All samples were prepared in a coating machine (ZZS800, Chengdu Nanguang Machinery Co. Ltd., China). This device was equipped with four thermal evaporators and two ceramic heaters. In the coating process a base pressure about 1.2 × 10⁻⁴ Pa was provided in the coating chamber with a cryopump set that ensured an oil-free environment for all depositions. The ultimate vacuum pressure was 8.0 × 10⁻⁵ Pa. In order to obtain a better vacuum and reduce the influence of water vapor on the properties of thin films, the vacuum coating machine with a polycold was a good choice. BK7 glasses (Φ 25.4 × 4 mm) with a root-mean-square surface roughness of approximately 0.5 nm were used as substrates and fixed above the evaporators. The distance between the evaporation source and substrate was about 450 mm. For deposition, the BK7 substrates were cleaned manually with alcohol and acetone. Tungsten-boat was used for the evaporation of high purity Al (99.999%), while MgF₂ grains (Merck) was deposited by molybdenum-boat evaporation. The deposition rate and physical thickness for each of the thin films were controlled by a quartz-crystal monitor, which was fixed at the top center of the deposition chamber. Deposition rates of 10 nm/s and 0.2 nm/s were for Al and MgF₂ layers, respectively. Deposition angles for both materials were smaller than 20°. A series of Al mirrors (each with a thickness of 80 nm) over-coated with a thin MgF₂ layer was deposited by a 3-step coating process. The main step of the method had been well outlined and documented in other literatures.^[9,11,15] In order to enhance the reflectance at 121.6 nm, the total thickness of MgF₂ layer was set to be 25 nm. For the thin film design, optical constants of Al were obtained from the literature.^[19] Parameters of MgF₂ films were determined from spectrophotometric measurements as presented in Ref. [20]. The substrate temperature and the holding time were varied for preparing the second MgF₂ layer. The deposition temperatures ranged from 25 °C to 350 °C. The holding time was set to be 0 and 120 min. The prepared sample types were summarized in Table 1. In the heating process, temperature and pressure profiles versus time are exhibited in Fig. 1. As temperature increased, the pressure significantly increased, which was caused by outgassing from the walls of the chamber. After fabrications, all samples were cooled to room temperature, and then the vacuum chamber was vented.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. Plot of substrate temperature and pressure versus time in coating chamber.

Table 1.

Table 1.

Table 1. Overview of prepared sample types. . Type dAl/nm Tsub/°C d1 MgF2/nm Tsub/°C d2 MgF2/nm Tsub/°C Heating time/min Holding time/min M1 80 25 25 25 — — 0 0 M2 80 25 5 25 20 200 23 0 M3 80 25 5 25 20 250 42 0 M4 80 25 5 25 20 300 73 0 M5 80 25 5 25 20 350 118 0 M6 80 25 5 25 20 300 73 120

Table 1. Overview of prepared sample types. .

The reflectance for each of all Al mirrors was measured in the spectral range 115 nm–220 nm in steps of 0.1 nm and at an incident angle of 10° by using a high-precision ML6500 VUV spectrophotometer (Laser Zentrum Hannover, Germany). The nominal measurement accuracy of the VUV spectrophotometer is ±0.3% for the reflectance. In order to avoid the absorption of oxygen, moisture and hydro-carbon contamination, the spectrum was measured under vacuum environment with a pressure below 1.0 × 10⁻³ Pa. The surface morphology was characterized by using a Dimension 3100 atomic force microscope, performing in a tapping mode for imaging surface over scanning size of 5 μm square with 256 × 256 data points. Crystalline structure was evaluated with a Philips X’Pert-MRD x-ray diffraction. The start and end incident angle were 10° and 80°, respectively, in steps of 0.01°. Moreover, in order to study the humidity sensitivity of the as-deposited Al mirrors, samples were accelerated the ageing process in a custom humidity control chamber held at temperature of 50 °C and relative humidity of 95 ± 3% for 12 h, 24 h, and 36 h, respectively. After humidity experiments, the spectral performances of Al mirrors were measured and corresponding degradation of the reflectance was determined.

Figure 2(a) shows the measured reflectance spectra of the MgF₂ over-coated Al mirror with different deposition temperatures. To reduce the influence of stochastic variation in the layer thickness and random measurement error on the result of optical performance ranking, in each case, a spectrally averaged reflectance has been obtained for ranking purposes. Therefore, the reflectance at 121.6 nm is calculated by averaging the measured spectra from 121 nm to 122 nm (with 11 data points) and exhibited in Fig. 2(b), while the VUV reflectance is averaged in the wavelength region of 115 nm–220 nm (Fig. 2(c)). When the deposition temperature is between 25 °C and 250 °C, the spectral performance is improved with the increase of substrate temperature. Above 250 °C, the application of the higher deposition temperature leads the reflectance to significantly drop down. The experimental result is in agreement with that in Ref. [9].

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. (a) Spectral performance, (b) reflectance (121.6 nm), and (c) average reflectance (115 nm–220 nm) of MgF2 over-coated Al mirrors with various substrate temperatures.

Generally, the reflectance of MgF₂ over-coated Al mirror in the VUV is influenced mainly by the absorptance and scatterring of the prepared layer system. Absorptance is due to the fact that both MgF₂ and alumina have non-negligible absorption in the VUV region. As mentioned in Refs. [11], [21], and [22] a high substrate temperature is preferred to prepare the low-absorption MgF₂ film. Obviously, the absorption of MgF₂ film is not the main reason for reducing the reflectivity of Al mirrors prepared with a substrate temperature above 250 °C. On the other hand, the waiting period for the substrate to reach the expected temperature in between depositions of Al layer and MgF₂ layer will cause the fresh Al surface to be oxidized and contaminated due to its interacting with gas released from the chamber wall. The influence of released gas on reflectivity of Al mirror will be discussed and experimentally verified below. Moreover, scatterring strongly depends on the structure and the surface quality of the layer. In order to comprehensively analyze the source of the discrepancy in optical performance of samples prepared at different deposition temperatures, characterizations of crystalline structure and surface topography are necessary.

Traditional TE technique is used to prepare MgF₂ film with a columnar growth and a low material packing density. It provides a large free volume for accommodating the contaminations such as water. Condensed water not only leads the film to enhance the absorption and reduce the reflectance but also causes its lifetime to shorten. Therefore, it is necessary to improve the quality of the Al mirrors under harsh humidity conditions. The measured average reflectance and the relative reflectance degradation as a function of humidity test time of all Al mirrors are presented in Fig. 5. It is found that for all samples the optical performance in VUV range degrade dramatically. The averaged reflectance values of as-deposited Al mirrors at room temperature and 300 °C are as high as 79.0% and 81.8%, and decrease to 66.1% and 78.9% after 36 h of aggressive humidity testing, respectively, indicating 16.4% and 3.5% less reflectivity degradations for the two samples tested in the same environments. Moreover, as indicated in Fig. 5(b), a higher deposition temperature will be favorable for obtaining better humidity stability due to denser microstructure in the MgF₂ layer than lower deposition temperature. For the films degraded by harsh humidity test, ultraviolet ozone and high temperature annealing treatments can be used to restore their performances as reported in Ref. [9]. Taking into consideration the optical property and environmental durability of Al mirrors for the VUV applications, the best deposition temperature of the outermost MgF₂ film is in a range between 250 °C and 300 °C.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. Plots of average reflectance (115 nm–220 nm) (a) and relative average reflectivity degradation (b) versus humidity test time.

As already well documented in Ref. [11], the 3-step method for Al mirror preparation in VUV is based on the following considerations: firstly, Al layer grown with a high temperature is quickly contaminated by forming alumina and tends to rough surface morphology as crystal grain increases. Thus, the substrate at room temperature favors the fabrication of Al film. Secondly, MgF₂ film deposited at high temperature produces a higher-density structure with lower absorption. Last but not least, between depositions of Al film and outermost dielectric layer, a very thin MgF₂ layer is employed to cover the as-deposited Al film as soon as possible in order to protect it from forming oxidation and large crystal grain. Based on the previous experimental results, it is clear that neither the absorption of MgF₂ layer nor scattering in coating system is the main cause for the decrease in optical property of Al mirror prepared at temperature above 250 °C. The remaining reason is that the 5-nm-thick MgF₂ layer deposited by the second step may not completely prevent the fresh Al layer from being contaminated. As presented in Fig. 1, the higher the deposition temperature, the longer the heating time and the higher the pressure in the chamber is. It means that the probability of residual gas pollution increases. To further confirm the observation, the outermost MgF₂ layer with a holding time of 120 min (type M6 in Table 1) is prepared during deposition. In the case, an obvious spectral drop of the mirror is determined and the performance in the VUV spectral range is significantly worse than that of M4 as shown in Fig. 6. Due to the insertion of a constant temperature step in the heating process, the averaged reflectivity decreases from the initial 81.8% to 77.5%, even worse than the M5 reflectivity (79.2%). Therefore, improvement of Al mirrors fabricating process should focus on an optimal trade-off between intrinsic absorption of the MgF₂ film and protection of the metal surface from interacting with contaminants. In future studies, many process parameters must be optimized, including MgF₂ layer thickness in step 2 of the coating process, heating time for the substrate to reach the target temperature, residual gas species and pressure in the chamber, etc.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. Spectral performance of MgF2 over-coated Al mirrors deposited at 300 °C for different holiding times.