Open Access
Add to BibSonomy
Abstract
Imaging at H Ly-α (121.6 nm), among other spectral lines in the short far UV (FUV), is of high interest for astrophysics, solar, and atmosphere physics, since this spectral line is ubiquitously present in space observations. However, the lack of efficient narrowband coatings has mostly prevented such observations. Present and future space observatories like GLIDE and the IR/O/UV NASA concept, among other applications, can benefit from the development of efficient narrowband coatings at Ly-α. The current state of the art of narrowband FUV coatings lacks performance and stability for coatings that peak at wavelengths shorter than ∼135 nm. We report highly reflective AlF3/LaF3 narrowband mirrors at Ly-α prepared by thermal evaporation, with, to our knowledge, the highest reflectance (over 80%) of a narrowband multilayer at such a short wavelength obtained so far. We also report a remarkable reflectance after several months of storage in different environments, including relative humidity levels above 50%. For astrophysics targets in which Ly-α may mask a close spectral line, such as in the search for biomarkers, we present the first coating in the short FUV for imaging at the OI doublet (130.4 and 135.6 nm), with the additional requirement of rejecting the intense Ly-α, which might mask the OI observations. Additionally, we present coatings with the symmetric design, aimed to observe at Ly-α, and reject the strong OI geocoronal emission, that could be of interest for atmosphere observations.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The development of narrowband coatings tuned in the far ultraviolet (FUV, 100<λ<200 nm) is demanded for the future generation of space instruments, since key spectral lines of basic constituents for atmospheric physics, solar physics and astrophysics lie in this range [1,2]. The available FUV bandpass optical elements with high out-of-band rejection present a moderate throughput; therefore, this enabling technology is currently included in the NASA tier 2 strategic technology gaps [3]. This technology is also useful for many other fields, which include petawatt-laser beamlines [4], excimer laser optics [5], with, for example, Ar2 laser operation at 126 nm [6], and Kr2 at 146 nm [7], high-order harmonics, as the seventh harmonic of 800 nm emission of Ti:sapphire lasers is placed at ∼114 nm [8], thermonuclear fusion reactors [9], or the semiconductor industry, including wafer inspection in the sub-200 nm range [10].
Particular missions that might have benefited from the availability of narrowband coatings is the selected Solar Terrestrial Probes Missions of Opportunity by NASA: GLIDE (Global Lyman-alpha Imagers of the Dynamic Exosphere), recently renamed Carruthers Geocorona Observatory. This mission will make unprecedented measurements of the FUV light emitted by hydrogen atoms in the Earth’s exosphere. Measurements will be made outside the exosphere, where only a handful of comparable ultraviolet light images have previously been made. These measurements will serve as a tracer of exospheric density and spatial structure, an interesting advance for the understanding of upper atmospheric physics [11]. In addition to the Ly-α line, of interest for these and other missions, spectral lines such as NV at 123.8 and 124.2 nm and CIV at 154.8 and 155.0 nm are interesting to trace gases at temperatures of a few hundred thousand Kelvins, in between the temperature regimes covered by optical, and X-ray observations. The OI doublet (130.4 and 135.6 nm) is also relevant to a wide range of astrophysics targets, which include exoplanet search [12]. Advances in FUV narrowband mirrors would enable high-throughput observations of mission-specific bandpasses while limiting background contamination.
There is currently a wave of FUV astrophysics proposals. As a relevant example, the 2020 Astronomy and Astrophysics Decadal Survey [13] recommended an Infrared (IR)/Optical (O)/Ultraviolet (UV) Large Telescope optimized for observing habitable exoplanets and general astrophysics as the highest priority future space observatory. One of the instruments of the former, LUMOS (LUVOIR Ultraviolet Multi-Object Spectrograph) will have three channels: a FUV multi-object spectrograph (MOS), a NUV/VIS MOS, and a FUV imager covering the 100-200 nm range with narrow- and medium-band filters based on the technology presented in this work [14,15].
FUV narrowband coatings typically consist of multilayers (MLs) formed by several bilayers of two materials with low absorption and contrasting refractive indices [16,17]. The two materials are usually referred to as H (high refractive index material), and L (low index material), and the periodic multilayer design is usually presented as (H/L)m/substrate.
The high absorption of materials, a critical feature in the shorter part of the FUV, and the uncertainty on their optical constants in this part of the spectrum limit the efficiency of FUV optics, which, in turn, have a negative impact on the imaging performance of FUV optical systems. The materials that keep their transparency deeper in the FUV are some metal fluorides. Lithium fluoride (LiF), aluminum fluoride (AlF3), and magnesium fluoride (MgF2) present the shortest cutoff wavelengths, at ∼102 nm, ∼110 nm, and ∼113 nm, respectively. LiF has not been usually considered as a FUV ML material because of its hygroscopic nature, and the limited knowledge of its optical constants, even though it has been exceptionally used as a broadband mirror protection [18,19], and also in narrowband FUV MLs [20]. MgF2 and AlF3 have been typically used in FUV MLs as low-refractive index materials, along with LaF3 as the high-index material, which is transparent down to a relatively short FUV wavelength (∼120 nm) and has a good contrasting refractive index with the aforementioned materials. The most widely reported combination of fluorides is MgF2/LaF3 [21,–30]. However, most of the presented MLs are tuned at long FUV wavelengths. This may be due to the increased absorption even of the above fluorides with reducing FUV wavelengths, which limits the efficiency of coatings in the short FUV. The technical difficulties are especially critical in the transition range between the FUV and EUV, sometimes defined by the astrophysics community as the Lyman Ultraviolet (LUV: 91.2-121.6 nm) because it includes the important Lyman spectral series of hydrogen, starting with the ubiquitous and intense H Ly-α line at 121.6 nm [2]. Figure 1(a) shows the optical constants, (n,k), of MgF2 and LaF3 in the FUV and LUV [31], and Fig. 1(b) shows the predicted reflectance of ideal (MgF2/LaF3)15 MLs in the same range.
Fig. 1. (a) Optical constants (n,k) of MgF2 and LaF3 in the FUV and LUV. (b) Predicted reflectance of (MgF2/LaF3)15 MLs in the same range.
Regarding all-dielectric filters, there are few reported (MgF2/LaF3)m MLs peaked at wavelengths shorter than ∼157 nm, which are at ≈135 nm and ≈130 nm [21,22,28,32] and ≈121.6 nm [28,33,34]. The latter with a maximum peak reflectance of ≈65% set by Rodríguez-de Marcos [28]. The other typical combination of fluorides, AlF3/LaF3 MLs, has been explored mostly at long FUV wavelengths (∼193 nm) [24,35–39].
There are few narrowband coatings (not necessarily all-dielectric) have been reported in the literature for the LUV range, which consist of MgF2/LaF3 MLs tuned in 121.6 nm [28,33,34], a narrowband coating at H Ly-β (102.6 nm) based on Al, LiF, and SiC [20], an Acton Research’s filter based on Al/MgF2/Os [40], as well as a narrowband coating at 121.6 nm of unknown design and materials [41], with the maximum reflectance set by [28] at ≈65% with MgF2/LaF3, including commercial filters. Importantly, the environmental stability of the narrowband coatings has been scarcely documented [28,30,32,36,37,42,44]. Among the few works reporting multilayer stability, the majority of the multilayers significantly degraded over time.
The objective of this research is to improve the results of fluoride narrowband coatings tuned at Ly-α (121.6 nm) in view of the limited performance of the coatings in the shorter part of the FUV. For that reason, we will explore both multilayer types based on AlF3 or MgF2 with LaF3 tuned at 121.6 nm. Our results include coatings with a single requirement, to be peaked at a desired wavelength, and coatings with the additional requirement of rejecting an undesired secondary wavelength. We present narrowband multilayers peaked at H Ly-α in regard to the scientific interest of this spectral line, which is ubiquitous in the sky and is the strongest line in the solar spectrum. We also developed coatings for the OI doublet (130.4 and 135.6 nm) for exoplanet search, among other goals, which need the additional requirement of rejecting the close Ly-α since the latter’s strong intensity could mask the OI observations [43]. The complementary design, consisting of high reflectance at Ly-α and strong rejection at the OI doublet, which may be useful for Ly-α observations with imaging instruments like GLIDE that need to avoid the OI geocoronal emission, was also investigated. The environmental stability of some multilayers will be discussed.
The paper is organized as follows. Section 2 describes the experimental techniques used. Section 3 presents the reflectance of coatings centered at Ly-α or the OI doublet, based on periodic or aperiodic combinations of MgF2/LaF3 and AlF3/LaF3 MLs. Long-term coating stability in different environments is also investigated.
2. Experimental
Fluoride MLs with LaF3 and AlF3 or MgF2 were deposited in a high-vacuum chamber by thermal evaporation using W boats onto heated substrates. Fluorides with highest FUV transparency are deposited by vacuum evaporation. This transparency is enhanced when the coating is deposited on a heated substrate [48]. The latter is related with a reduced porosity, and hence with an increase of the packing density. Chamber pumping system, pressure, evaporated materials purity, substrate cleaning and heating and cooling after deposition are described in [30,44]. Deposition rates were ∼1.2 nm/s for AlF3, ∼0.6 nm/s for LaF3, and ∼0.7 nm/s for MgF2. Layer thicknesses were monitored with a quartz microbalance and an automatic controller that drives a shutter. One of the presented samples was deposited in a different chamber, a 75-cm diameter, 100-cm height cylindrical stainless steel placed in an ISO6 clean room. This chamber is pumped with a Velco 250A cryo system. Fore vacuum is made with a diaphragm pump and a turbomolecular pump. Before deposition, the latter chamber is heated to ∼180°C for ∼30 minutes to accelerate surface outgassing.
FUV reflectance was measured in GOLD’s reflectometer (GOLD is the Spanish acronym for Thin Films Optics Group, Madrid, Spain [45]), which operates in ultra-high vacuum (UHV) conditions. Details on the reflectometer are explained in previous works [30,44].
Hot-deposited fluoride MLs usually present high tensile stress, from which thermal stress is expected to be predominant when fluorides are deposited at elevated temperatures on low-expansion substrates, such as fused silica (FS) [46]. This stress is caused by the difference in CTE (Coefficient of Thermal Expansion) between the substrate and the film. It also depends on the difference between the deposition temperature and the application temperature [38,47]. One successful way to reduce stress in fluoride MLs is the use of a higher CTE substrate, such as CaF2 [44]. Nevertheless, in this work we used FS and float glass substrates, that, even though they are not the most suitable substrates to reduce stress, they conform as space-instrumentation substrate standards due to their mechanical, thermal and radiation resistance and their ease to polish and shape. Total stress on a coating can generate cracks and/or delamination of the surface, which is highly undesirable; these destructive effects are correlated with the Force Per Unit Width (FPUW), which, in its simplest form, can be calculated as stress times thickness. A previous investigation led to determining the total maximum thickness before crack generation of MgF2/LaF3 and AlF3/LaF3 MLs deposited on different substrates [30,44]. This thickness threshold for MLs deposited on FS at ∼250°C, the average substrate deposition temperature in this research, is around ∼650 nm and ∼1000 nm for MgF2/LaF3 and AlF3/LaF3 MLs, respectively [30,44]. In this work, we present samples with total coating thickness below the aforementioned thresholds, since they are centered at shorter wavelengths, which results in that each layer is thinner in proportion to the target wavelength. Surfaces were examined with optical microscopy in dark field mode to verify that there exist no cracks. The coating total thickness was measured with profilometry and ellipsometry.
Structural and compositional analysis of the multilayers systems was carried out by Raman spectroscopy using a confocal Raman microscopy instrument (Witec ALPHA 300RA) with a Nd:YAG laser excitation at 532 nm. Confocal Raman Microscopy is an improved model of Raman spectroscopy, because it gives information on the depth and structures in thin films, due to the possibility to filter in the lateral and vertical axis, to resolve up to 1 µm. The piezoelectric scanning table allows three-dimensional displacements in steps of 3 nm. The optical resolution of this confocal microscope is, approximately, 250 and 700 nm in the longitudinal and transversal directions, respectively. Raman spectra were recorded in the spectral range of 0–1175 cm−1 by using a grating of 1.800 lines per millimeter resulting in a spectral resolution of the system below 0.02 cm−1. The samples were mounted in a piezo-driven scan platform having a positioning accuracy of 4 nm (lateral) and 0.5 nm (vertical). The reflected laser light at each point was collected using a multimode optical fiber of 25 μm in diameter by a very sensitive charge-coupled device detector. Raman images of samples were measured with a size of 10 μm × 10 μm for surface scanning, and of 4 μm × 3 μm for in-depth scanning, and each image contained 10000 Raman spectra. Raman was measured at room temperature using an objective of 100× with a numerical aperture of 0.95. The acquired spectra were analyzed by using Witec Project 2.02 program and WitecControl Plus Software.
3. Results
Depending on the requirements, two strategies have been followed in the design of ML coatings. The first strategy consists of designing narrowband coatings with no further constraint; the second strategy adds the requirement to minimize ML reflectance at a secondary wavelength. The two strategies are presented in subsections 3.1 and 3.2, respectively. All samples were deposited on hot substrates in order to increase the transparency of the metal fluoride layers [48]. The temperature range was ∼230-270°C, i.e., close to the optimal deposition temperature of 250°C that was found for MgF2/LaF3 MLs [30]. A similar temperature was also reported as the fluoride optimal deposition temperature for broadband Al/MgF2 [49] and Al/AlF3 mirrors [50] to obtain a fluoride density close to the bulk density, and was successfully applied on AlF3/LaF3 MLs [44].
ML design involves the reflectance calculation of various interfaces. Let us summarize the basics of such calculation. We assume a plane wave that impinges on a flat boundary from media N2 to N1, which in general may absorb light. The electric field of the incident, reflected, and refracted wave can be decomposed into two components: the p component in the incidence plane and the s component perpendicular to that plane. The electric field in the reflected wave, normalized to the incident wave, is given by Fresnel coefficients [51]:
Such ratio of the electric field will be referred to as amplitude reflectance, to distinguish it from intensity reflectance. When at least one material at the interface absorbs radiation, Fresnel coefficients are complex numbers, which means that reflection produces both an amplitude change but also a phase shift, and the two are given by the modulus and phase, respectively, of Fresnel coefficients.
When we go from one interface to the various interfaces of a multilayer, the accumulated amplitude reflectance due to interferences among radiation reflected at all interfaces is calculated iteratively, starting with the innermost interface [52]. Hence, if the accumulated reflectance from the innermost interface (numbered as first) up to the i-th interface is given by ρi, the accumulated reflectance up to the i + 1-th interface is given by:
A ML consists of a stack of several layers with optimized thicknesses to maximize the optical throughput, in the present case, the normal-incidence reflectance at a given wavelength λ. Each interface provides a relatively small contribution to the full reflectance, but the combination of all interface contributions is aimed to provide a larger reflectance. For slightly absorbing layer materials, the design that maximizes reflectance is the quarterwave (QW) design, in which the optical path through each layer thickness equals λ/4.
The above calculation assumes smooth interfaces. The presence of interface roughness results in some specular reflectance reduction at each interface and the generation of scattered light. Such loss, assuming it is small, can be evaluated with Stearns formalism that approximates this loss by multiplying the Fresnel reflection coefficients with a correction function that is the Fourier transform of the derivative of the interface profile function [53,54]. In this research, we assumed that the interface profile can be described with an error function, so that the correction function is a Gaussian function, whose width parameter is the RMS interface roughness [53].
3.1 Narrowband multilayers
A QW periodic ML design of two fluorides of high (H) and low refractive index (L), (H/L)m, where m is the number of bilayers, was chosen to make narrowband reflectors nominally tuned in the FUV at 121.6 nm. Ten ML coatings were prepared, some of them with a variation of the number of layers, to determine the optimal number of bilayers that contribute to reflectance. The different deposition parameters for each coating are presented in Table 1.
Table 1. Design and deposition parameters.
Figure 2(a) compares the reflectance measured for (MgF2/LaF3)m and (AlF3/LaF3)m MLs ranging from 10 to 19 bilayers. Measurements were made after a few hours of contact with the atmosphere. We can observe that the peak reflectance is about 10% larger for all of the (AlF3/LaF3)m samples compared to the (MgF2/LaF3)m ones. This can be explained by the slightly shorter cutoff of AlF3 and a higher index contrast expected for AlF3 with LaF3 compared to MgF2, that also leads to some bandwidth widening.
Fig. 2. (a) Near normal reflectance as a function of wavelength of multilayers with m = 10, 13, 16, and 19 bilayers (AlF3/LaF3), and m = 10, 13, 15, 19 bilayers (MgF2/LaF3) (samples LY_A_10,13,16,19 and LY_M_10,13,15,19 in Table 1, respectively), each set deposited on a common FS substrate during the same vacuum cycle. (b) Reflectance at Ly- α vs the number of bilayers of the ML, m. Measurements were performed after a few hours of contact with the atmosphere.
Figure 2(b) shows the correlation between the reflectance at Ly-α and the number of bilayers. Both (AlF3/LaF3)m and (MgF2/LaF3)m samples show similar behavior with the number of bilayers: the reflectance slightly increases from 10 to ∼15 bilayers, and then slightly decreases when the number of bilayers increases to 19. Effective reflectance of a ML can be understood as a balance between the contribution of additional bilayers and the trend of surface and interface roughness to grow with total thickness [38,55,56], which increases scattering losses and/or reduces the contribution of each additional interface through transmittance enhancement. Thus, 15 bilayers seem to be a good trade-off to obtain high reflectance. Fresh reflectance at Ly-α of ≈86% and ≈76% was obtained for the (AlF3/LaF3)16 and (MgF2/LaF3)15 MLs, respectively, which is considerably larger than the maximum reported reflectance for a narrowband mirror at Ly-α of about ≈ 65% [28].
ML RMS (root mean square) roughness was measured for several samples with Scanning Electron Microscopy to obtain the roughness evolution with the number of bilayers or total thickness, and these results are reported elsewhere [57]. That research reports that roughness is slightly higher for MgF2- than for AlF3-based MLs, and it also reports that roughness increases with the number of bilayers and with total thickness, in agreement with [38,55,56], with a sharper increase for MgF2-based MLs. Such evolution of roughness explains that ML reflectance vs. number of bilayers has a maximum in Fig. 2 and it decreases after that, and it may also explain part of the larger efficiency of the AlF3- vs. MgF2-based MLs.
We calculated the peak reflectance for MLs tuned at 121.6 nm with m bilayers, m ranging from 3 to 25. The calculation took into account the increasing roughness reported in [57] and used the optical constants in [31] and [58]. Results of the increasing roughness and the predicted peak reflectance with the bilayer number, m, for (MgF2/LaF3)m and (AlF3/LaF3)m MLs can be found in Fig. 3. The trend of the peak reflectance with the number of bilayers is similar for both the calculations and the measurements presented in Fig. 2(b).
Fig. 3. Calculated peak reflectance and RMS roughness as a function of the bilayer number, m, of (AlF3/LaF3)m and (MgF2/LaF3)m MLs. Calculations took into account the increasing roughness with m. RMS roughness was assumed as per measurements in [57].
The highest reflectance of AlF3-based compared to MgF2-based MLs can be explained in part by the AlF3 shortest cutoff compared with MgF2 and by the higher AlF3-LaF3 refractive index contrast compared with MgF2-LaF3, and, in part to the lower RMS roughness of AlF3 compared to MgF2 [38].
Reflectance at longer wavelengths was also measured for some samples. Figure 4 shows the near-normal reflectance of the (AlF3/LaF3)16 ML, from the FUV through the visible and the near Infrared (NIR). In the visible and NIR, reflectance stays under 10%. Measured reflectance includes the contribution from the back-substrate surface. Often such contribution, which is undesired, can be eliminated, such as through non-polished, frosted or black back face, the use of antireflection coatings or using wedged substrates. To evaluate the reflectance without such contribution, it was calculated using the substrate optical constants [31] and subtracted from the measured reflectance. The result is also plotted in Fig. 4.
Fig. 4. Near normal reflectance as a function of wavelength extended to the visible and the near IR of AlF3/LaF3 MLs, corresponding to samples LY_A_16 (a), and LY_A_X (b) in Table 1. The x-axis is scaled linearly up to 180 nm, and then logarithmically up to 2500 nm. The dashed line represents the calculated reflectance if the substrate back face did not contribute to reflection. Insets correspond to an image of the LY_A_X sample.
3.2 Narrowband multilayers with the extra requirement of rejection at a secondary spectral line
The development of narrowband coatings sometimes involves paying special attention to rejecting a specific spectral line or band that, being more intense than the target spectral line, it may mask the information in the passing band. Reflecting one spectral line and suppressing another one turns more difficult when the two have close wavelengths. This is the case for the O I doublet lines (130.4 and 135.6 nm), which are close to Ly-α, and for sapecific observations at one spectral line, either of them could mask the other. Due to the ubiquity and intensity of the Ly-α line, some specific applications, through imaging oxygen traces [43], might need to enhance the throughput at the O I doublet while rejecting Ly-α. Conversely, instruments aimed to observe in the atmosphere, such as GLIDE, may need high throughput at Ly-α while having a strong rejection at the O I doublet.
The above target of separating Ly-α and the O I doublet was addressed with ML coatings with high reflectance at one spectral line or doublet and a low reflectance at the other. To prepare coatings with the aforementioned extra requirement, it was found necessary to introduce some aperiodicity in the ML design, mostly to reduce some of the lateral lobes in the reflectance spectrum. An approach for minimizing the lateral lobes can be found in [59]. Hence the design [(0.5 H)L(0.5 H)]m can reduce the lateral lobes for wavelengths shorter than the central band, and the alternate design [(0.5 L)H(0.5 L)]m can minimize the lateral lobes at wavelengths longer than the central band. Both designs result in some reflectance sacrifice at the peak. We observed that designs with an odd number of 0.5 H or 0.5 L thicknesses, such as 1.5 L(H/L)m or 0.5 L(H/L)m would also work to decrease the lateral lobes in one or the other side of the band. Examples of calculations with this approach for Ly-α are presented in Fig. 5.
Fig. 5. Calculation of the reflectance as a function of wavelength of an (AlF3/LaF3)15 ML tuned at 121.6 nm for the QW design [plotted as (H/L)15] and variations of the QW design that reduce the lateral lobes on one or the other side.
To obtain coatings that separate Ly-α from the O I doublet, several approaches were attempted with AlF3/LaF3) and (MgF2/LaF3) MLs. The design to reflect Ly-α and reject the OI doublet consisted of a 1.5 L(HL)15 with H = LaF3 and L = AlF3; this design may be advantageous to avoid very thin layers (0.5 H at Ly-α would be around 6 nm thick). The coating aimed to reflect the OI doublet consisted of an 8-bilayer MgF2/LaF3 periodic ML with two bottom and one top free-thickness layers that were optimized with Binda Genetic Algorithm of XOP-IMD software [54,60]. Samples of these two designs are plotted in Fig. 6. The central wavelength, averaged reflectance at the peak, bandwidth in FWHM and rate between the rejection and the peak of the samples presented in sections 3.1 and 3.2 are presented in Table 2.
Fig. 6. Experimental near-normal reflectance as a function of wavelength of aperiodic (AlF3/LaF3) and (MgF2/LaF3) MLs, (samples LY_A_X and OI_M_x in Table 1), tuned at 121.6 nm and at the O I doublet, respectively.
Table 2. Central wavelength λ0, averaged reflectance at the peak $\bar{{\boldsymbol R}}$(λ0), bandwidth in FWHM, and rate between the desired rejection and the peak.
Figure 6 indicates that the two goals can be achieved: a reflectance above 70% was obtained at the target wavelength while a low reflectance was obtained at the single or double wavelength to be rejected by placing the minima between ML side lobes close to the lines to be rejected. Reflection/suppression ratios even up to ∼100 were obtained.
Figure 4 also plots the reflectance at longer wavelengths measured for the AlF3/LaF3 ML reflecting Ly-α and rejecting the OI doublet (LY_A_X sample in Table 1). We again add the calculated reflectance without the contribution of the back of the substrate.
When larger rejections are necessary, multiple reflections on a series of mirrors can provide it, while keeping a significant throughput at the target wavelength. The measured reflectance elevated to the number of reflections for the ML peaking at 121.6 nm and rejecting the O I doublet is plotted in Fig. 7. The figure also includes the calculated reflectance without the contribution of the back of the substrate, which is relevant at long wavelengths. With just two reflections, a reflection-suppression ratio of ∼6800 between Ly-α and 130.4 nm and ∼230 between Ly-α and 135.6 nm (the two O I doublet lines), respectively, is obtained. Between Ly-α and the visible, a reflection-suppression ratio of ∼180 and ∼700 for a substrate with and without the back contribution, respectively, can be achieved with these two reflections. In a two- or multiple reflection system, the imperfect rejection due to the lateral lobes can be minimized by slightly shifting the lobes between the two or more mirrors. Overall, a two-system optics could give a remarkable freedom to obtain a desired optical profile.
Fig. 7. Near normal reflectance as a function of wavelength of an (AlF3/LaF3)m ML (LY_A_x, Table 1) elevated to the number of reflections, both in the FUV (a) and the FUV-vis-NIR (b). In (b), the x-axis is scaled linearly up to 180 nm, and then logarithmically up to 2500 nm. The dashed line represents the calculated reflectance if the substrate back face did not contribute to reflection.
3.3 Narrowband multilayer aging
Some samples presented in section 3.2 were also measured after several months of storage in a desiccator. The majority of the samples presented a decay in reflectance of about ≈3-5% after ≈5-6 months of storage. To better understand the nature of the decay, which is expected to be related to the adsorption of contaminants and/or oxidation of the materials, an (AlF3/LaF3)15 ML was simultaneously deposited on two cuts of the same float glass substrate during the same vacuum cycle. One of the pieces (LY_A_15_a) was stored in vacuum, and the other piece (LY_A_15_b) was stored in a desiccator with silica gel. The hygrometer displayed a RH (Relative Humidity) below 20% in the desiccator at all times. Reflectance was measured after less than ∼1 hour in contact with the atmosphere for both samples, and then after two months of storage in vacuum or in the desiccator. The results, shown in Fig. 8(a), present less than a ∼1% reflectance decay at Ly-α for the sample stored in vacuum, while the sample stored in the desiccator, decayed a 1% more. In spite of these decays, both samples kept high performance after two months. The moderate decay could be due to contaminants, mostly water, that can be adsorbed on the coating pores over time. Yet, the small reflectance difference of the two aged samples implies that storage in a desiccator keeps reasonable performance over time with the benefit of its simplicity.
Fig. 8. (a) Near normal reflectance as a function of wavelength of fresh and aged AlF3/LaF3 MLs, corresponding to samples LY_A_15_a and LY_A_15_b in Table 1. Samples LY_A_15_a and LY_A_15_b were measured after two months of storage in vacuum, and in a desiccator with RH below 20%, respectively. (b) Measurements of the near normal reflectance of sample LY_A_15_a after 15 days in environments of 50-60% and 60-70% RH.
To further study the degradation of the samples caused by humidity, after the two-month storage in vacuum, sample LY_A_15_a was later stored in a container with relative humidity (RH) of 50-60% for 15 days, and then remeasured. We repeated the process a second time by placing the same sample in a RH of 60-70% for 15 days, and measuring it again. Reflectance after the humidity cycles is presented in Fig. 8(b). We observed a further reflectance decay at Ly-α of ≈1% after the storage at 50-60% RH, and then an additional ≈1% after the storage at 60-70% RH. We can conclude that reflectance decay in humid environments is faster than in a desiccator, and vacuum provides a more stable environment for long-term storage. However, there are no significant reflectance losses in humid environments up to 70% RH in short periods, being the accumulated losses always below 5%. This relatively low reflectance decay can be attributed to the limited porosity of the fluorides [48], particularly the outermost LaF3 layer when deposited on a hot substrate, which reduces the adsorption of water.
In order to obtain a better understanding of the adsorption of contaminants in fluoride MLs, structural and compositional analysis of the (AlF3/LaF3) MLs was carried out by Raman spectroscopy using a Confocal Raman Microscope with a Nd: YAG laser light source (532 nm). Raman spectroscopy was carried out in the presented samples stored in different environments, to clarify the effect of humidity both on the ML surface and in-depth. For this purpose, the samples stored in vacuum, at 60% of RH, and a sample stored at 99% of RH, were analyzed. Figure 9(a) shows the average Raman spectra of these samples measured on the surface. For AlF3/LaF3 MLs, only the LaF3 can be observed by Raman spectroscopy, since the Raman signal corresponding to AlF3 has a lower intensity and overlaps with that of LaF3, being masked by the latter. LaF3 possesses 17 Raman modes (5A1g + 12Eg), although only a few of them are revealed for nonpolarized incident light [61]. Figure 9(a) shows 4 Eg symmetry modes at 199 cm-1, 297 cm-1, 311 cm-1, and 365 cm-1, and only one A1g symmetry mode at 390 cm-1. No Raman shift is observed for the samples stored in a humid environment compared with the sample stored in vacuum; this means that adsorbed water does not distort the molecular bonding, nor the crystalline structure of the layers. Water content can be observed at higher wavenumbers, through the O-H bonds, in the range of 1800-3600 cm-1, especially above 3200 cm-1.
Fig. 9. (a) Raman spectra in 100-4000 cm-1 of samples stored in vacuum, and with 60% and 99% RH. Raman images for scanning in surface [images (b), (c), and (d)] and in-depth profile [images (e), (f), and (g)] for samples stored (b),(e) in vacuum, (c),(f) with 60% RH and (d),(g) with 99% RH, where the phase distribution can be observed. The blue color indicates O-H bonds corresponding to water, and the red color corresponds to La from LaF3 layers.
For the sample stored in vacuum, no significant contribution of water is observed, only a negligible band around 3400-3700 cm-1, corresponding to O-H bonds, that can be attributed to the short contact with room humidity. For the sample stored in 60% RH, a wide band is registered at ∼3614 cm-1 corresponding to free O-H bonds, i.e. free water is adsorbed on the surface but with no bonding to it [62]. As for the sample stored in 99% RH, two clear O-H wide bands are observed at 3234 cm-1 and 3410 cm-1, corresponding to the symmetric and antisymmetric O-H stretching modes, respectively. These bands are well-defined and correspond to the binding with the surface of the sample. Moreover, a less intense band appears at 1612 cm-1 corresponding to O-H bending mode. Aside from water, a band around 2616 cm-1 corresponding to C-H stretching mode is observed, which may derive from hydrocarbons typically present in laboratory atmosphere.
Figures 9(b)–(g) show the Raman images obtained from scanning in surface and in-depth profile for samples stored in vacuum, in 60% and 99% RH, respectively. In these Raman images, a phase distribution can be observed, where the blue color indicates O-H bonds corresponding to water, and the red color corresponds to the presence of LaF3, and in fact it reveals the full coating, since the technique cannot resolve the layers within the ML; the technique integrates a total thickness of ∼500 nm, which is approximately coincident with the total thickness of the deposited MLs. For the sample stored in vacuum, the main observed phase is LaF3 (through the detection of La) with only a few points observed on the surface image corresponding to water, while no water appears in the depth profile. In the case of the sample stored at 60% RH, a larger water content is observed, and it appears as small isolated regions on the superficial Raman image; this agrees with the wide Raman band observed at ∼3500 cm-1. In the depth profile, only isolated blue points appear, indicating that the water contribution deep in the ML is negligible and that the water regions on the surface can be probably eliminated through vacuum outgassing, and/or post-annealing. Finally, in the sample stored with 99% RH, a great amount of water is observed both on the surface and in the depth profile, presenting a more homogeneous distribution along the surface. The depth profile shows a thick layer of water located on the surface, that can penetrate the sample. This is observed since the LaF3 and water phases overlap. The reflectance of the sample stored at 99% RH decayed by around 50%.
To conclude, the sample stored in an intermediate RH (∼60%) hardly undergoes deterioration due to water adsorption, since the water accumulated on the surface corresponds mainly to free O-H radicals, which are easily removable (through vacuum degassing and/or post-annealing), and consequently, optical properties are hardly affected, in resemblance to the vacuum stored sample. Contrarily, for the coating in contact with an extreme humidity such as 99%, the adsorption of water is more pronounced, producing high surface damage, and even cracks where the water accumulated. This occurred with a reflectance decay for this sample by around 50%.
4. Conclusions
Narrowband multilayer coatings peaked at the short FUV wavelengths of H Ly-α (121.6 nm) and the O I doublet (130.4 and 135.6 nm) have been developed. They consist of MgF2/LaF3 and AlF3/LaF3 MLs deposited by thermal evaporation on a substrate at ≈230-270°C. Fresh AlF3/LaF3 and MgF2/LaF3 MLs reflected at Ly-α up to ≈87% and ≈75%, respectively,. To the best of our knowledge, this result on AlF3/LaF3 ML is the highest reflectance of a narrowband ML at this wavelength obtained so far, with a strong reflectance increase from 65% in the literature.
MLs have also been developed with the additional requirement of having a strong rejection at a close spectral line for applications in which such line can mask imaging at the target wavelength. We present a ML with high reflectance simultaneously at the OI lines, of ≈65% and ≈73% at 130.4 and 135.6 nm, respectively, while the reflectance at Ly-α is below ≈2%. Alternatively, we present a coating centered at Ly-α with strong rejection at the close O I doublet lines of 130.4 nm and 135.6 nm, which could be beneficial to avoid parasitic light in observations for space instruments like GLIDE; reflectance of ≈74% at Ly-α and ≈1% and ≈5% at 130.4 and 135.6 nm, respectively, is presented.
We also investigated the effect of different storage environments on ML reflectance. Samples stored in vacuum presented a decay of ≈1% after two months (AlF3/LaF3), while the reflectance decay of samples stored in a desiccator was ≈2% and ≈4% after two and six months (AlF3/LaF3) and ≈3% after six months (MgF2/LaF3). Finally, some samples were stored in relatively humid environments (RH between 50% and 70%) for one month, resulting in a reflectance loss at the band peak (121.6 nm) of ≈2% (AlF3/LaF3). Summing up, all of the observed reflectance losses are below 5%, even in relatively humid environments such as 50-70% of RH.
Some samples were studied with RAMAN spectroscopy to understand the water adsorption process. Results show that for samples stored in vacuum, desiccator, or RH from ≈50% to ≈70%, the accumulated water in the surface corresponds to free O-H radicals, which are expected to be easily removable through vacuum outgassing.
Funding
Agencia Estatal de Investigación (BES-2017-081909, ESP2016- 76591-P, PID2019-105156GB-I00); Ministerio de Economía y Competitividad (BES-2017-081909).
Acknowledgments
We gratefully acknowledge José L. Bris and Joaquín Campos for spectrophotometer measurements. LRM acknowledges CRESST II cooperative agreement supported by NASA under award number 80GSFC21M0002.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. K. Sembach, “Technology Investments to Meet the Needs of Astronomy at Ultraviolet Wavelengths in the 21st Century,” Astro2010 Technology Development White Paper. Available at https://www.astro.princeton.edu/∼dns/Theia/Sembach_UVtechnology_EOS.pdf (2010).
2. J. Tumlinson, A. Aloisi, G. Kriss, et al., “Unique Astrophysics in the Lyman Ultraviolet,” arXiv, arXiv:1209.3272 (2012). [CrossRef]
3. Cite https://apd440.gsfc.nasa.gov/tech_gap-descriptions.html
4. C. Danson, C. Haefner, J. Bromage, et al., “Petawatt and exawatt class lasers worldwide,” High Power Laser Sci. Eng 7, e54 (2019). [CrossRef]
5. S. Kubodera and W. Sasaki, “Vacuum Ultraviolet Ar Excimer Emission Initiated by High Intensity Laser Produced Electrons,” in Proceedings of the 2000 Symposium on Nuclear Data, JAERI, N. Yamano and T. Fukahori, eds., 2006, pp. 285–288.
6. P. J. M. Peters, L. Feenstra, and H. M. J. Bastiaens, “Long-pulse ArF and F2 excimer lasers,” Proc. SPIE 4184, 338–347 (2001). [CrossRef]
7. D. J. Eckstrom, H. H. Nakano, D. C. Lorents, T. Rothem, J. A. Betts, M. E. Lainhart, and D. A. Dakin, “Characteristics of electron-beam-excited Kr* 2 at low pressures as a vacuum ultraviolet source,” J. Appl. Phys. 64(4), 1691–1695 (1988). [CrossRef]
8. R. Forbes, P. Hockett, and R. Lausten, “Efficient generation of the 7th harmonic of Ti: sapphire (114.6 nm) vacuum ultraviolet pulses with 60 fs duration by non-collinear four-wave mixing in argon,” Opt. Lett. 47(10), 2410–2413 (2022). [CrossRef]
9. P. Gohil and D. D. Burgess, “A comparison between laser induced fluorescence at Balmer-alpha and at Lyman-alpha for the measurement of neutral hydrogen densities in magnetically contained fusion plasmas,” J. Plasma Phys. 25(10), 1149–1167 (1983). [CrossRef]
10. N. Gambino, B. Rollinger, D. Hudgins, and R. S. Abhari, “Spectral emission properties of a laser-produced plasma light source in the sub-200 nm range for wafer inspection applications,” J. Micro/Nanolith. MEMS MOEMS 14(3), 034002 (2015). [CrossRef]
11. G. D. Krebs, “GLIDE,” Gunter’s Space Page. Retrieved December 15, 2021, from https://space.skyrocket.de/doc_sdat/glide.htm
12. J. C. Green and K. France, “SubLymE: The Sub-Lyman α Explorer,” Proc. SPIE 9144, 914405 (2014). [CrossRef]
13. National Academies of Sciences, Engineering, and Medicine, “Decadal survey on astronomy and astrophysics 2020” (2021).
14. K. France, B. Fleming, G. West, S. R. McCandliss, M. R. Bolcar, W. Harris, and J. Tumlinson, “The LUVOIR ultraviolet multi-object spectrograph (LUMOS): instrument definition and design. In UV, X-ray, and gamma-ray space instrumentation for astronomy XX,” Proc. SPIE 10397, 1039713 (2017). [CrossRef]
15. National Aeronautics and Space Administration, “LUVOIR: Final Report. The LUVOIR Final Report” (nasa.gov) (2019).
16. E. Spiller, “Interference in thin films: theory,” Soft X-Ray Optics, (SPIE, 2011), pp. 101–137.
17. J. I. Larruquert, Optical Properties of Thin Film Materials at Short Wavelengths, A. Piegari and F. Flory, eds. (Woodhead Publishing, 2018).
18. D. W. Angel, W. R. Hunter, R. Tousey, and G. Hass, “Extreme Ultraviolet Reflectance of LiF-Coated Aluminum Mirrors,,” J. Opt. Soc. Am. 51(8), 913–914 (1961). [CrossRef]
19. B. Fleming, M. Quijada, J. Hennesy, A. Egan, J. del Hoyo, B. A. Hicks, J. Wiley, N. Kruczek, N. Erickson, and K. France, “Advanced environmentally resistant lithium fluoride mirror coatings for the next generation of broadband space observatories,” Appl. Opt. 56(36), 9941–9950 (2017). [CrossRef]
20. L. V. Rodríguez-de Marcos, J. I. Larruquert, J. A. Méndez, M. Vidal-Dasilva, S. García-Cortés, N. Gutiérrez-Luna, L. Espinosa-Yáñez, C. Honrado-Benítez, and J. Chavero-Royán, “Lyman-β narrowband coatings with strong Lyman-α rejection,” Opt. Express 26(19), 25166–25177 (2018). [CrossRef]
21. M. Zukic, D. G. Torr, J. F. Spann, and M. R. Torr, “Vacuum ultraviolet thin films 2: Vacuum ultraviolet all dielectric narrowband filters,,” Appl. Opt. 29(28), 4293–4302 (1990). [CrossRef]
22. M. Zukic and D. G. Torr, “Multiple reflectors as narrow-band and broadband vacuum ultraviolet filters,” Appl. Opt. 31(10), 1588–1596 (1992). [CrossRef]
23. H. Bernitzki, H. Lauth, R. Thielsch, H. Blaschke, N. Kaiser, and K. R. Mann, “Current status of radiation resistance of dielectric mirrors in the DUV,,” Proc. SPIE 3578, 105–116 (1999). [CrossRef]
24. R. Thielsch, “Optical coatings for the DUV / VUV,” in Optical Interference Coatings, N. Kaiser and H.K. Pulker, eds. (Springer, 2003), pp. 257–277
25. J. E. Rudisill, A. Dupparre, and S. Schroeder, “Determination of scattering losses in ArF* excimer laser all-dielectric mirrors for 193 nm microlithography application,,” Proc. SPIE 5647, 9–22 (2005). [CrossRef]
26. S. Günster, D. Ristau, A. Gatto, N. Kaiser, M. Trovó, and M. Danailov, “Storage ring free-electron lasing at 176 nm-dielectric mirror development for vacuum ultraviolet free-electron lasers,,” Appl. Opt. 45(23), 5866–5870 (2006). [CrossRef]
27. G. H. Liu, Q. L. Xiao, Y. X. Jin, W. L. Zhang, H. B. He, and Z. X. Fan, “Mechanical stress in 355 nm LaF3/MgF2 high reflectors with various layer-pair number and methods for reduction,,” Vacuum 84(6), 778–781 (2010). [CrossRef]
28. L. Rodríguez-de Marcos, J. I. Larruquert, J. A. Méndez, and J. A. Aznárez, “Multilayers and optical constants of various fluorides in the far UV,” Proc. SPIE 9627, 96270B (2015). [CrossRef]
29. N. Gutiérrez-Luna, P. López-Reyes, C. Honrado-Benítez, L. Espinosa-Yáñez, Á Ríos-Fernández, L. V. Rodríguez-de Marcos, M. A. Quijada, and J. I. Larruquert, “Mirrors for improved FUV space observations,” Proc. SPIE 11444, 1144478 (2020). [CrossRef]
30. P. López-Reyes, B. Perea-Abarca, C. Honrado-Benítez, N. Gutiérrez-Luna, A. Ríos-Fernández, L. V. Rodríguez-de Marcos, and J. I. Larruquert, “Optimization of the deposition parameters of MgF2/LaF3 narrowband reflective FUV multilayers,” Opt. Mater. Express 11(6), 1678–1691 (2021). [CrossRef]
31. L. V. Rodríguez-de Marcos, J. I. Larruquert, J. A. Méndez, and J. A. Aznárez, “Self-consistent optical constants of MgF2, LaF3, and CeF3 films,” Opt. Mater. Express 7(3), 989–1006 (2017). [CrossRef]
32. C. Guo, M. Kong, D. Lin, and B. Li, “Fluoride coatings for vacuum ultraviolet reflection filters,,” Appl. Opt. 54(35), 10498–10503 (2015). [CrossRef]
33. J. Park, M. Zukic, M. W. McColgan, C. E. Keffer, D. G. Torr, and R. B. Hoover, “Design and fabrication of multilayer reflective filters for a Ritchey-Chretien Lyman-alpha telescope,” Opt. Eng. 35(5), 1479–1482 (1996). [CrossRef]
34. X. Wang, B. Chen, Y. Liu, T. Huo, and H. Zhou, “Design and fabrication of narrowband 121.6 nm minus filters,” Proc. SPIE 10691, 106911K (2018). [CrossRef]
35. H. Schink, J. Kolbe, F. Zinmermann, D. Ristau, and H. Welling, “Reactive Ion-Beam-Sputtering of fluoride coatings for the UV/VUV range,” Proc. SPIE 1441, 327–338 (1991). [CrossRef]
36. J. Kolbe, H. Kessler, T. Hofmann, F. Meyer, H. Schink, and D. Ristau, “Optical properties and damage thresholds of dielectric UV/VUV-coatings deposited by conventional evaporation, IAD and IBS,” Proc. SPIE 1624, 1221 (1991). [CrossRef]
37. J. Sun, X. Li, W. Zhang, K. Yi, and J. Shao, “High-reflectivity mirrors by Al2O3, LaF3 and AlF3 for 193 nm application,” Opt. Laser Technol. 56, 65–70 (2014). [CrossRef]
38. R. Thielsch, A. Gatto, and N. Kaiser, “Mechanical stress and thermal-elastic properties of oxide coatings for use in the deep-ultraviolet spectral region,” Appl. Opt. 41(16), 3211–3217 (2002). [CrossRef]
39. A. Ode, “Ion beam sputtering of fluoride thin films for 193 nm applications,,” Appl. Opt. 53(4), A330–A333 (2014). [CrossRef]
40. R. B. Hoover, T. W. Barbee Jr, P. C. Baker, J. F. Lindblom, M. J. Allen, C. E. DeForest, and A. B. Walker II, “Performance of compact multilayer coated telescopes at soft x-ray/EUV and far ultraviolet wavelengths,” Opt. Eng. 29(10), 821 (1990). [CrossRef]
41. N. Narukage, M. Kubo, R. Ishikawa, S. N. Ishikawa, Y. Katsukawa, T. Kobiki, and J. Choi, “High-reflectivity coatings for a vacuum ultraviolet spectropolarimeter,” Sol. Phys. 292(3), 40 (2017). [CrossRef]
42. X. D. Wang, B. Chen, H. F. Wang, F. He, X. Zheng, L. P. He, B. Chen, S. J. Liu, Z. X. Cui, X. H. Yang, and Y. P. Li, “Design and fabrication of far ultraviolet filters based on π-multilayer technology in high-k materials,” Sci. Rep. 5(1), 8503 (2015). [CrossRef]
43. A. Vidal-Madjar, J. M. Désert, A. Lecavelier des Etangs, G. Hébrard, G. E. Ballester, D. Ehrenreich, R. Ferlet, J. C. McConnell, M. Mayor, and C. D. Parkinson, “Detection of Oxygen and Carbon in the Hydrodynamically Escaping Atmosphere of the Extrasolar Planet HD 209458b,” ApJ 604(1), L69–L72 (2004). [CrossRef]
44. P. López-Reyes, C. Honrado-Benítez, N. Gutiérrez-Luna, Á Ríos-Fernández, L. V. Rodríguez-de Marcos, and J. I. Larruquert, “Far-UV reflectance and stress of narrowband AlF3/LaF3 multilayers,” Opt. Mater. Express 12(2), 489–502 (2022). [CrossRef]
45. “GOLD-IO-Lab,” https://gold.io.csic.es/.
46. R. Abermann, R. Kramer, and J. Mäser, “Structure and internal stress in ultra-thin silver films deposited on MgF2 and SiO substrates,” Thin Solid Films 52(2), 215–229 (1978). [CrossRef]
47. J. Ullmann, H.-G. Keck, R. Thielsch, H. Uhlig, and N. Kaiser, “Mechanical stress in fluoride coatings,” Proc. SPIE 3738, 136–147 (1999). [CrossRef]
48. O. R. Wood, H. G. Craighead, J. E. Sweeney, and P. J. Maloney, “Vacuum ultraviolet loss in magnesium fluoride films,” Appl. Opt. 23(20), 3644–3649 (1984). [CrossRef]
49. L. V. R. De Marcos, J. I. Larruquert, J. A. Méndez, N. Gutiérrez-Luna, L. Espinosa-Yáñez, C. Honrado-Benítez, J. Chavero-Royán, and B. Perea-Abarca, “Optimization of MgF2 –deposition temperature for far UV Al mirrors,” Opt. Express 26(7), 9363–9372 (2018). [CrossRef]
50. N. Gutiérrez-Luna, B. Perea-Abarca, L. Espinosa-Yáñez, C. Honrado-Benítez, T. de Lis, L. Rodríguez-de Marcos, J. A. Aznárez, and J. I. Larruquert, “Temperature dependence of AlF3 protection on far-UV Al mirrors,” Coatings 9(7), 428 (2019). [CrossRef]
51. E. Hecht, Optics, 4th Edition (Addison Wesley, 2002).
52. L. G. Parratt, “Surface studies of solids by total reflection of X-rays,” Phys. Rev. 95(2), 359–369 (1954). [CrossRef]
53. D. G. Stearns, “X-ray scattering from interfacial roughness in multilayer structures,” J. Appl. Phys. 71(9), 4286–4298 (1992). [CrossRef]
54. D. L. Windt, “IMD—software for modeling the optical properties of multilayer films,” Comput. Phys. 12(4), 360–370 (1998). [CrossRef]
55. D. Ristau, S. Günster, S. Bosch, A. Duparré, E. Masetti, J. Ferré-Borrull, G. Kiriakidis, F. Peiró, E. Quesnel, and A. Tikhonravov, “Ultraviolet optical and microstructural properties of MgF2 and LaF3 coatings deposited by ion-beam sputtering and boat and electron-beam evaporation,” Appl. Opt. 41(16), 3196–3204 (2002). [CrossRef]
56. S. Schröder, H. Uhlig, A. Duparré, N. Kaiser, F. Angewandte, and D. Jena, “Nanostructure and optical properties of fluoride films for high-quality DUV / VUV optical components,” Proc. SPIE 5963, 59630R (2005). [CrossRef]
57. “Nanostructure influence on high reflectivity of fluoride Multilayers in the Far-UV: electronic microscopy and high-resolution RBS characterization,” To be submitted soon
58. N. Gutiérrez-Luna, private communication.
59. J. A. Dobrowolski, “Optical properties of films and coatings,” Handbook of Optics I, 42–43 (1995).
60. P. D. Binda and F. E. Zocchi, “Genetic algorithm optimization of X-ray multilayer coatings,” Proc. SPIE 5536, 97–108 (2004). [CrossRef]
61. M. M. Ahmad, Y. Yamane, and K. Yamada, “Structure, ionic conduction, and giant dielectric properties of mechanochemically synthesized BaSnF,” J. Appl. Phys. 106(7), 074106 (2009). [CrossRef]
62. Y. Tominaga, A. Fujiwara, and Y. Amo, “Dynamical structure of water by Raman spectroscopy,” Fluid Phase Equilibria 144(1-2), 323–330 (1998). [CrossRef]
