High performance hierarchical nanoporous antireflective films by a facile sol–gel process

Abstract

Single layer antireflection (SLAR) sol–gel films on super white glass substrates were prepared by hierarchically introducing hollow silica nano-spheres (HSNs) and F127 templated nano-pores. HSNs and F127 templated nano-pores were formed by base and acid catalyzed sol–gel processes, respectively. The dispersibility of HSNs was ameliorated by Ph-TES surface modifications. The porosity of the SLAR films was engineered by a step-by-step addition of nano-pores and HSNs so that the refractive index could be tuned to be as low as the desired value of 1.24. It accordingly improved the transmittance of the films with an optimized maximum transmittance of 98.21% in the visible spectral region. Furthermore, the concentration of modified HSNs was adjusted to modulate the peak transmittance wavelength from 550 nm to 680 nm. The nano-scale surface embossments due to the addition of HSNs was also contributory to the antireflective effect of the films, especially at 45° or 5° incident angle. As a result, the average transmittance of the double side coating super white glass was increased by 7.10% at 400–1000 nm with good wear resistance. The porosity directed mechanical durability varied from 6H to 3H (pencil hardness). It displays considerable AR application prospects for suitable optical and optoelectronic devices.

---

1. Introduction

Anti-reflection (AR) coating plays a pivotal role in a wide variety of optical technologies.^1–8 It can suppress the reflection of light on the surfaces or interfaces of transparent substrates and improve optical properties (transparency and visibility). In particular, for solar systems, it is able to improve the efficiency of solar power generation by increasing the absorption of certain spectral response ranges in solar collectors.⁹ In addition, many applications require an increase in the transmittance in the front surface reflection of transparent materials.

The amplitude of reflected light at an interface formed by media with refractive indices of n₁ and n₂ can be represented by the reflection coefficient given by

R = (n₁ − n₂)²/(n₁ + n₂)² (1)

at normal incidence. The refractive index of the commercial super white glass is approximately 1.5, so an R of ∼4% would be generated on each side and the transmittance would be ∼92%. In order to suppress the reflection of an AR coating, at least two strategies could be deployed: one is to insert graded refractive index layers between the air and the substrate; the other is to generate a destructive interference of the reflection light by a λ/4 thick coating, where λ is the wavelength of the incidence light. For the insert of a single-layer AR (SLAR) coating, the optimized refractive index is given by the following equationwhere n_a and n_s are the refractive indices of the air and the substrate, respectively. Thus the refractive index of 1.24 is ideal for SLAR coatings.^10–12 Among various used materials for AR coating, silica is the most ideal due to its low reflective index about 1.45, outstanding environmental resistance and excellent durability. But the reflective index of a compact silica film is still too high to meet the SLAR coating requirement. The introduction of nano-porous structures could efficaciously reduce the reflective index of films by increasing the percent porosity, P, according to the following equation:¹³

n_p² = (n² − 1)(1 − P/100) + 1 (3)

where n_p and n are the reflective indices of the porous and the nonporous materials, respectively. Ultimately, by holding the thickness of λ/4 and destructively interfering the reflection, the SLAR coating can be further optimized to be with the minimum reflection.¹⁴

On the basis of aforementioned principles, various methods have been applied to fabricate AR coating. The preparation of AR coatings can be achieved either by dry deposition methods such as chemical vapour deposition (CVD), or by wet chemical methods such as self assembly method,¹⁵ template removal method and sol–gel method. The deposition method was ever predominantly used by industrial production due to its facility in controlling thickness and stacking multilayer. However, high-quality deposited coatings, especially in order to get complicated nanostructures such as butterfly wings,¹⁶ helix morphology¹⁷ and periodic pyramid,¹⁸ are expensive and also suffer from fundamental material limitations.¹⁴ Alternatively, the mesoporous coating structures were reported in recent literatures by using the wet chemical methods, such as self assembly method, template removal method and sol–gel method. A single layer antireflection (SLAR) coatings self assembled by SiO₂ nanoparticles could lower its reflection as low as 0.5%.¹⁵ In order to increase the porosity of the coating, hollow SiO₂ nanospheres (HSNs) could further be introduced into the porous film to get an extremely high porosity and refractive indices as low as 1.2.¹⁹ Porous polymer films have also been extensively employed for AR films usually by template removal methods.²⁰ But HSN self-assembly films usually show high surface roughness to suppress their resistance to mechanical abrasion, and porous polymer films by removing templates have very poor thermal stability and poor anti-scratch property.²¹ Recently, sol–gel for coating is becoming one of the most attractive method because it is as facile as CVD to control the coating parameter (component, thickness), low cost with roller coating, and green initiative by avoiding to use or produce some toxic gas. Furthermore, it can be combined with various pore formers, such as HSNs, block copolymers,²² nano polymer particles,^23,24 surfactant,²⁵ etc. Meanwhile, it could induce subwavelength-scale nanostructures on the surface of sol–gel coating. That is another effective way of inducing low reflectance. N. Mizoshita et al.¹⁹ have reported such type of hierarchically nanoporous thin films constructed by mesoporous silica nanoparticles (MSNs) and a template removal mesoporous silica matrix. The mesoporous matrix filling the inter-particle voids between MSNs was evidenced to enhance the mechanical durability of the coating. A low surface roughness (e.g., <100 nm) with low aspect ratios was also proved to guarantee wear resistance of the coatings. As a result, the optimal nanoporous coating exhibited high transparency (91.5–97.5%) and low reflectance (<2.2%) over the whole range of visible light wavelengths, and sufficient wear resistance. Hence, this kind of nanoporous AR films are expected to improve with increasing structural and performance controllability by hierarchically mesoporous matrix and mesoporous nanoparticles. Nevertheless, a bottleneck problem still maintain on how to solve the compatibility between mesoporous matrix and mesoporous nanoparticles. It is crucial to achieve the stability of precursor sols and product coatings when applied to practice.²⁶ To solve this problem, the surface-modification is necessary for mesoporous and microporous nano-silica to uniformly disperse in the precursor sol an final AR coating and to form a more stable and high-reproducible composition.^27,28

Multilayer coatings have also been explored to achieve graded refractive index in AR coating for wide range of wavelengths in the visible region.²⁹ However, the preparation of multilayer coatings by the sol–gel process is difficult as it requires the optimization of optical properties and thicknesses of more than two layers. Compared to multilayer coatings, single-layer antireflective (SLAR) coating have been applied extensively because it is more easy to scaled up.¹¹ An amount of hard work has been explored to develop SLAR films with the desired functional characteristics (abrasion-resistant, broadband antireflective and self-cleaning properties). For example, D. B. Mahadik et al.²⁹ prepared nano-porous SLAR on glass substrates using tetraethoxysilane (TEOS), polyethylene glycol (PEG) and Triton X-100 were used as the polymeric porogens. M. Boudot et al.³⁰ explored the possibility to prepare AR films by curing mesostructured templated silica-based layers by ammonia vapors to avoid high temperature treatment. However, some complex or toxic processes were presented methods mentioned above, such as the CTAB removal in mesoporous silica nanoparticles, the complicated multi-step coating, the strong corrosive ammonia catalytic process and sub sequential template-removal using water or plasma. At the same time, improvement of light output in relation to angle of incidence, which is becoming more and more important in our real life, was rarely discussed in these papers. Hence, it is necessary to fabricate hierarchical SLAR coating with a safer and simpler way and try to explore other additional optical property.

In this paper, a series of SLAR films with hierarchical nano-porous structure are prepared by a facile sol–gel process. F127-templated nano-pores and Ph-TES modified Hollow Silica Nanospheres (HSNs) are designed to be homogenously dispersed in silica films. Our work is mainly focused on the surface modification of the hollow silica nanospheres in order to keep it compatible with the mesoporous silica matrix by removing F127 templates. The porosity of the films can be engineered by the amount of surfactant and hollow silica nano-particles during sol–gel processes. After calcination, the organic groups of HSNs can be decomposed with F127 together. The optimized SLAR film shows excellent transmittance as high as 98.21% in the visible region under vertical incidence and simultaneously keep a mechanical durability better than the pencil hardness of 3H. Simultaneously, the reflection at the 45° incident angle is also suppressed intensively with such hierarchical nano-porous AR films.

2. Materials and methods

2.1 Materials

Tetraethyl orthosilicate (TEOS, 98%), absolute ethanol (EtOH), hydrochloric acid (38%), polyacrylic acid (PAA, M_w = 3000), Pluronic F127 (EO₁₀₆PO₇₀EO₁₀₆, M_w = 12 800), phenyltriethoxysilane (Ph-TES) were purchased from Aladdin Company. Super white glasses (2 mm thick) were bought from Hebei Optical Instrument Company. All chemicals were of analytic grade and used without further purification. Ultrapure deionized water with a resistivity higher than 18.2 MΩ cm was used in all the experiments, which was obtained from a three-stage Millipore Mill-Q Plus 185 purification system (Academic).

2.2 Preparation of phenyl group modified hollow silica nano-spheres (Ph-HSNs)

In a typical procedure, 0.36 g of PAA dissolved in 4.0–6.0 mL aqueous ammonia was mixed with 100 mL of absolute ethanol. This was followed by the dropwise of 1.0–4.0 mL TEOS under vigorous magnetic stirring at room temperature. After 12 hours, the sol containing HSNs were formed. Then, ammonia was removed by stirring the sol in a ventilating cabinet at 80 °C for 2 hours. At the next step, Ph-TES and deionized water were slowly added to the HSN suspended sol and stirred at 40 °C for 4 h to form Ph-HSNs. Finally, white Ph-HSNs were obtained by centrifugation and water-washing.

2.3 Preparation of the coating sol

Ph-HSNs, TEOS, F127, 0.1 M hydrochloric acid and water were used to prepare various types of precursor sols (sol A; sol B; sol C; silica sol) for the fabrication of AR films with different nano-porous structures. Sol A was obtained by the redispersion of Ph-HSNs in the mixture of TEOS (0.5–1.5 mL), 0.1 M hydrochloric acid (600 μL) and deionized water (600 μL). It was kept at room temperature for 12 h for aging. Scheme 2a shows sol A could be the precursor of the AR film dispersed with relatively small pores (<10 nm). Sol B was obtained by the mixture of F127 (0.25 g), TEOS (0.5–1.5 mL), 0.1 M hydrochloric acid (600 μL) and deionized water (600 μL). It was ready for use after 2 days aging in a sealed glass container. Scheme 2b shows sol B could be the precursor of the AR films dispersed with relatively large pores (10–100 nm). Sol C could be generated by merging sol A and sol B, as shown in Scheme 2c, to fabricate the films containing both kinds of pores mentioned above. For comparison purpose, we also prepare a simple silica sol simply by the hydrolysis of the same amount of TEOS into deionized water under the similar acid condition. The silica sol could be coated on the glass substrate to fabricate the simple silica AR films. It is noted that the concentration of Ph-HSNs could be adjusted and quantified by the ratio of the TEOS amount to produce HSNs of the total TEOS amount in the precursor sol. Accordingly, three different sols C were prepared with different Ph-HSN concentration (10%; 30%; 50%). Surprisingly, the hybrid sol was steady for at least 3 months when stored at room temperature. This sol–gel process provides a way for the preparation of stable sol and antireflection coating, which have potential to be applied to solar cells, low-e coating and other electronic devices.

Scheme 1 The sol–gel process of phenyl-modified hollow silica nano-spheres to improve the dispersibility.

Scheme 2 The sol–gel processes of the SLAR films from (a) sol A containing Ph-HSNs, (b) sol C containing F127 and (c) sol C containing both Ph-HSNs and F127.

2.4 Preparation of AR coating

By coating different precursor sols (Table 1), the SLAR films (AR-SiO₂; AR-F127; AR-HSN; AR-10; AR-30; AR-50) were coating on the glass substrates. Firstly, a super white glass substrate was sonicated in deionized water for at least 15 minutes, and then treated by oxygen plasma under an oxygen flow of 800 mL min⁻¹. Then, a cleaned substrate was immersed in the precursor sol for 30 seconds, and withdrawn at a speed of 1–5 mm min⁻¹ on a SYDC-100 dip-coating machine. Finally, the glass with the dip-coated film was dried for 2 minutes at room temperature (25 °C) and calcined at 450 °C for 1 hour. All the glass substrates were coated on both sides.

Table 1 Coating method of the investigated nano-porous anti-reflection films

Coating sol	Anti-reflection film
a The concentration of Ph-HSNs could be adjusted and quantified by the ratio of the TEOS amount to produce HSNs of the total TEOS amount in the precursor sol.
Pure silica sol	AR-SiO2
Silica sol containing F127	AR-F127
Silica sol containing HSNs	AR-HSN
Silica sol containing F127 and 10% HSNs^a	AR-10
Silica sol containing F127 and 30% HSNs^a	AR-30
Silica sol containing F127 and 50% HSNs^a	AR-50

---

2.5 Characterization

The hollow nano-spheres and pore structure of the coating was investigated with a JEOL-2010 transmission electron microscopy (TEM) at 200 kV at high vacuum condition. The surface morphology and roughness of the films was determined with AFM in noncontact mode. The cross-section morphology of the films was observed using scanning electron microscopy. The Fourier transform infrared (FTIR) spectra of the hollow silica nano-spheres were recorded using a Nicolet (Thermo Fisher Scientific) system equipped with a KBr detector. An ellipsometer was used to measure the refractive index and thickness and to collect specular reflection spectra at oblique incidence angles of the films. Optical transmission and reflection measurements were performed using a U-4100 UV-vis-NIR spectrophotometer. The transmittance spectra from 240 to 2600 nm, at normal incidence, 5° or 45°, were measured using a Hitachi U-4100 spectrometer, where the slit was set to 4 nm. The particle size distribution of Ph-HSNs was determined in EtOH by the dynamic light scattering method. The pencil hardness test was performed by using pencils (B-3084T3) with hardness ranging from 6B (the softest) to 6H (the hardest). The coatings were scratched with the flat end of the pencil placed at a 45° angle to the surface with a loading of 750 g. The existence of the scratch was evaluated using an optical microscope.

3. Results and discussion

3.1 Surface modification with phenyl group

HSNs have been synthesized with a soft templating methods. The PAA, which was mixed with NH₄OH, can be templates for HSNs in ethanol. The shell thickness and size of HSNs can be conveniently controlled by adjusting amounts of TEOS and NH₄OH, respectively.³¹ By adjusting the amount of PAA, we can control the particle size to be about 10 nm to about 100 nm, as shown in the SEM image (Fig. 1a). The particle size distribution (Fig. 1b) represents a relatively wide nano-scale range. Thus, such kind of HSNs is suitable to coat several hundred nanometer thick AR films and is practical to build hierarchical nano-porous structure to fulfill both low refractive indices and good structure stabilities. The XRD pattern (Fig. 1c) manifests the HSNs are amorphous. The TEM observation (Fig. 1d) reveals all the HSNs have hollow sphere structures with about 10 nm thick shell layers. However, it also exposes that the HSNs are easily aggregated to each other. That is adverse for HSNs to be homogenously dispersed in AR films.

Fig. 1 SEM image (a), particle size distribution (b) and XRD patterns (c) of the hollow silica nano-spheres (HSNs); TEM images of HSNs before (d) and after (e) the surface modification by Ph-TES.

In order to improve the dispersibility of HSNs in the precursor sol, we develop a facile method to modify the surface states of the HSNs. Ph-TES is a very useful silane coupling agent. Scheme 1 clearly depicts the surface modification process of Ph-TES where three –SiOC₂H₅ groups on Ph-TES are firstly hydrolyzed and then polycondensed with three –Si–OH groups on the surface of the HSNs so that phenyl groups are bonded on the surface of the HSNs. Under weak acidic condition, the hydrolysis reaction proceeds completely before the polycondensation reaction in acetic acid-catalyzed system.³² So we can firstly hydrolyze Ph-TES and then let phenyl adsorb on the HSNs by aging. By implementing the surface modification method, a good dispersibility can be achieved for the Ph-HSNs, as observed by TEM in Fig. 1e. Compared with original HSNs, the phenyl groups anchored on the surface of HSNs may act as a great lubricant between random two particles, so the Ph-HSNs will not pile up under electron microscope observation. It was observed that the silica sol containing Ph-HSNs can be stored for at least 3 months, whereas the HSNs without modification by phenyl groups aggregated gradually in the silica sol after 15 days, thus indicating that such a modification process is effective to prevent agglomeration.³³

Phenyl groups are clearly proved to be bonded on the surface of HSNs by the FTIR spectra (Fig. 2). For both HSNs before and after the modification, the vibration bands of –OH, the anti-symmetric stretching vibrations, the symmetric stretching vibrations and the bending vibrations of Si–O–Si bonds are observed around 3440 cm⁻¹, 485 cm⁻¹, 792 cm⁻¹ and 1094 cm⁻¹, respectively. But for Ph-HSNs, additional bands appear at 742 cm⁻¹ and 1430 cm⁻¹, corresponding to the characteristic Si–C and Si–phenyl stretching vibrations, respectively.³⁴ These results suggest that the hydroxyl groups on the silica particles were replaced partially by phenyl groups and the condensation between Ph-TES and HSNs occurred.

Fig. 2 FTIR spectra of HSNs before (a) and after (b) the surface modification by Ph-TES.

3.2 Microstructures and mechanical properties

The microstructures of the hierarchical nano-porous films could be regulated by the ratio of the pore formers (F127 and Ph-HSNs) and the binders (silica sols). It will further determine the mechanical properties principally by the porosity factor. Fig. 3 and S2† show SEM surface morphologies of the films dipped with different ratio of Ph-HSNs while F127 is fixed to a constant concentration. After calcinations, Ph-HSNs produce relatively large pores (10 nm to 100 nm) clearly observed as white grey sphere domains in Fig. 3b and c, and F127 just create relatively small pores (<10 nm) appearing as tiny dark points in Fig. 3a–c. The density of Ph-HSN pores in the films increase with the augment of Ph-HSNs in the precursor sols, but closed pores cannot well-formed until the ratio of Ph-HSNs reaches 50% (AR-50). Especially, the gelation time of the films dramatically became longer with the increase of Ph-HSN ratio. Broken pores in AR-10 and AR-30 may be related with too quick gelation speed for these two samples. Clear acid silica sol can be well converted into transparent gel by a typical sol–gel process under certain gelation speed, while too quick gelation speed usually introduces stress in the materials and lead to undesired cracks.²² Here the gel process can be accelerated after introducing a small number of Ph-HSNs, but the gelation speed would fall back when feeding in a large number of Ph-HSNs. Because a small number of HSNs could be act as the active sites to promote the gel process, but a large number of HSNs would generate spatial steric hindrance large enough to slow down the gel process.³⁵ Although these values are probably underestimated because the AFM probes cannot enter inter particle voids entirely, the RMS coming from AFM images that presented in Fig. S1† can be consistent with the SEM images in some degree.

Fig. 3 SEM images of different AR films: (a) AR-F127; (b) AR-10; (c) AR-30; (d) AR-50.

The further observations by the electron microscopy (Fig. 4) also affirm that the AR films have a nano-porous with a hierarchical pore distribution. In Fig. 4a, the 5.0 nm pores are formed by F127 and other pores with large sizes (>10 nm) originate from Ph-HSNs. It also shows that the Ph-HSN pores are well bonded to the mesoporous matrix. Fig. 4b show that the thickness of the AR-50 film is 161 nm and the ellipsoid shaped pores with dark contrast are dispersed in the film. The thickness could also be estimated by cross section SEM of AR-30 and AR-10 as 147 nm, 120 nm, respectively. But we find they were not so match with the quarter wavelength of the peak transmittance. That may blame on embossment of films. On the one hand, such nanosize embossment could influence the precise of the thickness measurement. On the other hand, it could enhance the transmittance and reduce the thickness effect on the anti-reflection.

Fig. 4 TEM image (a) and cross-section SEM image (b) of the AR-50 nano-porous films.

3.3 Optical properties and AR performances

The transmittance of the AR films (Fig. 5a) can be progressively improved at least to higher than 95% with different AR coating methods (Table 1), where those containing hierarchic nano-pores from F127 and Ph-HSNs get the maximum up to 98.2%. The diffuse reflectance of the AR films (Fig. 5b) keep consistence with the transmittance (Fig. 5a), where the reflectance: AR-10 < AR-HSN < AR-F127 < AR-SiO₂ < bare glass and the transmittance: AR-10 > AR-HSN > AR-F127 > AR-SiO₂ > bare glass. For the aim to optimize the transmittance at a certain wavelength, λ, the refractive index, n_c, of a SLAR film should be according to the principle of the gradient changing to refractive index (eqn (2)), and the thickness, d, should also be adjusted to be λ/4n_c on the basis of the destructive interference mechanism. Based on the above two AR mechanisms: on the one hand, we adjust the speed of dipping to control some maximum transmittance at about 550 nm;³⁶ on the other hand, we introduce hierarchical F127-removal and Ph-HSNs nano-pores in the films to lower the apparent refractive index than the glass substrate to get good AR performances.³⁷ Fig. 5c shows bare glass, AR-SiO₂, AR-F127 and AR-HSN have stepwise decreasing refractive index, so the transmittance of them is heightened in turn. But the exception is the AR-10 films, which has the highest transmittance among all the films without the support by the lowest refractive index. That may be because it has surface much more rough due to the embossment structures (see Fig. 4, S1 and S2†) deriving from hierarchical F127 and Ph-HSNs nano-pores. According to AFM results (Fig. S1†), all the RMSs were below 20 nm. From SEM (Fig. 4 and S2†), the maximum embossments are about 30 nm, 40 nm and 40 nm for AR-10, AR-30 and AR-50, respectively. Furthermore, AR-10, AR-30 and AR-50 show much rough surface than AR-SiO₂ and AR-F127 with nano-embossment structures. Such embossment structures can trap the incident light by increasing reflecting times to decrease the angle of incidence.³⁸

Fig. 5 UV-vis-NIR transmittance spectra (a and d), reflectance spectra (b and e) and refractive indices (c and f) of different SLAR films.

By proportioning the Ph-HSNs concentration in the precursor sols, the maximum of the transmittance spectra can be well tuned to a specific wavelength, as shown in Fig. 5d. The films AR-10, AR-30 and AR-50 display the peak transmittance at 550 nm, 600 nm and 680 nm, respectively. Generally in dip-coating processes, different withdrawing speeds would lead to different peak transmittance wavelength owing to thickness determined destructive interface effects, where a faster speed is favored to get thicker coatings.³⁴ From cross section SEM images (Fig. 4 and S2†), AR-10, AR-30 and AR-50 coatings are evaluated as 120 nm, 147 nm and 161 nm thick, respectively. Obviously, the addition of Ph-HSNs simultaneously raise the porosity of the AR films, thus the thickness is enlarged distinctly. When applied to the formula of the destructive interface, the peak transmittances of AR-10, AR-30 and AR-50 should be located at 480 nm, 588 nm and 644 nm, respectively. But the experimental peak transmittances (Fig. 5d) locate at much longer wavelengths (510 nm, 620 nm and 670 nm, respectively). These difference can be understood by the effect of light trapping by embossments on the coating surfaces. The embossments may be formed by protuberant nanospheres. They are believed to be able to change large incident angle to small one and also increase optical path for anti-reflection, thus they provide a way like gradient refractive index to reduce reflectance. From AFM images (Fig. S1†) and cross section SEM (Fig. 4 and S2†) and, AR-10, AR-30 and AR-50 show different surface roughness (with the RMS values of 15.24 nm, 10.38 nm and 7.728 nm) and clearly observed embossment. Those maybe lead to an similar effect with the destructive interface for anti-reflection and eventually cause a red-shift peak transmittance wavelengths. Compared to bare glass, the maximum transmittance of such AR films containing Ph-HSNs can facilely preserve a 7.5–7.8% increment at various wavelength. The transmittance improvement can also be verified by the reflectance spectra (Fig. 5d). But the minimums of the reflectance in Fig. 5d locate at different wavelengths (560 nm, 690 nm and 860 nm for AR-10, AR-30 and AR-50, respectively), also related with the rough film surfaces formed by the hierarchic nano-pore and acid silica sol dominated sol–gel processes. The refractive indices of the AR films containing Ph-HSN nano-pores (Fig. 5d) apparently present a continuous dropping down owing to the introduction of Ph-HSNs step by step. All the refractive indices of films are between 1.00 (air) and 1.5 (bare glass), which is benefit to improve the AR performance.

The 45° and 5° specular reflectance spectra (Fig. 6a and b) further verify the transmittance enhancement of the AR films containing different ratio of nano-pores. In Fig. 6a, the 45° specular reflectance is found to be greatly suppressed from higher than 10% (bare glass) to lower than 5% (AR-50) by coating different nano-porous films, especially by coating those containing hierarchical nano-pores. In Fig. 6b, the 45° specular reflectance spectra display a similar evolution with those of the diffuse reflectance spectra (Fig. 5b and e), where bare glass, AR-SiO₂, AR-F127, AR-HSN and AR-10 have the similar minimum reflectance wavelengths around 550 nm, but AR-30 and AR-50 have different minimum reflectance wavelengths. The photograph in Fig. 6c intuitively reveal the excellent AR effect of the AR-50 films, where the left half was bare glass and the right half has been coated with the AR-50 films. Obviously, the logos are strong reflected by the left half bare glass, while the right half suppresses the reflected light a great extent. In a result, we cannot observe the inverted image of the logos in front of the glass but clearly see the logo behind the glass. Therefore, the films containing hierarchic F127 and Ph-HSN nano-pores show good performance as optical AR applications.

Fig. 6 The 45° specular reflectance spectra (a) and the 5° specular reflectance spectra (b) of different SLAR films; (c) photograph of an AR-50 double side coated super white glass.

3.4 Porosity and mechanical properties

The AR effect is the profit of the low refractive indices of the films, principally induced by the large porosity of the films, but it is contradicting to the retaining of mechanical properties such as hardness. This is because more void space would make the structure brittle. Hardness of the AR film is an important factor for its outdoor durability, so the serious deterioration of mechanical properties should be well forfended for the AR treatment of an optical film. Table 2 lists porosity factors and the pencil hardness grades of the nano-porous AR films, where porosity is according to eqn (3). With the introduction of the pore formers (F127 and Ph-HSNs) step by step, the produced AR films get higher and higher porosities up to 58.6%, while the pencil hardness is weakened from 6H to 3H. The adhesive tape applied on the cut did not cause any flakes at the intersections. The edges of the cut also remained smooth as shown in Fig. S4.† Although some of them can just meet partial pencil hardness requirements practically, these HSN-containing AR films with pencil hardness > 3H are competent for some special AR applications due to multi-angle antireflection and other matching properties.²³

Table 2 The pencil hardness grades and porosity factors of the nano-porous AR films

AR films	AR-SiO2	AR-F127	AR-HSN	AR-10	AR-30	AR-50
Porosity	31.1%	38.8%	43.1%	48.4%	50.7%	58.6%
Hardness	6H	5H	4H	4H	3H–4H	3H

---

4. Conclusion

A facile sol–gel process was applied for fabricating high performance hierarchical nano-porous antireflective films. Hollow silica nano-spheres (HSNs) were prepared under base catalyzed condition and subsequently modified by the hydrolysis and condensation of Ph-TES. The excellent nano-granular and dispersibility were obtained via simultaneously embedding Ph-TES modified hollow nano-spherical silica and F-127 template-removal nano-pores in the silica gel films. Primarily on the basis of a graded refractive index mechanism, the porosity of the SLAR films was controlled by the ratio of nano-pores and HSNs and the refractive index was well optimized to be as low as the desired value of 1.24. Secondarily, relying on a thickness-controlled destructive interference of the reflection, the peak transmittance wavelength could be adjusted from 550 nm to 680 nm. Furthermore, the nano-scale surface embossments due to the addition of HSNs was also contributory to the antireflective effect of the films, especially at 45° or 5° incident angle. The pencil hardness of the AR films still kept as better than 3H even when the porosity is higher than 50%. The optimized AR coatings on double sides of glass substrate thus exhibited a maximum transmittance of 98.21%. It qualifies such sol–gel process as an efficacious way to fabricate SLAR optical films with high stabilities, reproducible properties and high performances.

Conflict of interest

The authors declare no competing financial interest.

Acknowledgements

The authors gratefully acknowledge the financial support from the Program for International S&T Cooperation Projects of China (No. 2014DFB50100), Zhejiang Provincial Natural Science Foundation of China (No. LY16E020003), the Fundamental Research Funds for the Central Universities (No. 2016QNA4005; No. 2016FZA4007).

References

1. M. Thomas, Appl. Opt., 1986, 25, 1481–1483.
2. D. M. Bagnall and M. Boreland, Energy Policy, 2008, 36, 4390–4396.
3. G. San Vicente, R. Bayón, N. Germán and A. Morales, Sol. Energy, 2011, 85, 676–680.
4. S. Guldin, P. Kohn, M. Stefik, J. Song, G. Divitini, F. Ecarla, C. Ducati, U. Wiesner and U. Steiner, Nano Lett., 2013, 13, 5329–5335.
5. M. Dai, Y. Wang, M. Pan, S. Lin, G. L. Rempel and Q. Pan, Appl. Surf. Sci., 2014, 289, 209–217.
6. P. Buskens, M. Mourad, N. Meulendijks, R. van Ee, M. Burghoorn, M. Verheijen and E. van Veldhoven, Colloids Surf., A, 2015, 487, 1–8.
7. F. Chi, G. Wei, Q. Zhang, X. Sun, L. Zhang, X. Lu, L. Wang, F. Yi and X. Gao, Appl. Surf. Sci., 2015, 356, 593–598.
8. D. Li, Z. Liu, Y. Wang, Y. Shan and F. Huang, J. Mater. Sci. Technol., 2015, 31, 229–234.
9. C. H. Chen, S. Y. Li, A. S. T. Chiang, A. T. Wu and Y. S. Sun, Sol. Energy Mater. Sol. Cells, 2011, 95, 1694–1700.
10. H. A. Macleod, Thin-film optical filters, CRC press, 2001.
11. H. K. Raut, V. A. Ganesh, A. S. Nair and S. Ramakrishna, Energy Environ. Sci., 2011, 4, 3779–3804.
12. M. A. Aegerter and M. Mennig, Sol–gel technologies for glass producers and users, Springer Science & Business Media, 2013.
13. F. Guillemot, A. Brunet-Bruneau, E. Bourgeat-Lami, T. Gacoin, E. Barthel and J. P. Boilot, Chem. Mater., 2010, 22, 2822–2828.
14. J. Moghal, J. Kobler, J. Sauer, J. Best, M. Gardener, A. A. Watt and G. Wakefield, ACS Appl. Mater. Interfaces, 2012, 4, 854–859.
15. Y. Zhao, J. Wang and G. Mao, Opt. Lett., 2005, 30, 1885–1887.
16. L. Xu and J. He, ACS Appl. Mater. Interfaces, 2012, 4, 3293–3299.
17. L. Zhang, C. Lu, Y. Li, Z. Lin, Z. Wang, H. Dong, T. Wang, X. Zhang, X. Li, J. Zhang and B. Yang, J. Colloid Interface Sci., 2012, 374, 89–95.
18. X. Zhang, P. Lan, Y. Lu, J. Li, H. Xu, J. Zhang, Y. Lee, J. Y. Rhee, K. L. Choy and W. Song, ACS Appl. Mater. Interfaces, 2014, 6, 1415–1423.
19. N. Mizoshita, M. Ishii, N. Kato and H. Tanaka, ACS Appl. Mater. Interfaces, 2015, 7, 19424–19430.
20. W. Joo, Y. Kim, S. Jang and J. K. Kim, Thin Solid Films, 2011, 519, 3804–3808.
21. A. Nakajima, K. Abe, K. Hashimoto and T. Watanabe, Thin Solid Films, 2000, 376, 140–143.
22. T. Li, J. He, L. Yao and Z. Geng, J. Colloid Interface Sci., 2015, 444, 67–73.
23. W. Joo, Y. Kim, S. Jang and J. K. Kim, Thin Solid Films, 2011, 519, 3804–3808.
24. H. P. Ye, X. X. Zhang, Y. L. Zhang, L. Q. Ye, B. Xiao, H. B. Lv and B. Jiang, Sol. Energy Mater. Sol. Cells, 2011, 95, 2347–2351.
25. Q. Ye, S. M. Zhang, Q. Wang, L. H. Yan, H. B. Lv and B. Jiang, RSC Adv., 2014, 4, 35818–35822.
26. J. Zhang, P. J. Lan, J. Li, H. Xu, Q. Wang, X. P. Zhang, L. R. Zheng, Y. H. Lu, N. Dai and W. J. Song, J. Sol–Gel Sci. Technol., 2014, 71, 267–275.
27. Y. Hoshikawa, H. Yabe, A. Nomura, T. Yamaki, A. Shimojima and T. Okubo, Chem. Mater., 2010, 22, 12–14.
28. H. Kamiya and M. Iijima, Sci. Technol. Adv. Mater., 2016, 11, 044304.
29. D. B. Mahadik, R. V. Lakshmi and H. C. Barshilia, Sol. Energy Mater. Sol. Cells, 2015, 140, 61–68.
30. M. Boudot, V. Gaud, M. l. Louarn, M. Selmane and D. Grosso, Chem. Mater., 2014, 26, 1822–1833.
31. Y. Wan and S. Yu, J. Phys. Chem. C, 2008, 112, 3641–3647.
32. M. Kuniyoshi, M. Takahashi, Y. Tokuda and T. Yoko, J. Sol–Gel Sci. Technol., 2006, 39, 175–183.
33. J. Zhu, J. Tang, L. Zhao, X. Zhou, Y. Wang and C. Yu, Small, 2010, 6, 276–282.
34. V. Rao, A. B. Gurav, S. S. Latthe, R. S. Vhatkar, H. Imai, C. Kappenstein, P. B. Wagh and S. C. Gupta, J. Colloid Interface Sci., 2010, 352, 30–35.
35. J. Brinker and G. W. Scherer, Sol–gel science: the physics and chemistry of sol–gel processing, Academic press, 2013.
36. X. Du, Y. Xing, X. Li, H. Huang, Z. Geng, J. He, Y. Wen and X. Zhang, RSC Adv., 2016, 6, 7864–7871.
37. J. Sun, C. Zhang, C. Zhang, R. Ding and Y. Xu, RSC Adv., 2014, 4, 50873–50881.
38. G. Aben, T. Kockelkorel, Y. Li, K. Matloka, M. Plaum, R. de Rijk, H. Schoot and N. Voicu, http://www.dsm.com/corporate/marketsproducts/markets/energy/khepricoat.html, 2015.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20524k

---