Study of Optical Performance and Structure of Yb/Al (1.5 wt.% Si) and Yb/Al (Pure) Multilayers Designed for the 73.6 nm Range

Abstract

Yb/Al multilayer films exhibit excellent theoretical reflectivity in the 54–90 nm wavelength range. This study attempted to incorporate 1.5% wt.% of Si impurities into Al to suppress the crystallization of Al, reduce interfacial roughness, and enhance the actual reflectivity of the prepared Yb/Al multilayer films. Internal microstructure changes in the film layers before and after Si impurity doping were investigated using GIXRR, AFM, and XRD techniques. The reflectivity of two types of multilayer films, Yb/Al (1.5 wt.% Si) and Yb/Al (pure), was tested to evaluate the effect of Si impurity on film performance. The reflectivity of Yb/Al (1.5 wt.% Si) multilayers compared to Yb/Al (pure) multilayers increased by approximately 4%.

1. Introduction

In the solar atmosphere, extreme ultraviolet radiation is observed via EUV imaging instruments to understand solar activity [1]. Numerous extreme ultraviolet solar observation satellites have been launched internationally, including the Solar and Heliospheric Observatory (SOHO) [2], the Solar Terrestrial Relations Observatory (STEREO) [3], the Solar Dynamics Observatory (SDO) [4], and the Solar Orbiter [5], which have all advanced relevant research. Multilayer mirrors are the core components of extreme ultraviolet solar observation systems. Above the 115 nm wavelength range, multilayer films with high reflectivity have been fabricated using MgF₂ and Al, or MgF₂ in combination with other fluorides (such as BaF₂, LaF₃, and LiF) [6,7,8]. Below the 50 nm range, high-reflectivity multilayer films such as Sc/Si [9,10,11] or Mg/SiC [12,13] have been developed. However, within the 54–90 nm range, due to strong absorption by materials, the fabrication of highly efficient mirrors is challenging. Lanthanide materials exhibit relatively low absorption within the 54–90 nm range, which makes the production of multilayer mirrors feasible within this spectrum. Research by Windt et al. has led to the development of Si/Tb, SiC/Tb, Si/Gd, and Si/Nd multilayer films, which have demonstrated peak reflectivity values between 12% and 27% in the 55–69 nm range [14,15,16,17].

Among lanthanide materials, ytterbium (Yb) has a lower absorption coefficient in the 55–90 nm range [18]. Yb/Al multilayer films can enhance reflectivity in this spectrum. In 2009, Vidal-Dasilva et al. fabricated Yb/Al/Yb/SiO thin films using a thermal evaporation method, achieving peak reflectivity values of 27.6% at 80 nm and 24.7% at 85 nm [19]. After two years of storage in a drying cabinet, the peak reflectivity at 80 nm decreased to about 20%. The inclusion of a SiO-diffusion barrier layer improved the reflectivity of Al/SiO/Yb/SiO films to 19.1% at 91 nm, and the reflectivity of Yb/SiO/Al/SiO/Yb/SiO films to 18.3% at 78 nm; this also enhanced storage stability. The addition of a small percentage of silicon (Si) into aluminum (Al) can be an effective method for modifying the crystalline structure of the Al layer in multilayer films. Scientists like Zhong and colleagues, by doping Al with around 1% Si, aimed to reduce the grain size in the Al layers, which led to a smoother interface between the layers [20]. Between 2010 and 2014, Meltchakov et al. improved the deposition process by using Al targets doped with 1.5 wt.% Si or 2 wt.% Cu instead of pure Al. The resulting Al/Mo/B4C multilayers achieved a peak reflectivity of 55.5%, 49.8%, and 42.1% at 17.5 nm, 20.9 nm, and 30.1 nm, respectively [21,22,23,24].

In the spectrum of extreme ultraviolet (EUV) radiation of the solar transition region, several crucial spectral lines exist, such as O III (70.3 nm), O II (71.8 nm), O V (76.0 nm), Ne VIII (77.0 nm/78.0 nm), and O IV (79.0 nm), which fall within the range of 70–80 nm. Considering the measurement capabilities of the laboratory reflection spectrometer, Ne I (73.6 nm) was chosen as the target spectral line for thin-film research in this study. The multilayer films of Yb/Al, designed to operate at a target wavelength of 73.6 nm, were fabricated using pure aluminum and aluminum doped with 1.5 wt.% silicon layers. We aimed to enhance the reflectivity efficiency of Yb/Al through the introduction of Si impurities and to determine the reasons for this enhancement. Employing techniques such as grazing incidence X-ray reflectivity (GIXRR), X-ray diffraction (XRD), and atomic force microscopy (AFM), this investigation aimed to analyze the impact of silicon incorporation on the internal structure of Yb/Al multilayer films. Additionally, extreme ultraviolet (EUV) reflectivity measurements were conducted to evaluate the optical performance of the multilayer films. The Yb/Al (pure) and Yb/Al (1.5 wt.% Si) multilayers exhibited reflectivities of 11.9% and 15.4%, respectively, when subjected to a 4° incidence angle.

2. Measurements and Data Analysis

The Yb/Al multilayers were fabricated utilizing a high-vacuum direct-current magnetron-sputtering system operating in planet mode [16]. Modulation of the layer thickness was achieved by adjusting the speed of sample translation across the target. The distance between the target and the substrate was maintained at 7.9 cm for the Yb, Al (pure), and Al (1.5 wt.% Si) targets. Prior to deposition, the vacuum chamber’s background pressure was typically reduced to 1.0 × 10^-4 Pa. During deposition, ultra-high-purity Ar (99.999%) served as the sputtering gas; the Ar pressure was set to 1.2 mTorr. The sputtering sources operated under constant power conditions.

Structural characterization of the sample films was performed via grazing incidence X-ray reflectivity (GIXRR) utilizing an X-ray diffractometer (Bede, Durham, UK) operated in the 2theta-omega configuration. The incident X-ray source was provided by Cu-Kα radiation at a wavelength of 0.154 nm [25]. Reflectivity profiles obtained from GIXRR were subsequently modeled and fitted, employing the IMD software 4.0 to deduce the layer structures and interface quality. Crystalline properties of the multilayers were assessed via X-ray diffraction (XRD), conducted on a Bruker D8 Advance diffractometer. Diffraction patterns were acquired via symmetric reflection geometry, with the detector scan range set between 25° and 40° at a grazing incidence angle. This arrangement facilitated the detection of crystallographic planes oriented parallel to the surface of the multilayer. Crystallization features were investigated by matching the angular positions of the diffraction peaks to the standard powder diffraction files (PDFs) provided by the International Centre for Diffraction Data (ICDD). The mean grain size, oriented perpendicular to the direction of crystal growth, was estimated through the application of the Scherrer equation to broaden the diffraction peaks.

$$D=\frac{\mathrm{K}\lambda }{Bcos\left(\theta \right)}$$

(1)

where K is a dimensionless shape factor with a value of 0.89 (assuming spherical grains), λ is the X-ray wavelength (0.154 nm for Cu-Kα radiation), B is the full width at half the maximum (FWHM) of the XRD peak, and θ is the Bragg angle corresponding to the peak position [26].

The surface topography of the specimens was meticulously examined utilizing atomic force microscopy (AFM), employing a Bruker Dimension Icon apparatus, which was operated in tapping mode. The examination entailed scanning areas of disparate dimensions—namely, 10 × 10 μm², 4 × 4 μm², and 1 × 1 μm²—each with a resolution set at 256 × 256 pixels. To quantify the distribution of surface roughness across various spatial frequencies, the power spectral density (PSD) of the surface was determined using methods derived from Fourier analysis [27].

Reflectivity, as a function of angle, was measured using a specialized extreme ultraviolet reflectometer [28]. Utilization of this apparatus permitted the quantification of reflectivity within the extreme ultraviolet (EUV) spectral domain spanning 30–200 nm. The instrument was equipped with a chopper to facilitate the separation of the reference beam, thus allowing for the precise measurement of very low photocurrents down to the order of 10 pA. The measurement error for reflectivity was constrained to less than ±0.5%. This capability was critical for assessing the angular dependence of reflectivity with high accuracy, especially for samples that demonstrated weak photocurrent signals in the extreme ultraviolet (EUV) region.

3. Results and Discussion

3.1. GIXRR

The measured and fitted GIXRR curves of Yb/Al multilayers with two kinds of Al targets are shown in Figure 1, where the dotted lines represent the measurements and the solid lines represent the fitting result, obtained using the software IMD. The measured curves were fitted with a two-layer model.

Table 1. GIXRR fitting results of Yb/Al multilayers deposited with different Al layers.

Sample	SiC Thickness	Yb Thickness	Al Thickness	Average Roughness
Yb/Al (pure)	9.87 nm	19.28 nm	41.44 nm	1.95 nm
Yb/Al (1.5 wt.% Si)	9.68 nm	20.57 nm	38.78 nm	1.89 nm

presents the fitted structure of Yb/Al multilayers. Overall, the multilayer structure and performance of the Yb/Al (1.5 wt.% Si) multilayer films were significantly better than those of the Yb/Al (pure) multilayer films.

Fewer small reflection peaks were observed in the grazing incidence X-ray reflectivity (GIXRR) test curves for Yb/Al (pure) multilayer films. When the grazing incidence angle becomes larger than 0.6 degrees, the characteristic peaks of the Yb/Al (pure) multilayer films disappear more quickly due to the factors above, which degrade the X-ray reflectivity performance. The disappearance of peaks at higher angles is less pronounced in the Yb/Al (1.5 wt.% Si) multilayer films, which maintain their reflectivity characteristics over a broader range of angles, indicative of their superior interface quality. From the fitting results, it can also be observed that the roughness of Yb/Al (1.5 wt.% Si) is slightly smaller than the roughness of Yb/Al (pure).

3.2. AFM

Table 2. Root mean square (RMS) surface roughness at different test areas of AFM.

Sample	10 × 10 μm2	4 × 4 μm2	1 × 1 μm2
Yb/Al (pure)	4.85 nm	3.01 nm	2.41 nm
Yb/Al (1.5 wt.% Si)	1.81 nm	2.04 nm	2.09 nm

delineates the root mean square (RMS) surface roughness of Yb/Al multilayers, ascertained via atomic force microscopy (AFM) across disparate scanning regions (Figure 2). The surface morphology of Yb/Al (pure) multilayers is characterized by the presence of relatively large particulate features, with dimensions of approximately 0.1 μm. Such particles contribute to pronounced surface roughness, an observation that dovetails with the empirical data gathered from the grazing incidence X-ray reflectivity (GIXRR) assessments.

3.3. XRD

Figure 3 illustrates the X-ray diffraction (XRD) patterns of Yb/Al multilayer structures recorded in symmetrical reflection mode. The observed diffraction peaks showed angular distributions at 28.405°, 32.210°, and 38.897°, which may be attributed to the crystallographic planes of Yb (111), Yb₂O₃ (123), and Al (111), respectively [29,30,31,32]. In the comparative analysis of the diffraction peaks between the Yb/Al (pure) and the Yb/Al (1.5 wt.% Si) multilayers, a noticeable diminution in the intensity of the peaks for the latter was observed, particularly the peak corresponding to the Al (111) plane, which became remarkably subdued. Additionally, the peak associated with Yb (111) exhibited a slight reduction in intensity. Utilizing the Scherrer formula, which relates the FWHM of a diffraction peak to the size of the coherent diffracting domains or crystallites in a polycrystalline sample, the average grain size of the crystallographic phases was quantified. These calculations are presented in

Table 3. Average grain size of different crystal phases.

Sample	Yb (111)	Yb2O3 (123)	Al (111)
Yb/Al (pure)	7.636 nm	4.865 nm	8.730 nm
Yb/Al (1.5 wt.% Si)	7.895 nm	4.915 nm	9.318 nm

.

The introduction of Si inhibits the crystallization of Al. We hypothesize that the relatively weaker crystallization of Al enhances its diffusion with Yb, leading to a reduction in the crystallinity of Yb. The diffusion between Al and Yb counteracts the improvement in interfacial roughness resulting from the reduced crystallization of Al. Therefore, the difference in average roughness, obtained from the fitted GIXRR curves, between Yb/Al (pure) and Yb/Al (1.5 wt.% Si) is not significant.

3.4. Reflectivity

The angular dependency of the reflectivity of the Yb/Al multilayers was systematically investigated at 73.6 nm and the results are depicted in Figure 4, which encompasses incident angles ranging from 3° to 20°, incremented at 1° intervals. For each specified angle, the detection of the reflected light’s intensity, alongside that of a reference beam, was performed at a sampling rate of 512 Hz. Subsequent to this, a subtraction of the background noise from the signal was executed, ensuring that the data reflected the true reflectivity of the sample. To facilitate a comparison, the intensity of the reflected light was normalized against that of the reference beam.

Accompanying the experimental data, Figure 4 provides the theoretical reflectivity curves as computed by the IMD (Integrated Munster Data) software. These curves delineate the expected reflectivity of an ideally fabricated multilayer structure with negligible surface roughness. The Yb/Al (pure) and Yb/Al (1.5 wt.% Si) multilayers exhibited reflectivities of 11.9% and 15.4%, respectively, when subjected to a 4° incidence angle. Notably, the reflectivity of the Yb/Al (1.5 wt.% Si) surpassed that of the Yb/Al (pure) by approximately four percentage points. The measured reflectivity of the Yb/Al multilayers was observed to be inferior to the theoretical reflectivity values.

The observed discrepancy between the experimental and theoretical reflectivity values could be primarily ascribed to a combination of factors, including surface roughness, interlayer diffusion, and the onset of oxidation at the surface. Surface roughness can lead to increased scattering of light, thus diminishing reflectivity. Interlayer diffusion may result in a blurring of the interfaces, which can also adversely affect reflectivity properties. Furthermore, the formation of an oxide layer can alter the electronic structure at the surface, impeding optimal reflection. These phenomena underscore the challenges in the fabrication of multilayer structures where deviations from the ideal design parameters can significantly impact functionality. We attributed this discrepancy to roughness, interlayer diffusion, and surface oxidation.

4. Conclusions

In a direct-current (DC) magnetron sputtering system, two types of Yb/Al multilayer films were fabricated using pure Al targets and Si-doped Al targets. A suite of characterization techniques, including grazing incidence X-ray reflectometry (GIXRR), atomic force microscopy (AFM), X-ray diffraction (XRD), and angular dependence of reflectivity studies revealed significant differences in the structure and performance between the two multilayer materials.

The Yb/Al (1.5 wt.% Si) multilayer films demonstrated superior reflective performance. X-ray diffraction analysis indicated that the incorporation of Si effectively reduced the grain size within the Al layers and slightly decreased the crystallinity of Yb. These combined effects led to a decrease in interfacial roughness, which positively impacted the reflectivity and the outcomes of the rocking curve analysis. Notably, the reduction in interfacial roughness played a critical role in enhancing specular reflection and reducing non-specular scattering.

Furthermore, power spectral density (PSD) analysis corroborated the improvement in surface and interfacial quality due to Si doping, as reflected in the lower surface roughness and more optimized interlayer transitions.

In summary, the introduction of Si, by fine-tuning the structural properties of the multilayer films, not only enhanced the specular reflectivity of the Yb/Al multilayer films but also reduced non-specular scattering. This led to a significant improvement in performance for optical applications such as high-precision mirrors and optical coatings. These findings underscore the importance of material engineering in the design of multilayer films, particularly with respect to the quality of optical interfaces in high-performance optical applications.