviolet lithography (EUVL) [8].In order to fulfill demands of high productivity, EUVL requires high power radiation sources of EUV.A promising technology for such sources is based on the emission from dense tin plasmas produced by pulsed CO 2 laser radiation with 10.6 μm wavelength [9][10][11][12].These plasmas emit, apart from useful in-band EUV (13.5 ± 0.1 nm) light, also unwanted, so called out-ofband (OoB) radiation, e.g.longer/shorter wavelength EUV, ultraviolet, visible and scattered infrared light from the laser [11,12].All sorts of OoB radiation are efficiently reflected by the mirrors in the optical system.To prevent problems such as mirror heating, wafer heating, resist flare and imaging contrast loss, strong mitigation of OoB radiation is highly demanded.
Along with a variety of different types of spectral purity filters [13][14][15][16][17][18][19][20], multilayer grating designs have been proposed earlier for suppression of OoB radiation in [6,7].As an advantage application of multilayer grating mirrors for the radiation filtering allows straightforward cooling unlike free-standing grids and foils, and these ideally could replace a standard multilayer optical element that allows to avoid introduction of additional elements in the existing optical system.These filter designs use a filtering principle based on the diffractive deflection of OoB radiation from the direction of EUV beam propagation.For instance, van den Boogaard et al. showed the potential of achieving full spectral separation of EUV and IR radiation by application of a Mo/Si multilayer coated blazed grating [6].However fabrication of such blazed multilayer grating possessing high EUV reflectance requires an application grating substrate with roughness adequate for deposition of Mo/Si coating [21], which is still a big technical challenge.Suppression of ultraviolet OoB radiation has been demonstrated recently along with high EUV reflection efficiency with a diffractive filter based on Mo/Si multilayer integrated in rectangular-groove grating [7].
In this paper we focus on the further development of a diffractive optical element to suppress the IR radiation using a rectangular grating rather than a blazed grating.Rigorous calculations of the optical response of metallic rectangular grating were used in order to determine optimal geometrical parameters of such a diffractive filter.Furthermore, we applied an alternative manufacturing procedure allowing a significant reduction of the grating period with respect to the previously reported method [7] thus improving the angular separation of "useful" and "parasitic" radiation.

Working principle
The design of the diffractive EUV filter is given schematically in Fig. 1.This hybrid structure consists of the rectangular-groove grating with an EUV-reflecting multilayer coating deposited on top.Mo/Si coating is intended to be used because of its high reflectance at the 13.5 nm wavelength.High lateral electrical conductivity of the Mo/Si coating causes efficient reflection of IR radiation from the multilayer.The lateral periodicity of the structure results in diffraction of the reflected radiation: besides the specularly reflected radiation, additional intensity is observed in the diffraction maxima at off specular angles.These diffraction angles are determined by the grating equation: ( ) where p is the grating period, λ is the wavelength of incident radiation, θ is the angle of incidence, m is an integer representing the m th diffraction order, and ϑ is the diffraction angle corresponding to the m th order.As can be calculated with Eq. ( 1) characteristic values of the diffraction angles corresponding to 13.5 nm and 10.6 μm wavelength's values differ by about three orders of magnitude.Therefore, when reflecting from the described diffraction filter the diffracted IR radiation will be significantly deflected from the direction of propagation of the reflected EUV radiation.Only the specularly reflected beam of IR radiation, corresponding to the 0th diffraction order, remains undeflected.By tuning the geometrical parameters of the grating (p, d, h) the 0th order can be suppressed for the IR light.The 0th order suppression principle can be clearly demonstrated when considering the optical response of the grating structure within the scalar theory of diffraction.Consider a 2dimensional infinite rectangular-groove grating illuminated with a normal-incident plane wave with wavelength λ.The grating is characterized by its period p, width d and depth h of the grooves.Analogous to the situation of a transmission grating, which is given in [22], one can derive the following expression for the n th order diffraction efficiency (the ratio of diffracted intensity to incident intensity): where R tot is the total reflected intensity, Γ = d/p and A is the phase shift factor For the particular case of the 0th order Eq. ( 1) results in ( ) From Eq. ( 4) it is seen that 0th order intensity becomes zero only when Γ = 0.5 and h = λ/4 + mλ/2, where m = 0,1,2,… When the 0th order reflection is suppressed, the total reflected radiation is distributed between the remaining diffraction orders: the off-specularly reflected waves.Note that for the specific case of Γ = 0.5 Eq. ( 2) results in zero intensity of even diffraction orders as well.The main part, about 80%, of the intensity of the reflected light is distributed between the 1st and the −1st diffraction orders (see Fig. 2).The actual value of the diffraction angle of these most intense diffraction orders is determined by the value of the grating period, as can be seen in Eq. ( 1).The choice of the period is discussed in the following section.Here we conclude that the smallest resonant groove depth h = λ/4 is the most convenient for filtering applications from the practical point of view.
Fig. 2. Calculated (Eq.( 2)) diffraction efficiencies of the first three diffraction orders as function of the h/λ ratio for the case Γ = 0.5.

Calculations
In this paragraph we present the results of rigorous calculations of the plane wave reflection from a metallic rectangular grating representing the above described diffraction filter.These calculations allow to accurately estimate the suppression rate of the specular reflectance in the infrared range.We also discuss illumination of the grating structure with finite size source in order to evaluate the impact of a limited spatial coherence on IR and EUV diffraction.Based on the results of these calculations we estimate the range of grating period values providing the required more than 100 × suppression of the reflectance at 10.6 μm wavelength.The calculations were performed by using the approach described in [23].The perfectconductor (PC) approximation works well for the long-wavelength infrared range for calculations of the electromagnetic wave reflection from metal gratings [24].This perfectconductor approximation also allows to consider reflection from the grating as a function of the dimensionless parameters p/λ and h/λ that is convenient for scaling procedures.All calculations were performed for both TE and TM polarization states of the incident plane wave, but the results below are presented in terms of half-sum for TE and TM reflectance, R = 0.5(R TE + R TM ), in order to simulate reflection of unpolarized light which is relevant for our applications.
First we consider the illumination of rectangular PC grating with a normal incident wave with wavelength λ.Grating parameters h and Γ are fixed at λ /4 and 0.5 respectively, while the p/λ ratio is varied in order to determine the optimal range of period values.Figure 3 shows the dependence of the 0th order reflectance (R 0 ) on the ratio p/λ.One can see periodical modulation of reflectance as a function of p/λ, which is mainly due to energy redistribution between waveguide modes in the groove cavities.However, the average reflectance gradually decreases with increasing of p/λ.When the ratio p/λ passes the value of 5.1, specular reflectance does not exceed 1%, thus meeting the requirement for the applications.However, practical applications require suppression of the reflectance not only for normal incidence, but within a finite range of angles.For the oblique incidence we calculated R 0 as a function of the angle of incidence (θ) for a set of the p/λ ratio values (see Fig. 4).From Fig. 4 one can see that specular reflectance can be suppressed more than 100 times for the angles up to 20°.where b stands for the characteristic radiation source size and z is the distance between the grating and the source.For the scattered CO 2 laser radiation in LPP sources the plasma volume can be considered as a secondary source of IR radiation with a characteristic source size b = 100-300 μm [25].Figure 5 (red lines) shows how the coherence radius at 10.6 μm wavelength, IR ρ , varies with the distance z within the length-scale relevant to the typical sizes of EUV source-collector systems [26].We have also used Eq. ( 5) for the estimation of the coherence radius values at 13.5 nm wavelength, EUV ρ , in order to evaluate the impact of the source size on the diffraction of EUV radiation (Fig. 5, blue lines).It is seen from Fig. 5 that the IR ρ values are in the centimeter range while EUV ρ is in the order of tens of micrometers.Now based on the calculations it is possible to outline the range of the geometrical parameters of the grating structure which would be suitable for the application purposes.As it was stated above the step-height h determines via the h = λ/4 requirement the resonant wavelength of the specular reflectance suppression in the long-wavelength range (IR in our particular case).The p/λ ratio governs the intensity of the specular reflectance R 0 at the resonance, i.e. the suppression of the 0th order intensity improves with the increase of the p/λ ratio.At the same time the angular deviation of the ± 1st orders of diffraction (which are the most intense) from the 0th order decreases as ( ) In practice, diffraction angles should be large enough to provide deflection of the ± 1st orders from the radiation acceptance aperture of a particular EUV optical system.At the same time the / IR p ρ ratio should be maximized to guarantee near-zero specular reflection at the given wavelength.These are the main limiting factors for the p/λ ratio.In practice, the choice of the p/λ ratio in the filter design will be a trade-off between the suppression of the specularly reflected wave intensity and the angular separation of the off-specularly diffracted waves.In our particular case of the 10.6 μm wavelength, p ~100 μm provides a satisfactory angular separation (≈5.7°) and a predicted suppression of the specular reflectance by more than two orders of magnitude for θ values up to 20° (see Fig. 4).Period values in this range are 2 orders of magnitude smaller than the IR coherence length scale (Fig. 5) thus providing efficient diffraction.Furthermore, the relatively small EUV coherence length (Fig. 5) destroys the EUV diffraction pattern preventing possible EUV diffraction losses when p ~100 μm .

Fabrication method
In this section, we describe the fabrication of the considered above EUV reflecting/IRsuppressing multilayer grating structures.We used a lift-off contact lithography process, as schematically shown in Fig. 6, to fabricate a silicon substrate with a lamellar grating design on top of which a ML was deposited.First, a silicon wafer is spincoated with a layer of UVsensitive lift-off photoresist of 3μm thick, which is slightly thicker than the aimed step height of h = λ/4.This resist is pre-baked at 95°C for 2 mins and subsequently exposed with UV light while in contact with a mask consisting of a 50 μm half-pitch grating.The period value of the resist mask, 100 μm, was chosen from the above described theoretical considerations.The substrate was then allowed to stabilize at room temperature for 10 mins.After stabilization, a reversal step consisting of a bake at 120°C for 2 mins and a UV flood exposure was performed.This reversal step gives rise to a trapezoidal mask profile, which is required for the resist strip chemicals to easily access the resist in the lift-off step [27].The resist was then developed, where the area is initially protected by the mask, i.e. not exposed to UV light in the first exposure step, were removed from the substrate.
After the preparation of the lift-off resist, a Si spacer layer was deposited (step 3 of Fig. 4).Magnetron sputtering deposition was employed, enabling layer growth with subnanometer roughness levels over the thickness of the spacer layer.This is a critical requirement in order to obtain high EUV reflectance from the Mo/Si multilayer deposited onto the spacer layer [28].The experimental geometry and working pressure (≈10 −3 mbar) resulted in local angle of incidence variations of the deposited atoms at the sample up to 10° from normal.In order to obtain a sharp projection of the mask structure into the deposited layer under these conditions, i.e. to prevent half-shadow regions, the mask thickness should be much smaller than the mask apertures.In the presented experiments the thickness/aperture ratio was 3/50.
After silicon deposition, the actual lift-off is performed by stripping the remaining resist from the substrate using ultrasonic baths of acetone and IPA.The initial silicon substrate and the deposited silicon structures thus form a grating pattern defined by the mask.Finally, on top of this silicon grating structure, a high reflecting multilayer was deposited using electron beam evaporation in combination with ion beam smoothing [21].The fabricated PsR structures were characterized with atomic force microscopy.The average step height of the grating was measured to be 2.35 ± 0.05 μm corresponding to 9.4 μm wavelength of the zero-order suppression spectral minimum.In spite of the mismatch with the targeted 10.6 μm wavelength this test multilayer grating structure can be used for the validation of the filtering principle.

Spectral characterization of samples
EUV reflectometry of the samples was performed at the EUV beam line of the Physikalisch-Technische Bundesanstalt (PTB) in Berlin.The measurements were performed at the angle of 1.5° from normal incidence and with a detector aperture of 0.5°.
Infrared reflection spectra of the sample were recorded with a Bruker Vertex 70v Fouriertransform infrared (FTIR) spectrometer.The measurements were performed in the 400 -2000 cm −1 spectral range.A pyroelectric detector was used in order to provide a broad dynamic range (four orders of magnitude) in the measurements.The spectra were recorded with a resolution of 4 cm −1 and in each case 64 spectra were averaged.An aluminium mirror was used as the reference sample for the reflectivity measurement.
Accurate measurements of the specular reflectance (0th order diffraction efficiency) from the grating structure are complicated due to relatively small diffraction angle (θ ≈5.7°) at the wavelength of interest.Typical numerical apertures of mirrors in commercially available reflection accessories for the used spectrometer exceeds this diffraction angle value resulting in collection of undesired off-specularly reflected radiation.Note that in our measurements we separated the off-specular from specular signal by decreasing the aperture of the mirrors with slit diaphragms oriented perpendicularly to the plane of diffraction.

Results and discussions
The results of the EUV reflectance measurements are shown in Fig. 7.The peak reflectance of a reference Mo/Si multilayer on a polished substrate, co-deposited during multilayer deposition on the Si grating substrate, is 69%.The reflectance of a Mo/Si multilayer deposited on the spacer layer outside the masked area, is 68%, the 1% reduction of the reflectance probably caused by roughness of the spacer layer.However, 61% has been measured on the final multilayer grating samples, representing a loss of 7%.The influence of structural imperfections of the multilayer are considered the major cause of the reflectance loss.Edge rounding effects lead to non perfect periodicity of the multilayer near the edges of the grating facets.As it was studied with transmission electron microscopy in detail in our previous work, the width of this distorted area scales with the thickness/aperture ratio of the deposition contact mask, and with the angular anisotropy of the deposition flux [29].Therefore the EUV reflectance loss could be minimized by employing lower mask thickness/aperture ratios, but in practice this requires increased pitch lengths [7].Alternatively, further optimization of the deposition anisotropy by collimated deposition techniques is considered promising [30].The measured minimum reflectance value of R min = 1.4% exceeds our theoretical expectations of R 0 = 0.3% for the p/λ ≈10 ratio (Fig. 3), which limits the suppression rate of the 0th order intensity.We suggest that the discrepancy is caused by our measurement method.The calculations are made for the case of the infinite-size grating illuminated with a monochrome plane wave, which has infinite spatial coherence length by definition.Illumination of the grating structure in the FTIR spectrometer with the finite spatial coherency radiation limits the number of periods of the grating that interfere destructively and, consequently, results in the reduction of the interference contrast.For our measurement scheme this has led to the increase of R min .The application of diaphragms for angular filtering required in our specific case also results in the reduction of the illumination spot size and consequently reduction of the number of the grating periods involved in the formation of the diffraction pattern.In the case of illumination of the described diffractive filter with the scattered CO 2 laser radiation in EUV sources we expect the suppression rate of the 0th order reflectance to be closer to our theoretical estimations due to the larger coherence length.If stronger suppression is needed, a possible way to meet this requirement is to increase the period of the structure.

Conclusions
We have developed a special combination of a high-reflectance multilayer mirror for EUV radiation and a lamellar grating substrate that provides spectrally selective suppression of longer-wavelength radiation.This suppression of the long-wavelength specular reflectance occurs due to quarter-wavelength phase-shift resonance.It is not affecting the high reflectance of the multilayer structure for EUV light.This system allows for example the suppression of unwanted radiation from laser plasma-based EUV and soft X-ray sources, having an undesired component of scattered IR laser radiation.Such spectral filtering is of high importance for emerging lithographic applications, high harmonic generation systems and water window microscopy.
A test mirror has been fabricated using a contact mask deposition of the grating substrate followed by the deposition of a Mo/Si multilayer designed for normal incidence reflectance at λ = 13.5 nm.The grating has been optimized for mid-IR suppression.A factor of 70 reduction has been experimentally demonstrated at 9.7 μm wavelength together with 61% EUV peak reflectance at 13.5 nm wavelength at near normal incidence.

Fig. 1 .
Fig. 1.Schematic design of EUV filtering system based on the rectangular multilayer grating.Colors: grey -grating substrate; blue/yellow -Mo/Si multilayer deposited on top; red arrows -IR radiation; violet arrows -EUV radiation.

Fig. 3 .
Fig. 3.The calculated 0th order reflectance (R0) from a perfectly conducting rectangular grating as a function of p/λ for the case of h = λ/4 and Γ = 0.5.

Fig. 4 .
Fig.4.The calculated 0th order reflectance (R 0 ) from a perfectly conducting rectangular grating as a function of θ.Grating parameters h and Γ are fixed at λ/4 and 0.5 respectively, p/λ ratio is varied.Note that the above given calculated results correspond to the case of an infinite grating with plane wave illumination.While in practice limited spatial and temporal coherence of the radiation source will affect diffraction from the considered grating structure.Obviously, a finite coherence size of the illumination spot on the grating will abridge the number of orders involved in the diffraction pattern.The resulting redistribution of the reflected intensity between the limited number of diffraction orders will determine the suppression efficiency of the IR specular reflectance.The characteristic coherence length in the illumination spot can be estimated via the relation for the radius of spatial coherence

Fig. 5 .
Fig. 5. Calculated dependencies of the coherence radius on the distance z between grating and source.IR ρ -coherence radius for 10.6 μm wavelength, EUV ρ -for 13.5 nm wavelength.

Fig. 6 .
Fig. 6.Step by step process description of the lift-off UV contact lithography procedure.

Fig. 7 .
Fig. 7. Blue squares -the measured EUV reflectance of the Mo/Si multilayer deposited onto the fabricated Si grating.Grey circles -the measured EUV reflectance of the reference Mo/Si multilayer on a super-polished Si substrate, co-deposited during the multilayer deposition onto the Si grating.

Figure 8
Figure8shows reflectance spectra of the sample measured with the FT-IR spectrometer.According to a polynomial fit of the measured data the spectral position of the reflectance minimum corresponds to 1030 cm −1 wavenumber (9.7 μm wavelength).The absolute value of the zero-order reflectance at the minimum is 1.4% corresponding to 71 times suppression of the IR light.The 9.7 μm wavelength value allows to estimate the grating groove depth as 2.43 μm via the h = λ/4 criterion, which is close to the AFM measured h value of 2.35 ± 0.05 μm.

Fig. 8 .
Fig.8.FT-IR reflectance measurements for the multilayer grating fabricated using a line pattern with a pitch of 100 μm.