Protective Magnetron Sputtering Physical Vapor Deposition Coatings for Space Application

Abstract

In this study, the use of Cr/CrN+CrCN/Cr-C:H, Cr/W-C:H, and Cr/CrN+Ag/Cr-C:H coatings deposited on copper beryllium were investigated. These protective coatings were prepared using the Magnetron Sputtering Physical Vapor Deposition (MSPVD) method. The tests were carried out in order to qualify the outer DLC (Diamond-Like Carbon) layers for use as the protective function and for regulating the thermo-optical properties. The objective of this study was to compare the properties of chromium and chromium nitride-based coatings. The microstructure, architecture, and chemical composition were studied using scanning electron microscopy (SEM), Photo Diode BackScattered Electrons (PDBS), and X-ray dispersion spectroscopy (EDX). The adhesion was evaluated using a scratch test and a peel and pull-off method. The level of protection against the cold welding effect was tested. Thermo-optical, microhardness, and surface electric resistivity tests were performed. It was found that in cases where increased resistance to cold welding is required, DLC2 and DLC3 proved to be the best solutions. An example of such an application is tubular boom antennas, which are stored in a rolled-up form until deployed in space. They are susceptible to cold welding due to vibration during rocket launch and subsequent exposure to high vacuum.

1. Introduction

High demands are placed on devices sent into space. They are related to economics and the requirement to minimize the mass of all elements, which reduces the cost of sending a satellite to a low orbit or a probe on a long journey to other planets. For the same reason, the reliability of solutions is also required [1,2]. Based on the tubular boom technology, various deployable structures and mechanisms can be built, providing a compact size and minimal weight when stowed and a significant size and stiffness when deployed [3,4]. However, the harsh space environment, especially vacuum and a wide temperature range combined with vibration during rocket launch, gives strict requirements for all space mechanisms. One of the best materials for tubular booms is copper beryllium (CuBe2), which is easy to shape in annealed conditions and which, by heat treatment, gains high strength and stiffness. Copper-beryllium alloy is one of the more interesting copper-based alloys. It is a recently developed alloy containing about 97% copper, 2% beryllium, and enough nickel to increase elongation. Its feature is that heat treatment improves its physical properties, increasing the tensile strength from 70,000 psi in the annealed condition to 200,000 psi in the heat-treated condition [5,6]. However, as most metals have to be protected against cold welding and wear due to the other mechanisms like thermal limitations, booms have to have proper thermo-optical properties. All of these requirements can be met by the application of hard coating with a suitable ratio of absorptivity (α) and emissivity (ε). Additionally, the selected coating has to be applicable on a thin CuBe2 tape, withstand thermal cycling in a wide temperature range, and be electrically conductive at least at the dissipative level. Advanced surface modification techniques allow for improving the material characteristics used in the aerospace industry. Recently, Cr-based multilayer coatings were studied for their better properties than a single-layer coating. Coatings based on chromium compounds are particularly valuable in space applications due to their heat and creep resistance, where parts are exposed to high and low temperatures. The proper combination of materials and design strategy results in higher hardness, wear, and oxidation resistance [7,8]. The most interesting methods for the space industry are PVD (physical vapour deposition) coatings with DLC layers [9,10,11,12,13,14,15,16]. PVD CrN and CrCN coatings can be improved by introducing an anti-wear DLC coating using the RF PACVD method. Within the DLC layers, a-C or ta-C can be distinguished. The differences consist mainly of the different contents of sp3 bonds. According to Vetter (2014), ta-C carbon has 50% sp3 bonds and a hardness of at least 50 GPa [17]. Coating development and testing were focused on fulfilling requirements and defining properties; hence, the statistical approach was limited. The overall approach to testing was based on predicted space environmental conditions and booms’ functionality, i.e., its manufacturing and loads during the launch. The coatings’ properties were compared between the initial state and after heat treatments and thermal cycling.

The tests were carried out in order to qualify the DLC (Diamond-Like Carbon) coating for use in the protective function and regulating the thermo-optical properties on the RWI (Radio Wave Instrument, Astronika, Warsaw, Poland) antennas. It is one of the measuring instruments developed by the Astronika company, who, together with the entire JUICE (Jupiter Icy moon Explorer) spacecraft in April 2023, started to research Jupiter’s Galilean moons and their surroundings. As part of the research, it was necessary to prove that both the selected coating and the antennas covered with it, made in the “tubular boom” technology, would survive the demanding launch of the rocket and the hostile space environment. The primary concern with cold welding in deployable tubular boom systems (Figure 1) is the potential for the coils of the boom to bond together, which would prevent successful deployment. In cases where cold welding occurs but does not prevent the system from opening, there may be surface damage, which does not pose further risks for standard applications. These systems (each of which extends to 2.5 m) must be resistant to wear due to frequent deployment and retraction in the environmental conditions in which they have to operate: very high vacuum, launch loads, vibrations, and thermal fluctuation.

This study considered various multilayer coatings based on chromium and chromium nitrides deposited on a copper beryllium substrate. There are no studies in the literature regarding coatings produced on the CuBe2 alloy. The architecture of Cr/CrN+CrCN/Cr-C:H coatings was due to the increased coating adhesion to the substrate due to the chromium underlayer. A middle CrN-CrCN composite provides cracking resistivity, while the utmost DLC layer ensures a low friction coefficient and good wear resistance [9,10]. In the case of the Cr/W-C:H coating, the technical underlayer of chromium increases coating adhesion to the substrate, while the utmost DLC layer ensures a low friction coefficient and good wear resistance [9,10,11,12]. In the last case, the technical underlayer of chromium increases coating adhesion to the substrate, and the middle CrN-Ag composite provides cracking resistivity and excellent electrical conductivity. At the same time, the utmost DLC layer ensures a low friction coefficient and good wear resistance [9,10].

2. Materials and Methods

2.1. Substrate and Layer Deposition

All coatings were deposited using Magnetron Sputtering Physical Vapor Deposition (MSPVD). The MSPVD chamber was custom-made for a DLC supplier based on some industrial and laboratory items; hence, no brand or type of chamber was available. The metallic elements were introduced as targets sputtered with the Ar⁺ ions. Other elements were introduced to the chamber as gaseous compounds, e.g., CH₄ as a source of carbon for DLC coatings. The following coatings were produced: (1) Cr/CrN+CrCN/Cr-C:H (marked as DLC1), (2) Cr/W-C:H (marked as DLC2), and (3) Cr/CrN+Ag/Cr-C:H (marked as DLC3). The use of 3 different types of coatings was intended to select the coating with the best properties for the space probe components. The bottom technical layer of pure chromium is intended to improve adhesion to the substrate. The middle composite coating consists of alternating chromium nitride and silver layers, intended to increase fracture resistance. In contrast, the top layer of hydrogenated and chromium-doped DLC should provide a low coefficient of friction and abrasion resistance.

A thin (80 μm) CuBe2 tape in 1/2 H condition was used as a substrate (Berylco B25 alloy tapes, NGK BERYLCO, Salford, UK). Before coating application, the tapes were chemically cleaned and then ion-etched with the Ar⁺ ions. Cleaning consisted of washing in alkaline detergents with ultrasound support (40 °C, 1 min) and rinsing several times in deionized water with ultrasound support (40 °C, 1 min). The tapes, measuring 40 mm × 1500 mm, were mounted on a rotating cylinder, causing the tapes to be cyclically and equally exposed to the metal targets.

2.2. Coating Characterization

Coatings were characterized in terms of basic properties and meeting the requirements given by the harsh space environment. Hence, the following aspects were tested: chemical composition (SEM), surface roughness, surface electrical resistivity, cold welding, and adhesion to the substrate. Moreover, the influence of thermal cycling and substrate tempering on the above parameters was tested.

2.2.1. Composition (FIB + SEM + EDS)

Due to the very low thickness of coatings, cross-sections were prepared using FIB (focus ion beam). To protect the investigated coatings, a thin layer of tungsten was sputtered on top of them. Then, the cross-sections were observed using SEM (Scanning Electron Microscope; 5500, HITACHI, Tokyo, Japan) in SE (secondary electrons) and PDBSE (Photo Diode BackScattered Electrons) modes. The chemical composition was checked with EDS (energy-dispersive spectroscopy).

2.2.2. Surface Roughness

Surface roughness was measured with an optical profilometer Wyko NT9300 (Veeco, Plainview, NY, USA). To check how coating application affects the surface roughness, samples of the bare substrate and each coating were tested. All samples were prepared in the same way—they were ground with nonwoven pads. This allowed for the removal of unidirectional scratch patterns present on the raw tapes. Additionally, one set of polished samples without coating was prepared.

2.2.3. Surface Electric Resistivity

Surface resistivity was measured through two linear, copper electrodes pressed to the samples by a constant force (10 N) (Figure 2). This configuration allowed for measuring surface resistivity directly in Ohm/sq. Measurements were taken with a calibrated multimeter.

Standard methods used for surface resistivity measurement are not suitable for this application since they require larger samples (4 inch square) and the application of 500 V for a prolonged time, which will dissipate a significant amount of heat with the expected low resistance.

2.2.4. Thermo-Optical Properties

The total reflectance spectra were in a solar wavelength range of 300–2500 nm and an infrared wavelength range of 2.5–20 mm. The solar absorptance (alpha) was calculated according to the ISO9845 standard [18] using the global solar spectrum for AM1.5. Solar reflectance was calculated as 1-alpha. The emittance value was calculated for blackbody radiation at room temperature.

Differences in the spectrum between AM1.5 (surface of the earth) and AM0 (space) were mainly in the near-IR and UV ranges. However, the difference was negligible between the coatings.

Measurements were taken using a Lambda 900 spectrophotometer (PerkinElmer, Waltham, MA, USA) with a 15 cm integrating sphere and Bruker FTIR (Bruker, Billerica, MA, USA) with a 7 cm integrating sphere.

2.2.5. Microhardness

The hardness of the coatings was measured using a nano-hardness tester (Hysitron Triboindenter, Bruker, Billerica, MA, USA). Due to the very low thickness, the load value was also very low (10 μN) to minimize the influence of the substrate’s hardness in the final result. The loading time was 10 s.

2.2.6. Cold Welding

The level of protection against the cold welding effect was tested in a material pair: bare CuBe2 pin against flat CuBe2 samples with tested coating. As the reference, two sets of non-coated samples were used— the as-delivered CuBe2 tape and a polished one. Each sample was glued to a steel disk to ensure proper fixation and a rigid form. All samples were subjected to the same heat treatment—tempering in an inert atmosphere.

The test was performed in accordance with the STM-279 standard [19]. The main test parameters were as follows: Hertzian pressure, 114 MPa (1 N at the pin with radius 25 mm); Fretting freq., 210 Hz, 50 μm stroke; adhesion force, measured every 10 s; 5000 cycles.

Hertzian contact pressure was defined based on the application of the tubular boom antennas developed by Astronika [4]. Two factors were determined in the test: maximal adhesion force and maximal lifetime, defined as the number of cycles before adhesion achieved 200 mN.

2.2.7. Scratch Test

The adhesion test was performed using the “Scratch test” method according to ASTM Standard G171 [20], using a CSM Revetest equipment (CSM Instruments, Peseux, Switzerland). Scratch was made with a diamond Rockwell intender (radius of 0.8 mm), progressive load in a range of 1 ÷ 100 N, with a speed of 50 N/min and length of 20 mm.

At each step of the investigation, three scratches were made per sample for statistics purposes. Critical values of load Lc1 and Lc2 were measured, where

• Lc1 is the load where the first initial single spelling of coating was observed (adhesive type of failure);
• Lc2 is the load where cracks inside the trace were observed (cohesive type of failure).

2.2.8. Peel and Pull Off

The test was performed on an MTS QTest 10 device (MTS, Eden Prairie, MN, USA) according to ECSS-Q-ST-70-13C [21]; all tests were started using the strongest tape 3M8915 (6.7 N/cm). During the test, force–displacement data were collected; before and after each test, the surface of the samples was observed under a magnification of 7× using a flash magnifier, and for each set of samples, a result sheet was prepared and filed. Each coating after each treatment was represented by 3 samples.

3. Results

3.1. Coating Characterization

3.1.1. Composition (FIB + SEM + EDS)

Cross-sectional SEM observations (Figure 3, Figure 4 and Figure 5) confirmed the expected coating composition. In most cases, cross-sections showed uniform thickness (

Table 1. Coatings’ thickness.

Coating	DLC 1	DLC 2	DLC 3
Thickness [µm]	1.04	0.6	0.87
SD	0.07	0.07	0.09

) and material distribution along the substrate’s surface. The thickness was determined based on the cross-sectional images (Figure 3, Figure 4 and Figure 5).

For all three DLC coatings, the carbon layer is deposited at the top of the interlayers. The chromium bottom layer allow for the increase in adhesion, and then others are introduced to enhance the desirable properties. DLC1 has a four-layer CrN-CrCN composite, which is poorly distinguished in the SE mode, as they differ only in the proportion of light elements (Figure 3). The DLC2 has a gradual transition from pure chrome, through pure tungsten, to a tungsten-doped carbon coating (Figure 4). The DLC3 between the chromium and carbon layers has a clearly visible layered CrN-Ag composition (Figure 5).

The results of the chemical composition test on the cross-section of the DLC1 coating are presented in

Table 2. EDS results from a selected measurement point on the DLC1-3 coatings.

Material	Atomic Concentration % of the Particular Elements
	Cr	Cu	W	Ag	C	O	P	N	Al
DLC1	32.3	8.5	—	—	33.8	—	—	25.1	0.3
DLC2	13.9	21.7	5.6	0.5	55.8	2.2	0.3	—	—
DLC3	46.6	12.9	—	8.1	28.9	3	—	—	0.6

. The chemical composition is consistent with the expected one. Aluminium in the composition is an artefact resulting from the measurement method. The visible copper peak comes from the substrate.

The chemical composition of the DLC2 coating is as expected—the presence of chromium from the coating, copper from the substrate, and tungsten (W-C:H) is visible (Table 2). The phosphorus and oxygen in the composition are probably artifacts resulting from the measurement method. Silver, on the other hand, may be both an artifact and a result of the contamination of one of the targets during the production of the DLC3 coating, which contains this metal.

The chemical composition of the selected point on the DLC3 coating’s cross-section shows chromium, silver, and copper from the substrate, which is consistent with expectations (Table 2). The aluminium and oxygen in the composition are most likely an artifact resulting from the measurement method.

3.1.2. Surface Roughness

The surface observations, made in the PDBSE mode, show minor differences between the various types of coatings (Figure 6). The initial state of the substrate is primarily responsible for surface topography. The only significant difference is the presence of small (1–3 µm) spherical chromium particles in the case of DLC coatings. They were formed during the application of a pure chromium technical underlayer. They show a reduced attachment to the substrate; however, even their removal leaves a thin but present chromium coating—the substrate is not fully exposed. This was confirmed by EDS, and there is still a chromium reading at the location of the spherical holes (Figure 7,

Table 3. Composition measurement results for selected points (4–6) on the DLC1 coating.

	Atomic Concentration of the Particular Elements [%] ± Error
Element	Measuring Point 4	Measuring Point 5	Measuring Point 6
C K	51.8 ± 1.8	10.3 ± 1.0	8.6 ± 1.2
N K	0 ± 0	8.7 ± 2.6	0 ± 0
Al K	0 ± 0	0.3 ± 0.1	0 ± 0
Si K	0 ± 0	0 ± 0	0.4 ± 0.1
Cr K	46.3 ± 0.4	9.0 ± 0.3	30.4 ± 0.5
Cu K	1.9 ± 0.3	71.7 ± 1.0	60.6 ± 1.0

).

In the case of other coatings, the surface morphology is primarily a reflection of the substrate; there are no noticeable differences or foreign objects.

3.1.3. Surface Resistance (Electrical) Ohm/sq

Despite small changes in surface resistivity after thermal cycling, all coatings show very good electrical conductivity (Figure 8). The good surface conductivity, due to the use of a radio antenna, must be better than in standard solutions for exposed satellite surfaces; expected R < 1385 Ω/sq (R—surface resistivity). The surface resistivity value requirement is based on internal analysis, which considers the expected radiation environment during the JUICE mission. The following level of resistivity allows avoiding negative effects related to surface charging of the RWI antennas. Even the highest value (2.8 Ohm/sq) provides sufficient discharging for the space environment. It is worth noting that low resistivity can be the effect of good conductivity through coatings’ thickness and metallic substrate instead of through the bulk of the coating. Comparing the coatings to each other, the best results are given by DLC1, both in the initial state and after heat treatment. In turn, the advantage of DLC3 over DLC2 without treatment may be the effect of the presence of silver in the layers. Ag is a very good conductor of electric current. In most cases, the highest resistances are recorded after thermal cycles and cause a slight decrease in conductivity. This may be due to changes in the structure of the coating itself, which affected the electrical properties. However, it was not decided to take action to determine the reasons for these changes in detail.

3.1.4. Thermo-Optical Properties

The most important parameter—the ratio between the absorptivity and emissivity (α/ε)—can be varied in a wide range. Compared to the bare substrate on one side, a reduction at a 40% level can be achieved by applying DLC3 (Figure 9). Additionally, it is worth pointing out that polishing itself can slightly decrease the a/e ratio (by 10%).

After thermal cycling, absorptivity did not changed significantly; however, for almost all samples, aside from DLS, emissivity changed up to four times for the bare substrate (Figure 9). Due to emissivity changes, the α/ε ratio significantly changed as well. The smallest change was observed for DLC3—only an 11% increase in the α/ε ratio—and the following DLC coatings had an ~30% change in α/ε.

3.1.5. Microhardness

Most of the coatings had similar microhardness to the substrate material (in the range of ±15%) (Figure 10). The lowest hardness was exhibited by DLC2, which was ~20% lower than that of the bare substrate. On the other hand, the highest hardness was measured for DLC1—it was two times higher than the substrate. The reduced young modulus only occurred for DLC1, similar to the substrate. For other coatings, E_r was significantly lower. The lowest values were for DLC2 and DLC3 (about 10% of the substrate). The DLC coatings constituted only the top layer of the composites referred to as DLC1 and DLC3. The mechanical properties, such as hardness, are largely due to the layered composite: CrCN-CrN and CrN-Ag, respectively. The composite zone was approximately 0.6–0.65 μm thick in both cases. The effect of replacing the hard CrCN layer with soft silver was, therefore, visible—both measured parameters were significantly lower in the case of DLC3. In addition, DLC2 was devoid of such an extensive internal structure. A carbon coating doped with this element was applied to a thin tungsten layer. DLC2 was also the thinnest of the three DLCs tested. Therefore, the results of microhardness measurements and the determined reduced Young’s modulus largely reflect the properties of the carbon coating itself.

3.1.6. Cold Welding

The best protection against cold welding and fretting was for DLC2 (Figure 11). This coating showed the lowest adhesion in contact with bare CuBe2 and a very high lifetime (329 cycles). The good adhesion was achieved in the case of the DLC 2 and DLC3 coatings. There was also a significant increase in the adhesion force caused by polishing—comparing polished and ground samples, where the polished one had ~75% higher adhesion. The main protective factors are the combination of abrasion resistance features and differences in the structure of the crystal lattice and the types of bonds of the cooperating materials. For this reason, ceramic or amorphous coatings give better results in cooperation with metals. It is also worth noting that the best results in the cold welding test can be associated with microhardness measurements or, more precisely, with the reduced Young’s modulus. The longest service life was shown by coatings with the lowest reduced Young’s modulus.

3.1.7. Scratch Test

In the initial state, the best results were achieved by DLC3 (42 N for LC1 and 90 N for LC2) and DLC1 (37 N for LC1, LC2 not observed) (

Table 4. Scratch test results—LC1 and LC2 forces for all coatings (if no value is given, the damage of the coating was not observed in the tested force range (up to 100 N)).

Sample	LC1 [N]			LC2 [N]
	Initial	Tempered	Thermal Cycled	Initial	Tempered	Thermal Cycled
DLC1	36.9	—	69.2	—	—	—
DLC2	25.1	53.1	76.4	60.0	—	—
DLC3	42.2	88.9	85.4	89.6	—	—

). In general, both tempering and thermal cycling caused a significant increase in forces LC1 and LC2. After tempering, LC1 increased at least two times (DLC2 and DLC3), and for DLC1, it is not defined because LC1 was above the tested limit. The coatings did not show cracks in the tested range.

After thermal cycling, LC1 increased for all coatings: at least two times for DLC1 and DLC2, and three times for DLC3. LC2 for all coatings was above the test limit.

The scratch test showed that heat treatment benefits the coatings in terms of increased resistance to cracking and adhesion to the substrate. In all tested coatings, the LC1 force, describing the force necessary to create the first noticeable cracks, increased at least twice, and in the case of the DLC1 coatings, it exceeded the measurement range (100 N). In addition, in the case of most coatings, after heat treatment, delamination in the scratch test no longer occurred. The probable reason for this behaviour of the coatings is a combination of two effects. On the one hand, the substrate is strengthened, increasing its stiffness. Smaller deformations provide better support for the coating. The second factor may be the removal of residual stresses and relaxing both the coating itself and the bond between it and the substrate. Another explanation may be the occurrence of compressive stresses in the coating, improving its cracking resistance.

3.1.8. Peel- and Pull-Off

During the test, none of the coatings showed any sign of delamination or peeling off. This shows that PVD coatings represent much higher surface adhesion in comparison to other types of adhesive coatings like paints or sprayed coatings.

Moreover, the results obtained from force measurements showed that the 3M8915 tape used in the test had adhesion values higher than those provided in the datasheet. The maximum pull-off—unloading force—in all tests was higher than 50 N, which is almost 100% higher adhesion of tape than the predicted 26.8 N on the measured surface contact of 4 cm².

Both the adhesion between the coating and the substrate, as well as the coating’s cohesion, exceeded the stress generated in this test despite the tape-to-coating adhesion being higher than required. No delamination or other damage were visible (Figure 12).

4. Summary and Conclusions

The tested coatings allow for the shaping of the thermo-optical properties of antennas in a wide range of α/ε ratio, which provides an opportunity to meet a wide range of mission requirements. In general, neither the heat treatment associated with the production of antennas nor the thermal cycles have negative effects on the mechanical properties of the coatings. This was visible in the case of the scratch test method, in which the coatings, after heat treatment, showed greater resistance to cracking and delamination. Moreover, they had improved resistance to mechanical factors, which was manifested in the increase in the critical forces LC1 and LC2 in the scratch test. However, thermal cycles changed, to a noticeable degree, the thermo-optical properties of both the base material and most of the coatings, except for DLC coatings. An additional advantage of all tested coatings is the high electrical conductivity, which, despite slight fluctuations, is sufficient to meet the requirements related to the discharge in space. All tested coatings showed high adhesion to the substrate, which was confirmed in the peel-and-pull-off test.

In cases where increased resistance to cold welding is required, DLC2 and DLC3 are the best solutions. An example of such an application is tubular boom antennas, which are stored in a rolled-up form until they are deployed in space, and which are susceptible to cold welding due to vibrations during rocket launch and subsequent exposure to high vacuum.