Indentation and Scratch Testing of Adhesion and Mechanical Properties of Multi-Layer Coatings for Concentrated Solar Power Systems

One of the most viable and already used sources of renewable energy are concentrating solar power systems (CSP) that convert solar thermal energy into electrical power [1-4]. Mirrors are used by parabolic trough solar collector (PTSC) technology to redirect and focus sunlight.

This sunlight heats up the transport fluid in the tube, which is subsequently used for producing electricity production (Figure 1). Over the last several decades, the PTSC technology has been effectively used around the world and it is now become the most advanced of all CSP designs in terms of technical and scientific research.

Although existing systems are designed to operate at a temperature of 400 °C, new systems are now being made with operating temperature of up to 600 °C. The new generation of systems should boost the performance by as much as 5 to 10% so that the technology continues to remain competitive in the green technology market. However, new degradation processes can occur that can reduce the system efficiency [5] at these elevated temperatures. CSP systems can provide a viable solution if they have a lifetime of about 20 to 25 years and therefore, understanding the degradation processes is very important to further develop these systems.

Schematic of the parabolic trough solar collector [4] and the collector tube.

Figure 1. Schematic of the parabolic trough solar collector [4] and the collector tube.

In order to ensure durability and efficiency of the PTSC technology, the absorber coating which is deposited on stainless steel tubes (where the working liquid circulates) represents an important factor. This coating warrants several key requirements which can be met only by using a multilayer structure [6-8].

Table 1 shows the composition of such multilayer absorber coating. While thorough investigation of the thermal stability and optical performance of the absorber coatings has already been made, very limited knowledge is available regarding the mechanical properties of these multilayer absorber coatings. Deterioration of optical properties can be indicated by changes in the elastic modulus and hardness of the coating. Such deterioration is attributed to degradation through thermal cycling day-night, which, in turn, can cause adhesive failure of the coating.

This article focuses on the characterization of adhesion and mechanical properties of standard CSP absorber coatings. The study was part of the FP7 NECSO project whose aim was to establish a standardized approach to evaluate the lifetime of the CSP absorber coatings through accelerated aging tests. Nanoscratch testing was performed to characterize the adhesion of the absorber coating, while low load nanoindentation was applied to characterize hardness and elastic modulus. All tests were performed on samples before the aging tests and after the accelerated aging tests.

The collector tube with absorber coating.

Figure 2. The collector tube with absorber coating.

Materials and Experimental Procedures

Test samples were provided in the form of cylinders that had an 80 mm internal diameter and a 2 mm wall thickness with the absorber coating applied on the outer side. The substrate was developed from AISI321 steel. IK4 Tekniker (Eibar, Spain) performed the deposition through the PVD magnetron sputtering method on long tubes measuring 4 m. These tubes are used in real CSP applications. Table 1 summarizes the structure of the coating. Next, the tubes were cut to about 20 cm long cylinders for the nanomechanical characterization and accelerated aging tests.

Table 1. Composition and optical properties of the tested absorber coating deposited by PVD magnetron sputtering at Tekniker.

Layer Composition Thickness [nm]
Anti reflective layer (AR) SiO2 75
Absorber layer (AB) Mo/Al2O3 60
Infrared reflective (IRR) Mo 200
Antidiffusion layer (AD) Al2O3 150
Substrate (AISI321 Steel)

For accelerated ageing tests, selected tubes were used in an exclusively developed chamber at 500 °C and 450 °C with 0.1 mbar of oxygen. It took one day to complete the tests in all cases. Since oxidation is one of the most critical factors in the deterioration of solar selective coatings, the level of oxygen is considered to be very important.

Nanoindentation

Anton Paar Ultra Nanoindentation Tester (UNHT) equipped with Berkovich diamond indenter was used to perform the nanoindentation tests. Using the ISO 14577 standard [9], data was analyzed to acquire the coating’s hardness (HIT) and elastic modulus (EIT). The maximum load (0.02 mN and 3.0 mN) was then chosen to characterize the entire coating or the thin top layers (AB and AR). The 0.02 mN and 3 mN indentations resulted in ~12 nm indentation depths and ~140 nm indentation depth, respectively. In either case, it is possible to prevent the effect of the substrate and underlying layers [10,11].

The indentation parameters were as follows: loading time 10 seconds, hold time at maximum force 5 seconds, unloading time 10 seconds.

Indentation with the UNHT on the tubes placed in the sample holder developed for the NECSO project.

Figure 3. Indentation with the UNHT on the tubes placed in the sample holder developed for the NECSO project.

Nanoscratch Testing

Using the Anton Paar Nano Scratch Tester (NST), scratch tests with length of 1.0 mm with gradually increasing normal load (1000 N/minute) from 2 mN to 500 mN were carried out by diamond sphero-conical indenter with a tip radius of 10 µm. The scratch tests involved a pre-scan and post-scan procedure so that the effect of sample inclination and/or surface profile can be eliminated. This method helps obtain the true penetration and residual depth of the scratch. Once the scratch test was completed, the sample was viewed with the connected optical microscope and a Panorama image was subsequently recorded. In addition, the Panorama feature included in the Scratch software allows the scratch image to be synchronized with the recorded signals (normal force, friction force, depth, etc.).

Nanoscratch tests on the tubes placed in specially developed sample holder.

Figure 4. Nanoscratch tests on the tubes placed in specially developed sample holder.

Critical loads (Lc) characterize the adhesion (and cohesion) of the coatings. Critical load Lc is the load that corresponds to the first appearance of a specific type of failure, for example delamination or cracks. A number of critical loads can be typically determined: the initial cracks in the coating (mainly cohesive damage), partial delamination (adhesive failure) and then complete delamination (catastrophic adhesive failure).

Optical Properties

For optimum performance of CSP absorber coatings, the absorbance of the coatings should remain over 90% while the emissivity should be below 10% throughout the lifetime of the coatings. Using a hemispherical reflectance spectrometer (Perkin Elmer Lambda 950), the optical properties of the CSP absorber coatings before and after the accelerated aging tests were determined and their relation to changes in mechanical properties was assessed.

Nanoindentation Results

Low load (0.02 mN) indentations: Figure 5a shows typical load-displacement curves for indentations with 0.02 mN maximum load on unaged coating. During these indentations, less than 12 nm was the maximum depths reached but this value is less than 10% of the thickness of the two upper most layers (AR+AB, 60+75=135 nm). In spite of such low indentation depths and loads, the resultant curves were repeatable with low level of noise. Moreover, the irregularities on the load-displacement curves most probably originated from the non-homogeneous structure of the two top layers containing pores and voids.

Superposition of three curves for each aging condition and three curves for the unaged sample are illustrated in Figure 5b. While some variations are seen in the curves and the resulting mechanical properties, a clear relationship between the aging conditions and mechanical properties of the samples cannot be established. Nevertheless, Table 2 shows that both elastic modulus and hardness of the AB and AR layers slightly reduced following the aging tests. Elastic modulus (EIT) decreased more considerably than in hardness (HIT). Such changes can be due to structural changes caused by high temperature as well as by oxidation. Nevertheless, the changes in HIT and EIT continue to remain relatively low.

Typical indentation (load-displacement) curves on individual antireflective (AR) and absorber (AB) layers (a), superposition of indentation curves with 0.02 mN maximum load on unaged and aged samples (b).

Typical indentation (load-displacement) curves on individual antireflective (AR) and absorber (AB) layers (a), superposition of indentation curves with 0.02 mN maximum load on unaged and aged samples (b).

Figure 5. Typical indentation (load-displacement) curves on individual antireflective (AR) and absorber (AB) layers (a), superposition of indentation curves with 0.02 mN maximum load on unaged and aged samples (b).

Table 2. Hardness HIT, elastic modulus EIT, maximum depth hm and elastic work to total work of indentation ratio (ηIT) of the two top layers (AR+AB) obtained with 0.02 mN maximum load indentation.

unaged 450 °C 500 °C
Hit
[MPa]
Average
St. Dev.
3017.7
497.8
2531.7
278.0
2872.1
546.0
Eit
[GPa]
Average
St. Dev.
70.4
13.1
57.6
11.1
53.0
10.7
hm
[nm]
Average
St. Dev.
11.9
1.3
14.0
1.2
13.4
2,1
nit
[%]
Average
St. Dev.
44.6
8.1
52.5
9.6
57.3
10.4

Comparison of hardness and elastic modulus of the two upper layers on unaged and aged samples (indentation with 0.02 mN maximum load).

Comparison of hardness and elastic modulus of the two upper layers on unaged and aged samples (indentation with 0.02 mN maximum load).

Figure 6. Comparison of hardness and elastic modulus of the two upper layers on unaged and aged samples (indentation with 0.02 mN maximum load).

High load (3.0 mN) indentations: To characterize the entire absorber coating, indentations with 3.0 mN maximum load were carried out. Approximately 140 nm was the maximum indentation depth at the 3.0 mN load, thereby fully penetrating the AB and AR layers as well as testing the underlying AD and IRR layers (~490 nm is the total thickness of the multilayer coating). Therefore, the mechanical properties of the whole absorber coating were determined while the effect of the substrate can still be taken as negligible. Superposition of three indentation curves on the samples in both aged and unaged states are shown in Figure 7.

Typical indentation (load-displacement) curves obtained with 3 mN maximum load indentations (a) and superposition of three indentation curves for unaged and aged samples (b).

Typical indentation (load-displacement) curves obtained with 3 mN maximum load indentations (a) and superposition of three indentation curves for unaged and aged samples (b).

Figure 7. Typical indentation (load-displacement) curves obtained with 3 mN maximum load indentations (a) and superposition of three indentation curves for unaged and aged samples (b).

Low noise and good repeatability were observed for all indentation load-displacement curves. A slightly smaller indentation depth is shown by the indentations on the aged sample, indicating higher hardness. Table 3 shows that the aged sample displayed slightly higher hardness and elastic modulus compared to the unaged samples. This is most probably caused by the compacting of the Molybdenum IRR layer owing to extended stay at higher temperature. Also, it indicates that the AB and AR layers provide enough protection for the AB Mo/Al2O3 and IRR Mo layers against high temperature oxidation. Such a condition would deteriorate their mechanical and thus optical properties.

Table 3. Comparison of mechanical properties of the entire absorber coating obtained with 3.0 mN maximum load indentation

unaged 450 °C 500 °C
Hit
[MPa]
Average
St. Dev.
9590.5
1081.8
12462.9
608.1
12116.4
1042.8
Eit
[GPa]
Average
St. Dev.
177.0
14.3
190.2
10.3
192.1
9.6
hm
[nm]
Average
St. Dev.
135.9
7.3
121.7
2.9
123.0
4.5
nit
[%]
Average
St. Dev.
37.0
2.6
43.2
1.3
41.3
3.5

Adhesion – Nanoscratch Testing

Scratch tests with progressively increasing load were performed in order to generate damage in the coating and on the coating-substrate interface. This was done to find out the adhesion and cohesion of the coating [12,13]. The Panorama feature can help save the whole scratch track together with the recorded signals and the results between the aged and unaged samples can be easily compared. This enabled the definition of critical loads a posteriori, making it possible to define all critical loads on all the samples in a similar way. Three critical loads (Lc1, Lc2 and Lc3) were established on all tested coatings (aged and unaged):

  • the first critical load (Lc1) relates to the first cracks seen in the scratch track
  • the second critical load (Lc2) relates to partial delamination (partial apparition of the substrate)
  • the third critical load (Lc3) relates to total delamination (apparition of the substrate on the entire width of the scratch track)

The scratch tests were accomplished in a number of regions on both aged and unaged samples (A minimum of four scratch tests in two to three regions on each individual sample). Only slight variations were observed between the scratch tracks on both aged (450 °C and 500 °C) samples. Also, the typical morphology relative to critical loads was extremely similar on these samples, thus showing only the usual result of scratch tests on the 450 °C sample.

Figure 9 depicts a classic example of a scratch on the unaged sample including a Panorama image (9a) and complete images of the critical load areas (Figure 9b). Among these, the most difficult to determine was the first critical load (Lc1): Lc1 corresponds to thin cracks originating from the sides of the scratch track. The second critical load (Lc2) is highlighted by the initial appearance of substrate in the scratch track (bright areas). The third critical load (Lc3) is highlighted by apparition of substrate via the entire width of the scratch.

Comparison of hardness and elastic modulus of the entire coating on the unaged and aged samples (indentation with 3 mN maximum load).

Comparison of hardness and elastic modulus of the entire coating on the unaged and aged samples (indentation with 3 mN maximum load).

Figure 8. Comparison of hardness and elastic modulus of the entire coating on the unaged and aged samples (indentation with 3 mN maximum load).

Typical scratch recording and Panorama image on the unaged sample (a), Panorama of the scratch track with detailed images of the corresponding Lc1, Lc2 and Lc3 areas (b).

Figure 9. Typical scratch recording and Panorama image on the unaged sample (a), Panorama of the scratch track with detailed images of the corresponding Lc1, Lc2 and Lc3 areas (b).

A typical scratch on the sample aged at 450 °C is shown in Figure 10. The first critical load is related to the coating’s cohesion, that is damage within the coating, whereas the second (Lc2) and third critical loads (Lc3) are associated with adhesion of the coating to the substrate. The scratch tests thus enabled direct comparison of the cohesion and adhesion of the coating to the substrate for the aged and unaged samples.

Typical scratch recording and Panorama image on the sample aged at 450 °C (a) and corresponding Lc1, Lc2 and Lc3 areas (b).

Figure 10. Typical scratch recording and Panorama image on the sample aged at 450 °C (a) and corresponding Lc1, Lc2 and Lc3 areas (b).

Table 4 shows the comparison between average critical load values acquired during nanoscratch testing on the unaged samples as well as samples aged at 500 °C and at 450 °C. The surprising fact was that the first and second critical loads on the aged samples were higher than on the unaged samples. Also, the third critical load was higher on the aged samples than on the unaged sample, albeit the difference was not as prominent as for the critical loads Lc1 and Lc2.

Table 4. Comparison of critical load values of unaged and aged samples.

unaged 450 °C 500 °C
Lc1 [mN]
St. Dev.
94.4
17.4
136.3
22.8
150.8
12.1
Lc2 [mN]
St. Dev.
177.1
49.2
236.2
13.2
241.1
28.9
Lc3 [mN]
St. Dev.
331.7
36.8
376.7
24.2
384.9
18.2

Optical properties: NECSO partner Tekniker determined the optical properties of the coatings for both aged and unaged samples. The influence of aging parameters on the absorbance was found to be insignificant and no direct correlation between the absorbance and aging parameters could be established (Figure 11).

Comparison of optical properties of the aged unaged and aged samples.

Figure 11. Comparison of optical properties of the aged unaged and aged samples.

Discussion

Indentation

Indentation results: the indentation results demonstrate that aging results in slight increases in the elastic modulus and hardness of the entire coating (Figure 8) and slight decreases of hardness and Young’s modulus of the top layers (AB and AR), as shown in Figure 6.

For the entire coating as well as the top layers, the change in mechanical properties due to aging was believed to be negligible but since elastic modulus and hardness of the entire coating did not change or increase following the aging tests, it can be safely concluded that the AR and AB top layers are capable of providing an efficient barrier against high temperature oxidation.

However, it must be noted that the mechanical properties of the layers, which were deposited by PVD magnetron sputtering method, largely rely on different structures produced by the application of various deposition parameters [9, 10]. Further, the hardness of the two top layers (including SiO2 layer) was found to be lower (~3 MPa) than that described in literature for bulk SiO2 (~ 10 MPa).

Scratch Testing

Each critical load feature (delamination and cohesive cracking) is vital for investigating the system’s performance. Cohesive cracks present in the layers produce preferential places for corrosion and other degradation processes and hence, it is important to identify crack formation for analyzing the system’s durability. Moreover, delamination results in complete loss of optical performance and functionality of the solar selective coating and hence should be avoided at all costs.

The nanoscratch tests demonstrated increased critical load values (Lc1, Lc2 and Lc3) for the aged samples compared to unaged samples (refer Table 4 for critical loads values of aged and unaged samples). This may be attributed to better cohesion (sintering) of the AD and IRR layers during the ageing tests. Moreover, the impact of aging conditions on the critical load values was found to be negligible and there was no clear trend (refer Table 4 for critical load values for samples aged at 450 °C and at 500 °C. The aging conditions were not severe enough to negatively affect the coating’s adhesion.

The results described above were also confirmed by only slight changes in the optical properties of the aged coatings: the absorbance is somewhat increased following the aging tests, while the emissivity either slightly increased or remained unchanged.

Conclusions

Adhesion and mechanical properties were analyzed for solar selective coatings, which are used in the concentrated solar power industry. The coatings were easily deposited on the 80 mm diameter collector tubes, making it possible to characterize real components. Using a dedicated sample holder, the large cylinders were easily mounted on the nanoindentation and nanoscratch instruments. Although the top layers had low thickness, their elastic modulus and hardness could still be determined (before and after the aging tests). In the same way, adhesion and mechanical properties of the entire coating can be characterized by nanoscratch and nanoindentation testing.

The key results can be summed up as follows:

  • The hardness and elastic modulus of the top layers were in the range of 3000 MPa and ~70 GPa, respectively. The aging tests caused a decrease of ~15% of both elastic moduli and hardness.
  • The hardness of the entire coating was in the range of 10’000 MPa and the elastic modulus was 180 GPa. The aging tests led to an increased hardness of ~20% and increased elastic moduli of ~5% with regard to unaged coating.
  • The nanoscratch tests of the entire coating showed an increase of ~40% for the first critical load Lc1 of the aged samples when compared to unaged samples. Also, an increase of ~30% for the aged samples was seen for the second (Lc2) and third critical load (Lc3) when compared to unaged samples (+15%).

This is an experimental program, which was crucial to characterize the effects of accelerated aging on the adhesion and mechanical properties of the absorber coatings. The article provides a better understanding about the methodology of characterization of the aging effects of solar selective absorber coatings. By using this methodology, degradation of such types of coatings can be routinely evaluated during and after the accelerated aging tests.

Acknowledgements

This work was financially supported by the NECSO (Nanoscale Enhanced Characterization of Solar Selective Coatings) project, grant agreement no. 310344.

References

[1] C.E. Kennedy and H. Price. NREL/CP-520-36997 Proceedings of ISEC2005 2005 International Solar Energy Conference.

[2] C.E. Kennedy. Review of Mid- to High-Temperature Solar Selective Absorber Materials. July 2002. NREL/TP-520-31267.

[3] M. Lanxner, Z. Elgat. Solar selective absorber coating for high service temperatures, produced by plasma sputtering. SPIE, Optical Materials Technology for Energy Efficiency and Solar Energy Conversion IX, vol. 1272 (1990).

[4] S.A. Kalogirou, Solar thermal collectors and applications, Progress in Energy and Combustion Science, vol. 30 231–295 (2004).

[5] J. Barriga, U. Ruiz-Gopegui, J. Goikoetxea, B. Coto, H. Cachafeiro, Selective Coatings for New Concepts of Parabolic Trough Collectors. Energy Procedia (2013).

[6] N. Sevlakumar, H.C. Barshilia, Review of physical vapor deposited (PVD) spectrally selective coatings for mid- and high-temperature solar thermal applications, Solar Energy Materials and Solar Cells, vol. 98 (2012).

[7] D. Barlev, R. Vidu, P. Stroeve, Innovation in concentrated solar power, Solar energy materials and solar cells, vol. 95 (2011).

[8] P. Nava, FLABEG, “What can be done from a value engineering point of view”, Conference: “Concentrating Solar Power and Chemical Energy Systems 2011” SolarPACES (IEA).

[9] ISO 14577-1:2015 Metallic materials - Instrumented indentation test for hardness and materials parameters - Part 1: Test method.

[10] A.C. Fischer-Cripps: The IBIS Handbook of Nanoindentation (Fischer-Cripps Laboratories Pty Ltd., Australia, Springer (2011).

[11] M.V. Swain and J. Mencik: Mechanical property characterization of thin films using spherical tipped indenters. Thin Solid Films 253, 204 (1994).

[12] P.A. Steinmann, Y. Tardy, H.E. Hintermann. Adhesion testing by the scratch test method: the influence of intrinsic and extrinsic parameters on the critical load. Thin Solid Films 154 (1987).

[13] S.J. Bull, E.G. Berasetegui. An overview of the potential of quantitative coating adhesion measurement by scratch testing Tribology International 39, issue 2 (2006).

This information has been sourced, reviewed and adapted from materials provided by Anton Paar TriTec SA.

For more information on this source, please visit Anton Paar TriTec SA.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit