2.1

Samples Description

The coating qualification campaign was performed on several samples of Al 6061-T651 in rolled plate form, the same aluminum alloy and forge currently foreseen for Ariel Telescope mirrors and supporting structure.

The samples are shaped as discs of 25mm of diameter and 6mm thick (Figure 1 shows one of the samples being held for visual inspection before the coating run). The samples were procured, polished and cleaned by MediaLario^† before delivery to CILAS for coating. Roughness of the optical surface was measured with a Taylor Hobson CCI White Light Interferometer on areas measuring 1.5mm × 1.5mm and 300 μm × 300 μm. All samples were within the 10 nm RMS specification.

Figure 1.

Picture of one of the aluminum samples during a visual inspection before coating.

The qualification campaign (in fact, a delta-qualification), consisted in a series of environmental and mechanical resistance tests to ensure that the coating will reach the end of life of the instrument without significant performance degradation, following the European Cooperation for Space Standardization (ECSS) Q-ST-70-17C standard.¹⁰ A brief description of the set of tests is reported in Table 1. The qualification program was overall successful.⁶

Table 1.

Summary of specifications of the coating qualification tests performed on the samples.

After the qualification campaing, a subset of the samples was retrieved by the authors and stored in airtight containers in a cleanroom environment (ISO6). The risk of exposure to excessive humidity is further minimized by placing silica gel dessicant bags inside the containers. The samples are being re-examined periodically in a normal laboratory environment.

Table 2 identifies the samples being monitored, coming from two subsequent coating deposition runs: a first test run, performed on April 3rd, 2019, and the actual qualification run, on December 12th, 2019. Both runs were conducted with the nominal coating procedure and produced equivalent results. The set of treatments to which each sample was subjected during the qualification is also reported in the table.

Table 2.

List of samples under test, with coating deposition date and qualification tests performed.

All verification and measurements described in this paper have been performed on all samples listed Table 2, but for the reminder of the treatment we will focus on Samples SN01 and SN12, that have been in storage for a longer period, Sample SN06M since it was already examined in details in a previous work⁷ and Sample SN08M because it was subjected to the entire set of qualification tests.

2.2

Coating Process

The coating process employed by CILAS for the samples is based on physical vapor deposition. The coating platform consists in a large magnetron sputtering chamber. Samples are inserted on a tray sliding beneath a set of cathodes.¹¹ The process is suited to optical substrates up to 2m by 2m of footprint and 0.4m of thickness.¹²

The protected silver coating described in this paper is on average 350 nm thick, with a thickness uniformity measured at 10 %.

The structure of the stack is illustrated in Figure 2 and consists of at least three layers: a NiCr adhesion layer of less than 10 nm of thickness, the reflecting silver layer, and a dielectric capping and protection layer. An additional intermediate adhesion layer may be present between the silver and the capping layer. The actual layers thicknesses and composition of the coating cannot be disclosed due to business confidentiality.

Figure 2.

Indicative structure of the multilayer coating stack. The total coating tickness is approximately 350 nm.

3.1

Relative Reflectance Measurements

Immediately after coating deposition, the reflectance of all samples was measured by the manufacturer with a Perkin-Elmer Lambda 950 spectrophotometer with a reflectometry accessory, in the waveband 500 nm to 2500 nm. Accuracy, as reported in the instrument datasheet, is ±0.6% from 500 nm to 890 nm and ±1 % above 890 nm.

Subsequent relative reflectance measurements were taken at the Institute for Photonics and Nanotechnologies of the National Research Council (CNR–IFN) in Padova with a custom built setup. Reproducibility of the measurements in the worst case has been determined to be better than ±1.2% over the 500 nm to 900 nm wavelength range, and better than ±3.8% over the 400 nm to 1000 nm wavelength range.

In order to cross-calibrate the two setups, which appear to have a measurement bias between each other (see Figure 5 in the Results section), we used one of the earliest measurement of Sample SN01 as reference point, since it presented the shortest temporal gap between measurements.

3.2

Visual Inspection

The ECSS standard suggests following Annex C of ISO 9211-4:2022 for Visual Inspection (VI), that mandates the use of two cool white 15W lamps positioned directly above the sample, and to look at the sample against a black matte background at a distance of ≤45cm and at a near grazing angle.¹³ The use of optical micrographs is suggested only in case of suspected degradation, to further qualify it. In our case it was employed to look for signs of oxidation, as explained in Section 4.1.

Darkfield imaging using a compact digital camera (Canon IXUS 220 HS) and a custom built LED lighting setup was also employed to highlight the presence of light scattering defects.

3.3

Atomic Force Microscopy

Besides optical imaging techniques, we employed an Atomic Force Microscope (AFM) for a qualitative analysis of surface topography and to measure surface roughness.

AFM scans were taken with a Park System^‡ XE-Series 70 microscope in non-contact mode and processed with the Gwyddion open source software¹⁴ (the processing pipeline consisted in removal of low spacial frequencies by fitting and subtraction of an x, y third order polynomial surface, rows alignment using “median” as statistic and “scars removal”).

4.1

Visual Inspection

The optical surface of all samples appeared visually unaltered after the storage period, as exemplified in Figure 3: no discernible signs of coating degradation, such as cracks, blistering or change in color/iridescence, nor other visible signs of delamination could be spotted. In particular, we could not detect signs of oxidation developing from surface grains, as described for example by Folgner et al. in their studies of protected silver coatings exposed to mixed flowing gas.^9,15

Figure 3.

Couples of photographs of each sample immediately after coating (left), and after the storage period (right). Note that the orientation of the sample is not consistent in each couple.

Apart from scratches and occasional dust particles, the most prominent features on the optical surface continued to be the glue residues from adhesion tests, especially evident on sample SN01.

It is worth noting that the uncoated back and lateral surfaces of the samples do exhibit a slight brownish coloration and faint whitish areas, compatible with the oxidation of bare aluminum, so the level of exposure to pollutants seems to be at least sufficient to cause slight degradation of the substrate.

Darkfield photographs also show a mostly uniformly dark optical surface, indicative of low scattering (Figure 4).

Figure 4.

Darkfield photographs of the four samples after the storage period. Orientation of the samples is consistent with each right image of Figure 3.

4.2

Reflectance Measurements

Figure 5 shows the results of the reflectance measurements performed on the samples at different points in time. Samples reflectance was measured after coating and after qualification tests by the manufacturer (solid gray line), who found no degradation in performance.⁶ One of the samples (SN01) was also measured by the authors early on during the qualification campaign, so this measurement could be used as reference for the cross-calibration of the two setups (dashed gray line).

Figure 5.

Reflectance measurements of four coated samples performed with two different setups, by the manufacturer and by the authors. One comparison measurement of sample SN1 (dashed gray line) was performed sufficiently close in time to the manufacturer’s measurement to be useful to assess the cross-calibration of the two setups, which appear to have a bias.

Eventually two comprehensive measurement campaigns could be performed six months apart (in October 2021 and April 2022, orange and blue lines in the plots).

Considering the difference between the initial measurement of the in-house setup and the manufacturer’s setup, and noting the estimated repeatability error, the results do not highlight a significant change in reflectance of the samples.

More precise measurements may be performed in the future to further confirm this preliminary result.

4.3

Atomic Force Microscopy

Two sets of AFM scans of representative areas of samples SN01 and SN06M are presented in Figures 6 and 7. For each set, the first scan (on top) was performed at the beginning of the storage period, after the sample underwent the qualification tests, and the second one (at the bottom), in August 2022.

Figure 6.

AFM scans (left) of sample SN01 before (top) and after (bottom) the storage period. For each scan, two orthogonal line profiles are also provided (plots on the right). The two scanned areas do not necessarily coincide.

Figure 7.

AFM scans (left) of sample SN06M before (top) and after (bottom) the storage period. For each scan, two line profiles are also provided (plots on the right). The two scanned areas do not necessarily coincide.

As discussed in Section 3.3, the scans were performed to provide a qualitative assessment of surface morphology variations and to measure surface roughness. Since the time in storage did not produce any visible signs of surface degradation nor delamination, the sampling location for the AFM was chosen to be reasonably representative of the central area, where reflectance was also measured, without aiming at specific surface blemishes or scratches. Please also note that measurements of the same sample do not image the exact same portion of surface.

A comparison of the AFM images does not indicate the appearance of new topological structures of relevance: most features are attributable to scratches and dents that were equally present before storage. The white areas in relief on SN01 are likely the residues of tape adhesive from the adhesion tests. These were in fact visible upon careful examination of the surface area affected by the test. Anecdotally, AFM scans of sample SN01 did result in frequent tip pollution that required replacement, possibly because of the residues.

RMS roughness measurements also do not appear to change significantly before and after the cycles.