Open Access
Add to BibSonomy
Abstract
Developing durable antireflection (AR) coatings with sapphire-like hardness and high transparency faces a significant challenge. Conventionally, achieving these requirements involves depositing thick, high-hardness nitride films. Here, we proposed an alternative approach that combines nanolaminate materials with optical design, overcoming the brittleness of thick nitride films. We selected Ta2O5/Si3N4 nanolaminates with similar refractive indices, improving tribological and optical performance through a unique optomechanical method. Our proposed AR coating exhibited a low reflectance of 0.8% (420-780 nm) and remarkable hardness of 22.8 GPa, and demonstrated the ability to withstand abrasion from steel wool up to 3,000 times on a glass substrate. This work successfully achieves a balance between hardness and toughness, opening new avenues for the development of highly durable coatings.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Antireflection coatings are widely used in optical components to minimize optical losses and mitigate the impact of reflected beams. However, current AR coatings often lack scratch and wear resistance compared to sapphire, which is favored for its high hardness as a cover glass in the electronics industry [1]. In recent years, silicon nitride has emerged as the leading choice for producing durable antireflection coatings due to its transparency, hardness, and other functional characteristics [2]. Paulson et al. proposed a hard AR coating consisting of a thick Si3N4 layer sandwiched between index-matching stacks on both sides to achieve a hardness exceeding 18 GPa [3]. The thickness of the hard middle layer determines this value. However, strength and toughness properties are mutually exclusive for most dielectric hard materials, such as Si3N4, SiOxNy, AlN, AlOxNy, and DLC films [4,5].
Most composite systems, such as nanolaminates [6], functionally graded coatings (FGCs) [7], and nanocomposite [8] coatings, have been extensively studied to surpass the limitations of single films and achieve superior mechanical performance. Oleksiy et al. proposed a transparent hard coating based on a Si3N4/BN periodic structure, which exhibited optical transparency (T > 85%), sapphire-like hardness (35 GPa), and excellent wear protection and flexibility properties [9]. Jõgiaas et al. investigated the optical and mechanical properties of various oxide coatings with nanolaminate structures, including HfO2/ZrO2, Al2O3/Ta2O5, and Al2O3/ZnO [10]. They demonstrated that adjusting thickness ratio could tailor the refractive index and hardness. However, there is limited research on integrating optical antireflection and mechanical properties using nanolaminate structures for optical industrial applications.
In this study, a Ta2O5/Si3N4 nanolaminate structure is first introduced into the fabrication of hard AR coating with high surface hardness and tribological resistance. The structure of the coatings is based on glass/index-matching stacks#1 /nanocomposite/ index-matching stacks#2/Air. Furthermore, we propose an optomechanical design method that can adjust coating thickness ranging from 500 nm to 2000nm, providing resistance against varying scratches. By harnessing the “hard” and “tough” physical properties of nanolaminate structures, we demonstrate that AR coatings based on Si3N4/Ta2O5 nanolaminate maintain a low residual reflectance of ∼0.8% while exhibiting an indentation hardness exceeding 22.8 GPa near the surface. The coating withstands over 3000 cycles of steel wool abrasion under a 500 g load. With its exceptional mechanical durability and optical transparency, the proposed protective AR coating holds promise for diverse applications, including solar cells, automobiles, and the electronics industry.
2. Experimental
Device fabrication: An inductively coupled plasma (ICP) assisted RF magnetron sputtering method was used to deposit hard AR coatings. Two different targets were used: (1) a tantalum metal target (99.999% purity), and (2) a silicon target (99.999% purity). Throughout these experiments, the ICP at an RF power of 3 kW was always sustained. By sputtering the Si and Ta targets, SiO2, Ta2O5, and Si3N4 thin films with standard stoichiometric ratios were obtained by reacting with thin metal layers according to filling O2 and N2 in the ICP source, respectively. Ta2O5/Si3N4 nanolaminate was prepared with a total thickness of approximately 1000 nm by depositing alternating thin layers of 50 nm Si3N4 and ultrathin 5 nm Ta2O5.Here, the thickness is determined based on the deposition rate of each individual layer. The substrate was not specially heated in the coating process, but the temperature would gradually rise because of the thermal radiation of ICP-assisted sputtering. Separate temperature measurements inside the substrate holder showed that the deposition temperature never exceeded 120°C.
Simulation and measurements: The reflectance measurement at normal incidence was conducted using a spectrophotometer (Olympus USPM-RU). The refractive index of the thin films was measured using the photometric method. That is, the optical constants and thickness of the films were obtained by fitting the measured transmittance and reflectance spectra based on the reverse engineering. Nanoindentation was performed using a high-precision hardness tester (Anton Paar, UNHT). A Berkovich pyramid indenter having a rounded end was used. According to the convention, the indentation depth should not exceed 10% of the coating thickness because the indentation results of the films cannot be treated independently of their substrates. Therefore, we used the average hardness value at a tip displacement (indentation depth) of 50 nm as a comparison criterion for SiO2, Ta2O5 and Si3N4 films. The indentation values were calculated using the Oliver-Pharr method from 20 indentations per film. The thickness of each multilayer film proposed in this paper can reach 1.5 microns. To characterize the hardness inside the entire film layer, Agilent G200 was used to obtain the load-displacement curve of the multilayer films and further determine the elastic modulus and hardness of the multilayer systems. Nano scratch tests were conducted with the XP head of the Agilent G200 nanoindenter equipped with the lateral force measurement option. In the progressive load scratch tests, the tip scratched the sample surfaces at a progressive load, increasing linearly from 0 to 500 mN over a distance of 50 to 350 µm.
3. Results and discussion
3.1 Fabrication and optical properties
Figure 1(a) depicts a schematic diagram of the proposed hard antireflection coating consisting of a thick nanolaminate structure sandwiched by two dielectric stacks. Because we aim to protect the underlying substrate against typical scratches, the total thickness of the hard AR coating is targeted at about 1-2µm. In our work, silicon nitride (Si3N4) is selected as the critical material due to its remarkable mechanical and optical properties, low absorption loss, and high refractive index over all visible wavelengths. Two dielectric stacks are comprised of alternative SiO2/Si3N4 films, serving as the index matching stacks to achieve better antireflection function. However, the biggest problem is that the hardness and crack-resistant properties of most nitride materials are incompatible. Nanolaminates are materials prepared on nano-scales, with structures consisting of alternate ultra-thin layers of materials. Here, a unique 5 nm Ta2O5(soft)/50 nm Si3N4(Hard) nanolaminate structure is established and considered an effective way to improve the toughness of brittle Si3N4 film. Although Ta2O5 is chosen for the demonstration in this study, other materials with lower elastic modulus and refractive index around 2.0, such as niobium pentoxide (Nb2O5) and titanium dioxide (TiO2), can also be used to attain a similar property. An inductively coupled plasma (ICP) assisted MF magnetron sputtering method was used to deposit hard AR coatings. The transmission and reflection spectra of our prepared hard coating were calculated by the transfer matrix method. The measurements were conducted using a spectrophotometer (Agilent Cary 7000). The photometry method was employed to determine the optical constants (seen in Fig. S1, Supplement 1) of the single films by fitting the measured reflectance and transmittance curves. We can confirm that the single-prepared films exhibit stoichiometric composition based on the obtained optical constants of SiO2, Si3N4, and Ta2O5 films.
Fig. 1. (a) Schematic diagram of the proposed hard antireflection coating deposited on glass substrate. (b) Cross-section image of the nanolaminate structure composed of alternating multilayers of 5 nm Ta2O5(soft)/50 nm Si3N4(hard). (c) Measured transmittance and absorption of our AR coating under light at normal incidence. (d) The measured reflection spectra of the proposed hard antireflection coating from the normal incidence to the oblique incidence of 10°, 20°, and 30°.
Figure 1(b) illustrates the cross-sectional image of the nanolaminate structure, which comprises 18 periodic alternating multilayers of 5 nm Ta2O5 (soft) and 50 nm Si3N4 (hard). The Si3N4 film shows a normal dispersion in the visible region and exhibits a small extinction coefficient leading to an acceptable absorption. Theoretically, the calculated average transmission over the 420-780 nm wavelength range reaches 98.7%. A distinction in the transmission between the calculated and measured spectrum can be observed, as is shown in Fig. 1(c). The experimental reflectance of ∼94.3% is always lower than the simulated one, which can be attributed to the absorption of the Ta2O5/Si3N4 nanolaminate. Absorption can be easily calculated using A = 1 – R – T, as indicated by the green curve in Fig. 1(c). The deposition rate was initially calibrated by measuring the thickness of the relatively thick single-layer coating. Therefore, the observed nanolaminate in Fig. 1(b), consisting of 5 nm Ta2O5 and 50 nm Si3N4 periods with a ratio of 1:2, is not due to deposition errors or similar factors but instead attributed to the presence of a transition layer between the Si3N4 and Ta2O5 nanolaminate composite. This process can significantly increase the absorption of the nanolaminate, which has been verified in the transmittance of AR#Nanolamimnate. Figure 1(d) shows the measured reflection spectra of the hard AR coating for an angle of incidence (AOI = 0-30°). The measured average reflectance ranging from 420 to 780 nm at normal incidence is 0.8%. Good behavior of angle robust could be obtained due to the high equivalent refractive index ∼2.0 of the nanolaminate structure. We provide the reflectance at a different angle of incidence using low-index material SiO2 as the intermediate layer, demonstrating a substantial increase in the spectral shift (seen in Fig. S2, Supplement 1).
3.2 Mechanical properties
We study the mechanical properties of three types of hard AR coatings with different hard layers (Ta2O5, Si3N4 layer and Si3N4/Ta2O5 nanolaminate) in the middle, respectively named AR#Ta2O5, AR#Si3N4, and AR#Nanolaminate. The structural diagram of the hard AR coatings is shown in Fig. 2(a). The optical properties of the three films are shown in Figs. S3-S5, Supplement 1. The structure of the proposed hard AR coatings can be divided into two parts: (1) the top and bottom stacks are used to match the refractive index of the middle thick layer, and (2) the thick spacer layer in the middle. The spacer layer can also be designed as a soft/hard nanocomposite to prevent brittle fracture of a single Si3N4 film. The detailed structural configurations of our AR coatings are summarized in Table 1. Two other AR coatings with different spacer materials of the hard layer are also prepared for comparison. Ta2O5 and Si3N4 with different elastic moduli were selected to engineer soft/hard nanocomposites to realize enhanced toughness. The toughening mechanisms are schematic illustrated in Fig. 2(b). As noted above, elastic modulus of Ta2O5 is small enough to resist Si3N4 from being broken due to crack propagation. Single layers of SiO2, Ta2O5, and Si3N4 films have been deposited on BK7 glass by sputtering to determine the hardness. A nanoindentation instrument, equipped with a diamond Berkovich tip, was used on the surface of the samples to measure the elastic modulus and hardness based on the well-established technique developed by Oliver and Pharr. Pure O2-prepared SiO2 films have a hardness (H) and elastic modulus (E) of 9.5 GPa and 75.2 GPa, respectively. Si3N4 films prepared with 100% N2 have a hardness and elastic modulus of 23.5 GPa and 233.3 GPa, respectively. Ta2O5 films have a lower hardness and elastic modulus of 8.5 GPa and 96.5 Gpa, respectively. These results are consistent with the references respectively [11–13].
Fig. 2. (a) Structural diagram of three types of hard AR coatings with Ta2O5, Si3N4 layer, and Si3N4/Ta2O5 nanolaminate in the middle. (b) Structural diagram of nanolaminate structure, which can strongly resist cracking compared to single material as a hard layer. (c) Indentation curves for three types of hard AR coatings and glass substrate. (d) The hardness of three types of hard AR coatings as a function of indentation depth.
Table 1. Structural configurations of the hard AR coatings. H and L represents Si3N4 and SiO2 respectively. The hardness of three types of hard AR coating.
The indentation curves of this hard antireflective (AR) coatings and the bare glass substrate are shown in Fig. 2(c). Furthermore, Fig. 2(d) indicates that these three samples exhibit similar hardness values near the surface (<200 nm). However, as the indentation depth increases, hardness values show a noticeable variation. These variations in hardness can be attributed to the different material hardness within the thin film structure. The near-surface structures of the three AR samples are identical, with middle layers consisting of Ta2O5, nanolaminate, and Si3N4, respectively, displaying increasing hardness in that order. The hardness value of AR#Si3N4 samples remains constant at 22.4 GPa, regardless of the indentation depth. AR#Ta2O5 exhibits a surface hardness of 17.9 GPa, but it exhibits a rapid decrease in hardness with increasing depth, reaching only 11 GPa at a depth of 450 nm. AR#Nanolaminate demonstrates favorable “strength properties” near the surface, with a hardness value of 22.8 GPa. However, as the indentation depth enters the nanocomposite material region, the hardness gradually decreases to 16 GPa.
3.3 Unique opto-mechanical design method
We employed a proposed optomechanical approach to achieve arbitrary variations in the thickness of the AR coating within the range of 500 nm to 2000nm while maintaining its optical performance unchanged. This approach offers the advantage of tailoring the AR coating to different levels of scratch resistance and durability by varying its thickness. Figure 3(a) depicts the schematic configuration of the proposed hard AR coating incorporating a varying thickness of nanolaminate. The total thickness of our hard AR design can be modified through an optical design method: buffer layer strategy [14]. Because the thickness of the buffer layer can be changed arbitrarily, this strategy significantly increases the design freedom of the mechanical properties of thin films. The refractive index profile of the design is presented in Fig. 3(b). The optical properties can be further investigated using the optical admittance diagram. Figure 3(c) shows the calculated reflection spectra for our hard AR coating with different thicknesses of the buffer layer, which have strikingly similar reflectance vs. wavelength performance exhibited by the coating. The designed average reflectance remains unchanged, at 1% or lower, as the spacer layer thickness varies from 0 to 2000nm. Figure 3(d) shows the optical admittance locus of our AR coating. Theoretically, a perfect index match to the air (i.e., the admittance ends at air (1,0)) can achieve zero reflection. With BK7 glass as a substrate, the starting point of the admittance trajectory is (1.52,0). Near the substrate side, the SiO2/Si3N4/SiO2/Si3N4/SiO2 combination forms a typical index-matching profile of broadband AR coatings for the glass substrate. The combined admittance of glass and index-matching stacks #1 equals N (here N = 2.05, refractive index of Si3N4). In this condition, when a thin film of refractive index N is deposited on it, no matter how thick the film is, it will not affect the reflectance/ transmittance. The function of index-matching stacks #2 is to transform the combined admittance N to the refractive index of air, so that the thin film can obtain low reflectance. Figure 3(e) shows the optical admittance locus of the nanocomposite. The refractive index of the two materials Si3N4 and Ta2O5 is comparable, which can be considered a single layer in optics. Although the thickness ratio of Si3N4/Ta2O5 50nm:5 nm is chosen for the demonstration in our work, other thickness ratios can also be used to attain a similar property.
Fig. 3. (a) The total thickness of our hard AR design can be modified by a unique opto-mechanical design method. (b) Refractive-index profile of the proposed hard AR coating. (c) The calculated reflection of AR coating with the thickness of the buffer layer varies, ranging from 0 to 2000nm. (d) The optical admittance of the hard AR coating at 550 nm with buffer layer and two impedance matching stacks. (e) The optical admittance of the buffer layer composed of 5 nm Ta2O5/50 nm Si3N4 nanocomposite.
3.4 Scratch and tribological performance
The scratch test was carried out for the three hard coatings. Figure 4(a) show the penetration depth profile under loading and after elastic recovery of the scratches performed in AR#Si3N4, AR#Ta2O5, and AR#Nanolaminate coatings. In the progressive load scratch tests, the tip, moving at a speed of 10 um/s, scratched the sample surfaces at a progressive load, increasing linearly from 0 to 500 mN over a distance of 350 µm. The residual groove refers to the indentation or groove left on the material's surface after the scratch, while the scratch groove refers to the actual mark or groove produced during the test. The critical load LC corresponds to catastrophic damage of the film or extensive exposure of underlying substrate (as seen in Fig. 4(a)). By quantitatively measuring the residual grooves and observing the point of films exposed to substrate using microscopic images (as seen in Fig. 4(b)), we can provide a comprehensive assessment of the failure behavior of the film under applied stress. The failure mechanisms of AR#Ta2O5 and AR#Si3N4 coatings are quite different. Note that the AR#Ta2O5 coating exhibits severe plastic deformation. The residual scratch curves of the AR# Si3N4 coating change abruptly. As the load increases above 270 mN, delamination due to cohesive and adhesive failure may occur in AR#Si3N4 coating, which shows eggshell flaking after scratches. AR#Nanolaminate shows significant improvement in scratch resistance against severe damage. It was found that the delay of the occurrence of the cracks was primarily due to the enhanced toughness stored in the Ta2O5/Si3N4 nanolaminate structure, which effectively suppressed the initiation of cracking and delamination.
Fig. 4. (a) A progressive load scratch tests, the tip scratched the sample surfaces at a progressive load, increasing linearly from 0 to 500 mN over a distance of 50 to 350 µm. (b) 3D microscopic images of the coating surfaces after nano scratch test. (c) Micrographs of the coating surfaces after being abraded with steel wool for 3000 times under the weight of 500 g, respectively. (Magnification: 100X).
The tribological behavior of the AR#Ta2O5, AR#Si3N4, and AR#Nanolaminate coatings was studied by sliding them 3000 times with steel wool under the weight of 500 g. We compare the results with the behavior for the bare substrate in Figs. S6-S7, Supplement 1. Numerous scratches were observed on the AR#Ta2O5 surface due to its low hardness (Fig. 4(c)). The wearing tracks were visible on the AR#Si3N4 coating. The microscopic image of the AR#Si3N4 coating shows that the film is peeling away from the substrate due to crack propagation. After conducting the abrading test of 3000 times, we provided additional information on the optical properties of the proposed AR coatings in Fig. S8, Supplement 1. On the other hand, this phenomenon did not occur in the case of AR#Nanolaminate coating, indicating that they have obvious crack resistance and relatively high hardness. We also made falling sand abrasion tests on different samples to demonstrate the resistance to cracking, as is shown in Figs. S9 and S10, Supplement 1. The improved wear performance of multilayer coatings is determined by optimizing their architecture to achieve the best combination of hardness and toughness. Multilayered AR coatings, made of different materials, can also be vulnerable to high temperatures due to mismatches in thermal expansion coefficients. This vulnerability can result in delamination or cracking, which can greatly compromise the coating's performance and alter its optical properties. Therefore, we further demonstrated the thermal characteristics alongside the optical properties when considering the durability of AR coatings (Figs. S11 and S12, Supplement 1).
4. Conclusion
In summary, we have successfully prepared a highly wear-resistant, sandwich-like hard AR coating on glass substrate with a low average reflection of∼0.8% and a high hardness of ∼22.8 GPa through ICP-assisted RF magnetron sputtering. By using a Ta2O5/Si3N4 nanocomposite structure, we enhanced wear resistance significantly owing to their multilayer toughening so that the conflict of hardness and toughness can be balanced more effectively. Through a unique opto-mechanical approach, the coating thickness can be adjusted thick enough to resist different degrees of scratches. Good wear performance can be maintained after being abraded with steel wool 3000 times under the weight of 500 g. Our proposed hard AR coating has a great advantage in low-cost fabrication and mass production.
Funding
National Natural Science Foundation of China (501100001809).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the corresponding author upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. C. Gödeker, U. Schulz, N. Kaiser, and A. Tünnermann, “Antireflection coating for sapphire with consideration of mechanical properties,” Surf. Coat. Technol. 241(25), 59–63 (2014). [CrossRef]
2. S. Bruns, M. Vergöhl, T. Zickenrott, and G. Bräuer, “Deposition of abrasion resistant single films and antireflective coatings on sapphire,” Surf. Coat. Technol. 290(25), 10–15 (2016). [CrossRef]
3. C. A. Paulson, J. J. Price, K. W. Koch, C. G. Kim, J. H. Oh, L. Lin, A. N. Subramanian, B. Zhang, J. Amin, A. Mayolet, C. Kosik Williams, and S. D. Hart, “Industrial-grade anti-reflection coatings with extreme scratch resistance,” Opt. Lett. 44(24), 5977 (2019). [CrossRef]
4. R. O. Ritchie, “The conflicts between strength and toughness,” Nat. Mater. 10(11), 817–822 (2011). [CrossRef]
5. B. Merle and M. Göken, “Fracture toughness of silicon nitride thin films of different thicknesses as measured by bulge tests,” Acta Mater. 59(4), 1772–1779 (2011). [CrossRef]
6. V. Chawla, M. Ruoho, M. Weber, A. A. Chaaya, A. A. Taylor, C. Charmette, P. Miele, M. Bechelany, J. Michler, and I. Utke, “Fracture mechanics and oxygen gas barrier properties of Al2O3/ZnO nanolaminates on PET deposited by atomic layer deposition,” Nanomaterials 9(1), 88 (2019). [CrossRef]
7. R. Fathi, H. Wei, B. Saleh, N. Radhika, J. Jiang, A. Ma, M. H. Ahmed, Q. Li, and K. K. Ostrikov, “Past and present of functionally graded coatings: Advancements and future challenges,” Appl. Mater. Today 26(1), 101373 (2022). [CrossRef]
8. J. M. Paiva, G. Fox-Rabinovich, E. Locks Junior, P. Stolf, Y. Seid Ahmed, M. Matos Martins, C. Bork, and S. Veldhuis, “Tribological and Wear Performance of Nanocomposite PVD Hard Coatings Deposited on Aluminum Die Casting Tool,” Materials 11(3), 358 (2018). [CrossRef]
9. O. V. Penkov, M. Khadem, and D. E. Kim, “Hard, Flexible, and Transparent Nanolayered SiNx/BN Periodical Coatings,” ACS Appl. Mater. Interfaces 11(10), 9685–9690 (2019). [CrossRef]
10. T. Jõgiaas, R. Zabels, A. Tamm, M. Merisalu, I. Hussainova, M. Heikkilä, H. Mändar, K. Kukli, M. Ritala, and M. Leskelä, “Mechanical properties of aluminum, zirconium, hafnium and tantalum oxides and their nanolaminates grown by atomic layer deposition,” Surf. Coat. Technol. 282(25), 36–42 (2015). [CrossRef]
11. V. Jambur, M. Molina-Ruiz, T. Dauer, D. Horton-Bailey, R. Vallery, D. Gidley, T. H. Metcalf, X. Liu, F. Hellman, and I. Szlufarska, “Temperature effects on the structure and mechanical properties of vapor deposited a-SiO2,” J. Non-Cryst. Solids 587(1), 121588 (2022). [CrossRef]
12. H.-M. Piirsoo, T. Jõgiaas, P. Ritslaid, K. Kukli, and A. Tamm, “Influence to hardness of alternating sequence of atomic layer deposited harder alumina and softer tantala nanolaminates,” Coatings 12(3), 404 (2022). [CrossRef]
13. M. Vila, D. Cáceres, and C. Prieto, “Mechanical properties of sputtered silicon nitride thin films,” J. Appl. Phys. 94(12), 7868–7873 (2003). [CrossRef]
14. J. Mouchart, “Thin film optical coatings. 5: Buffer layer theory,” Appl. Opt. 17(1), 72 (1978). [CrossRef]
