**2. Upright nanopyramid structured cover glass for solar cells**

### **2.1. Fabrication of upright nanopyramid structures on a glass substrate**

surfaces that enhance the light collection and increasing the effective optical path length of the light within the absorber layer of a solar cell [1]. Various texturing methods have been carried out such as texturing at the rear side [2] or the front side of a solar cell [3] or pre-texturing the thin film solar cell substrates [4, 5] in addition to the wide variety of light management schemes that are based on microscale structures that have been investigated to enhance the power conversion efficiency of solar cells. However, the use of nanostructures for improving the light absorption and trapping in solar cells is a more promising method compared to the traditional microsized surface texturing [6]. This is because of the lower

Nanostructures can be fabricated by various techniques, including electron beam lithography (EBL) [7], laser interference lithography (LIL) [8, 9], nanoimprint lithography (NIL) [10, 11], nanosphere lithography (NSL) [12] and block copolymer lithography (BCPL) [13]. Among them, the UV nanoimprint lithography (UV-NIL) is emerging as a powerful technique for fabricating nanoscale structures on large scale surfaces with simple, high-throughput, low-cost and high-resolution manufacturing capability [14]. Various nanostructures such as nanowires [15], nanorods [16], nanocones [17], nanopyramids [18], nanopillars [19] and metal nanostructures such as nanogrooves [20] and nanoparticle arrays [21] have been extensively studied. Despite their excellent light-trapping properties, texturing the active solar cell layer or introducing metal nanostructures within the cell results in poor charge carrier collection due to increased surface recombination. Fang Jiao et al. [22] demonstrated that the imprinting of moth-eye-like structures on the front side of monocrystalline Si solar cell surface enhanced the conversion efficiency by 19% compared to the planar solar cell through coupling the inci-

This chapter describes an approach of surface texturing which is different from other reported methods such as texturing the active material or using metal nanostructures. It is expected that nanopyramids coating approach might be enhanced solar cell performance without introducing additional surface recombination and excellent solar cell self-cleaning functionality.

Solar cell modules are installed in an outdoor environment for the vast majority of applications. Therefore, whatever the type of solar cell, glass is commonly incorporated as an encapsulation for preventing damage from dust, moisture and external shock [23, 24]. However, some of the incident light onto the solar cells will be lost through optical reflection due to the refractive index mismatch between the air and cover glass and through scattering or absorption by contaminants [25, 26]. In Section 2.3, it is shown that the amount of the incident light reaching the solar cell could be enhanced by incorporating antireflective and light-scattering nanostructures at the cover glass surface. Moreover, it is also demonstrated in Section 2.4 that the nanostructured cover glass has self-cleaning property and efficiently maintains the perfor-

The oblique light-scattering effect offered by the nanopyramids improves the light harvesting of the solar cells as a result of prolonged optical path length within the solar cells and thus, increasing the conversion efficiency [27, 28]. Several other research groups [29–31] have studied the use of cover glass that combines the antireflective and scattering effects with selfcleaning properties and examined their influence on the overall efficiency of the solar cells.

level of induced damage and the ease of coating different surfaces and materials.

dent light into the absorber layer.

26 Emerging Solar Energy Materials

mance of solar cells in harsh environments.

The UV nanoimprint lithography process is used for the replication of the upright nanopyramid (UNP) structures on glass substrates. The process flow diagram of the UV imprint is shown in **Figure 1**. First, the periodic inverted nanopyramid structures were formed on the Si master mold using laser interference lithography and subsequent pattern transfer process using reactive ion etching followed by KOH wet etching. Details of the fabrication process of the master mold are described in Ref. [9]. A UV curable resist (OrmoStamp) was spincoated onto a glass substrate and afterward the substrate was placed over the Si mold with the inverted nanopyramids inside the imprint tool. An imprint pressure of up to 4 mbar was applied to transfer the patterns with a

**Figure 1.** The schematic diagram of the overall fabrication process of upright nanopyramid structures on the glass.

UV light illumination of wavelength 365 nm. The upright nanopyramid pattern was successfully replicated from the Si mold onto the glass substrate with high fidelity. After the UV nanoimprint process, F13-TCS-based SAM was coated onto the upright nanopyramid patterned glass substrate in order to increase the hydrophobicity of the surface. More details of the UV nanoimprint process parameters and the tools which were employed can be found in Ref. [32].

The surface morphologies of the inverted and upright nanopyramid structures were examined by using scanning electron microscope (SEM) (JEOL 7000F FE-SEM) and atomic force microscope (AFM, DI3000). **Figure 2(a)** and **(c)** presents the SEM images and AFM image of the Si master mold with inverted nanopyramid structures, respectively. **Figure 2(b)** and **(d)** shows the SEM images and AFM images of the upright nanopyramid structured glass replicated from the master mold, respectively. As illustrated in **Figure 2(b)**, the inverted nanopyramid patterns on the Si master mold were transferred onto the UV curable resist coated glass substrate without any distortion and deformation using UV nanoimprint lithography. This is also confirmed in the AFM image in **Figure 2(d)**. As shown in **Figure 2**, the 450 nm wide and 310 nm high UNP structures with 125 nm separation were replicated uniformly over an area of 10x10 mm after the imprint process.

crystal orientation and resistivity of 0.5–1.0 Ωcm was used as the substrate. After piranha cleaning and 1:10 dilute HF dipping, a 200 nm thermal oxide was grown on the wafer using quartz tube furnace and dry/wet oxidation was done at 1000°C. The 10 mm × 10 mm individual cells were defined by photolithography, and buffered HF etching was performed to isolate the individual cells by opening windows in the oxide. The wafer backside was doped with boron dopant (B202 form Filmtronics) to create the back surface field effect. The emitter junction was formed by spin-on phosphorus doping processes using P509 dopant from Filmtronics. The diffusion was performed in a quartz tube furnace at 950°C for 30 min in a

**Figure 3.** The schematic representation of monocrystalline Si solar cell fabrication process.

removed using the diluted 10% HF solution. The 300-nm thick aluminum front and back contact were deposited by DC sputtering. In creating the top contact, top contact patterns were

The total and diffuse transmittance of UNP patterned glass or unpatterned bare glass were measured using a UV–visible spectrophotometer at room temperature with an integrating sphere over the wavelength range of 300–1200 nm. **Figure 4(a)** illustrates the comparison between the total and diffuse transmittance of the glass substrates with and without UNP patterns, which were measured using an integrating sphere with the incoming light entered from the patterned glass substrate side. As shown in **Figure 4(a)**, the total transmittance of the

environment. The phosphosilicate glass (PSG) on the wafer surface was

Nanopyramid Structures with Light Harvesting and Self-Cleaning Properties for Solar Cells

http://dx.doi.org/10.5772/intechopen.75314

29

20% O<sup>2</sup>

and 80% N<sup>2</sup>

defined by photolithography before the metal deposition.

**2.3. Optical properties and device performance**

**Figure 2.** 30°-tilted view SEM images of (a) inverted nanopyramid Si master mold and (b) upright nanopyramid structured on the glass, and the inset images are the cross-sectional views of SEM images. Three-dimensional AFM images (c) inverted nanopyramid Si master mold and (d) upright nanopyramid structured glass.

#### **2.2. Solar cell fabrication**

Monocrystalline Si solar cells were fabricated by the process shown in **Figure 3**. A single-sidepolished, Czochralski (CZ) grown, 350 μm thick, boron doped p-type silicon wafer with <100> Nanopyramid Structures with Light Harvesting and Self-Cleaning Properties for Solar Cells http://dx.doi.org/10.5772/intechopen.75314 29

**Figure 3.** The schematic representation of monocrystalline Si solar cell fabrication process.

crystal orientation and resistivity of 0.5–1.0 Ωcm was used as the substrate. After piranha cleaning and 1:10 dilute HF dipping, a 200 nm thermal oxide was grown on the wafer using quartz tube furnace and dry/wet oxidation was done at 1000°C. The 10 mm × 10 mm individual cells were defined by photolithography, and buffered HF etching was performed to isolate the individual cells by opening windows in the oxide. The wafer backside was doped with boron dopant (B202 form Filmtronics) to create the back surface field effect. The emitter junction was formed by spin-on phosphorus doping processes using P509 dopant from Filmtronics. The diffusion was performed in a quartz tube furnace at 950°C for 30 min in a 20% O<sup>2</sup> and 80% N<sup>2</sup> environment. The phosphosilicate glass (PSG) on the wafer surface was removed using the diluted 10% HF solution. The 300-nm thick aluminum front and back contact were deposited by DC sputtering. In creating the top contact, top contact patterns were defined by photolithography before the metal deposition.

#### **2.3. Optical properties and device performance**

UV light illumination of wavelength 365 nm. The upright nanopyramid pattern was successfully replicated from the Si mold onto the glass substrate with high fidelity. After the UV nanoimprint process, F13-TCS-based SAM was coated onto the upright nanopyramid patterned glass substrate in order to increase the hydrophobicity of the surface. More details of the UV nanoimprint

The surface morphologies of the inverted and upright nanopyramid structures were examined by using scanning electron microscope (SEM) (JEOL 7000F FE-SEM) and atomic force microscope (AFM, DI3000). **Figure 2(a)** and **(c)** presents the SEM images and AFM image of the Si master mold with inverted nanopyramid structures, respectively. **Figure 2(b)** and **(d)** shows the SEM images and AFM images of the upright nanopyramid structured glass replicated from the master mold, respectively. As illustrated in **Figure 2(b)**, the inverted nanopyramid patterns on the Si master mold were transferred onto the UV curable resist coated glass substrate without any distortion and deformation using UV nanoimprint lithography. This is also confirmed in the AFM image in **Figure 2(d)**. As shown in **Figure 2**, the 450 nm wide and 310 nm high UNP structures with 125 nm separation were replicated uniformly over an area

Monocrystalline Si solar cells were fabricated by the process shown in **Figure 3**. A single-sidepolished, Czochralski (CZ) grown, 350 μm thick, boron doped p-type silicon wafer with <100>

**Figure 2.** 30°-tilted view SEM images of (a) inverted nanopyramid Si master mold and (b) upright nanopyramid structured on the glass, and the inset images are the cross-sectional views of SEM images. Three-dimensional AFM

images (c) inverted nanopyramid Si master mold and (d) upright nanopyramid structured glass.

process parameters and the tools which were employed can be found in Ref. [32].

of 10x10 mm after the imprint process.

28 Emerging Solar Energy Materials

**2.2. Solar cell fabrication**

The total and diffuse transmittance of UNP patterned glass or unpatterned bare glass were measured using a UV–visible spectrophotometer at room temperature with an integrating sphere over the wavelength range of 300–1200 nm. **Figure 4(a)** illustrates the comparison between the total and diffuse transmittance of the glass substrates with and without UNP patterns, which were measured using an integrating sphere with the incoming light entered from the patterned glass substrate side. As shown in **Figure 4(a)**, the total transmittance of the

**Figure 4.** (a) Measured total and diffuse transmittance spectra of the bare glass and the UNP patterned glass as a function of wavelength and (b) the optical haze spectra of the bare glass and the UNP patterned glass as a function of wavelength. Photographs of diffracted light patterns of the corresponding samples obtained from the green diode laser with a wavelength of 532 nm are also displayed in the inset.

The cross-sectional electric field distributions at different wavelength obtained for the incident light propagating from air to the glass substrate with and without upright nanopyramid structure are depicted in **Figure 6**. As shown in **Figure 6**, upright nanopyramid structures show wide angular range light-scattering patterns, especially in the wavelength range below 600 nm and provide oblique light transmission from air and the glass while there is no scattering of light for bare cover glass. These results demonstrate that the upright nanopyramid structured

**Figure 6.** The cross-sectional electric field distribution profiles at different wavelength by FDTD analysis: (a) upright nanopyramid structured glass with a period of 600 nm, base of 500 nm and height of 400 nm and (b) bare flat glass.

**Figure 5.** FDTD simulation model layout of the UNP structured glass substrate: (a) perspective view and (b) XZ view.

Nanopyramid Structures with Light Harvesting and Self-Cleaning Properties for Solar Cells

http://dx.doi.org/10.5772/intechopen.75314

31

UNP patterned glass was slightly lower than that of the bare glass in the wavelength range of 450–800 nm which may be caused by the diffraction losses due to the higher order diffracted waves [33, 34]. However, the diffuse transmittance of the UNP patterned glass was increased up to 24% in the visible wavelength region due to higher orders of diffracted waves in the transmission, whereas the bare glass substrate shows almost no diffuse transmittance over a wide wavelength range as illustrated in **Figure 4(a)**.

The haze value (*H*), which is determined by the ratio of the diffuse transmittance (*T*d) to the total transmittance (*Tt* ), that is*, H = T*d/*Tt* , indicates the light-scattering properties of the samples. When the incident light passed through the bare cover glass, the *H* value is close to zero as shown in **Figure 4(b)**. In contrast, the *H* value is significantly increased for patterned glass, especially, in the wavelength range 300–600 nm, which signifies that strong light scattering is achieved by UNP structured glass. This light-scattering behavior can also be confirmed in the insets of **Figure 4(b)**. For the bare cover glass, there is almost no light diffraction, whereas the UNP patterned glass shows high order diffraction patterns using a green diode laser at a wavelength of 532 nm. This scattering effect will result in changes in the propagation direction of light from normal to the oblique incidence in the solar cell. As a result, the optical path length of the incident light is elongated, and hence the light absorption in the active layer of the solar cell is also improved by the patterned glass. Indeed, the high haze optical property due to the light scattering effect would positively enhance the power conversion efficiency of the solar cells with the UNP patterned glass compared to the bare cover glass [35–37]. This is especially important for thin film devices.

Numerical simulations were performed using the finite-difference time-domain (FDTD) method by Lumerical solutions Inc. to illustrate how the incoming light couples with and without upright nanopyramid structure. **Figure 5** shows the FDTD simulation model layout of the UNP structured glass substrate in perspective view and XZ view. Perfectly matched layers (PML) and periodic boundary conditions were used in the perpendicular and horizontal directions.

Nanopyramid Structures with Light Harvesting and Self-Cleaning Properties for Solar Cells http://dx.doi.org/10.5772/intechopen.75314 31

**Figure 5.** FDTD simulation model layout of the UNP structured glass substrate: (a) perspective view and (b) XZ view.

The cross-sectional electric field distributions at different wavelength obtained for the incident light propagating from air to the glass substrate with and without upright nanopyramid structure are depicted in **Figure 6**. As shown in **Figure 6**, upright nanopyramid structures show wide angular range light-scattering patterns, especially in the wavelength range below 600 nm and provide oblique light transmission from air and the glass while there is no scattering of light for bare cover glass. These results demonstrate that the upright nanopyramid structured

UNP patterned glass was slightly lower than that of the bare glass in the wavelength range of 450–800 nm which may be caused by the diffraction losses due to the higher order diffracted waves [33, 34]. However, the diffuse transmittance of the UNP patterned glass was increased up to 24% in the visible wavelength region due to higher orders of diffracted waves in the transmission, whereas the bare glass substrate shows almost no diffuse transmittance over a

**Figure 4.** (a) Measured total and diffuse transmittance spectra of the bare glass and the UNP patterned glass as a function of wavelength and (b) the optical haze spectra of the bare glass and the UNP patterned glass as a function of wavelength. Photographs of diffracted light patterns of the corresponding samples obtained from the green diode laser with a

The haze value (*H*), which is determined by the ratio of the diffuse transmittance (*T*d) to the

ples. When the incident light passed through the bare cover glass, the *H* value is close to zero as shown in **Figure 4(b)**. In contrast, the *H* value is significantly increased for patterned glass, especially, in the wavelength range 300–600 nm, which signifies that strong light scattering is achieved by UNP structured glass. This light-scattering behavior can also be confirmed in the insets of **Figure 4(b)**. For the bare cover glass, there is almost no light diffraction, whereas the UNP patterned glass shows high order diffraction patterns using a green diode laser at a wavelength of 532 nm. This scattering effect will result in changes in the propagation direction of light from normal to the oblique incidence in the solar cell. As a result, the optical path length of the incident light is elongated, and hence the light absorption in the active layer of the solar cell is also improved by the patterned glass. Indeed, the high haze optical property due to the light scattering effect would positively enhance the power conversion efficiency of the solar cells with the UNP patterned glass compared to the bare cover glass [35–37]. This is

Numerical simulations were performed using the finite-difference time-domain (FDTD) method by Lumerical solutions Inc. to illustrate how the incoming light couples with and without upright nanopyramid structure. **Figure 5** shows the FDTD simulation model layout of the UNP structured glass substrate in perspective view and XZ view. Perfectly matched layers (PML) and periodic boundary conditions were used in the perpendicular and horizontal directions.

, indicates the light-scattering properties of the sam-

wide wavelength range as illustrated in **Figure 4(a)**.

wavelength of 532 nm are also displayed in the inset.

especially important for thin film devices.

), that is*, H = T*d/*Tt*

total transmittance (*Tt*

30 Emerging Solar Energy Materials

**Figure 6.** The cross-sectional electric field distribution profiles at different wavelength by FDTD analysis: (a) upright nanopyramid structured glass with a period of 600 nm, base of 500 nm and height of 400 nm and (b) bare flat glass.

glass enhances the diffuse transmittance of the cover glass or any substrate used for solar cells applications. Thus, these structures can lead to the power conversion efficiency enhancement of encapsulated solar cells due to the improved light harvesting in the absorption layer of the solar cells caused by the combined effects of strong light scattering and antireflection coating [37, 38].

In order to verify the effect of the periodic upright nanopyramid patterns, the patterned glass and the bare glass substrate were employed as a cover encapsulation glass on the monocrystalline Si solar cell. **Figure 7(a)** shows the current density-voltage characteristics of the encapsulated monocrystalline Si solar cell with and without UNP patterned cover glass. The monocrystalline Si solar cell performances are summarized in **Table 1**.

As shown in **Figure 7(b)**, the solar cells with upright nanopyramid patterned cover glass exhibited improved EQE values in comparison with the bare cover glass, particularly in the wavelength region of 400–600 nm. This is due to the increased photogenerated carriers generated by its higher haze properties. This result was precisely matched with the optical haze value result shown in **Figure 4**. From these results, the periodic upright nanopyramid patterned glass offers a better-graded index medium to the incident light compared to the bare glass [25]. Therefore, the patterned cover glass can reduce the Fresnel reflectance and scatter more incident light into the solar cells' emitter area and prolong the optical path length, there-

**Table 1.** Device characteristics of encapsulated monocrystalline Si solar cells with and without the UNP patterned cover

**Monocrystalline Si solar cells VOC (V) JSC (mAcm−2) FF (%) PCE (%)** Without cover glass 0.58 34.38 59.82 11.93 Bare cover glass 0.58 31.60 55.23 10.12 UNP patterned cover glass 0.58 32.39 59.76 11.23

Nanopyramid Structures with Light Harvesting and Self-Cleaning Properties for Solar Cells

http://dx.doi.org/10.5772/intechopen.75314

33

In real outdoor environments, the cover glass layer of the solar cell can easily be contaminated by dust particles which interfere with the incident light directed into the solar cell active layer and thus reducing the solar cells' performance. Therefore, the nanopyramids covered glass encapsulation has the advantages of acting as an antireflection layer and as self-cleaning surface and will maintain the solar cell performance under real outdoor environment condition [40, 41]. The water wetting behaviors of the samples with different morphologies were investigated. **Figure 8** shows (a) the photographs of a water droplet on (I) the bare glass, (II) UNP structured glass and (III) SAM-coated UNP structured glass and (b) sequential photographs of water droplet self-cleaning process for (I) the bare cover glass and (II) UNP structured glass. As shown in **Figure 8(a)**, the bare glass exhibited a hydrophilic surface with a water contact angle (θCA) of ~36° while UNP patterned glass showed a hydrophobic behavior with a θCA value of ~112°. This hydrophobic behavior is associated to the enhanced surface roughness of the UNP patterned glass, which can be demonstrated by the Cassie-Baxter Equation [42]. Moreover, F13- TCS-based SAM was coated onto the UNP patterned glass in order to enhance its hydrophobic surfaces. In this case, the contact angle of the SAM-coated UNP glass was increased to 132° as shown in **Figure 8(a)**. These contact angle (θCA) values are comparatively lower than those reported with superhydrophobicity (i.e., θCA > 150°) in previous studies [43, 44]. However, it can be observed that the black dust particles on the surface of UNP patterned glass were cleared away by the rolling down water droplets without any remaining dust particles or water droplets at the surface, as shown in **Figure 8(b)**. In contrast, the dust particles remained on the bare glass even with rolling down water droplets. Thus, the dust particles partially remained especially at the edge of the bare glass. The conclusion is that the UNP patterned cover glass has dual functionality of light-harvesting and self-cleaning properties and would enhance the

fore, improving the light trapping and increasing the overall conversion efficiency.

**2.4. Surface wettability and self-cleaning behaviors**

glass and the bare cover glass.

practicability of solar cells in real outdoor environments.

There was no significant change in the open circuit voltage (VOC), but a significant enhancement in the short-circuit current density (JSC) was observed as expected. The fill factor (FF) of monocrystalline Si solar cell was slightly enhanced from 55.23 to 59.76% with UNP patterned cover glass. Such improvement in FF could be attributed to the enhanced density of free carriers [39] induced by the increased number of photons entering the active layer of the solar cell and reducing the effective series resistance. The value of JSC for the monocrystalline Si solar cell without cover glass was 34.38 mAcm−2. This value was decreased to 31.60 mAcm−2 with a bare cover glass. This reduction in JSC indicates that the cover glass reduces the number of photons entering the active layer of the solar cell through reflection and absorption processes. However, by replacing the bare glass with UNP patterned cover glass, JSC value was increased to 32.39 mAcm−2 for encapsulated monocrystalline Si solar cell. Hence, the use of the upright nanopyramid patterned glass as a cover glass is an effective way to improve the power conversion efficiency (PCE). The encapsulated monocrystalline Si solar cell with patterned glass efficiency has increased by 10.97% compared to the encapsulated monocrystalline Si solar cell with bare cover glass. The experiment was repeated with commercially manufactured polycrystalline solar cell and similar trend was observed. This enhancement is mainly due to the strong light scattering effect via the upright nanopyramid structures.

**Figure 7.** (a) Current density-voltage characteristics and (b) external quantum efficiency (EQE) spectra of encapsulated monocrystalline Si solar cell with and without the upright nanopyramid patterned cover glass and bare cover glass.


**Table 1.** Device characteristics of encapsulated monocrystalline Si solar cells with and without the UNP patterned cover glass and the bare cover glass.

As shown in **Figure 7(b)**, the solar cells with upright nanopyramid patterned cover glass exhibited improved EQE values in comparison with the bare cover glass, particularly in the wavelength region of 400–600 nm. This is due to the increased photogenerated carriers generated by its higher haze properties. This result was precisely matched with the optical haze value result shown in **Figure 4**. From these results, the periodic upright nanopyramid patterned glass offers a better-graded index medium to the incident light compared to the bare glass [25]. Therefore, the patterned cover glass can reduce the Fresnel reflectance and scatter more incident light into the solar cells' emitter area and prolong the optical path length, therefore, improving the light trapping and increasing the overall conversion efficiency.

#### **2.4. Surface wettability and self-cleaning behaviors**

glass enhances the diffuse transmittance of the cover glass or any substrate used for solar cells applications. Thus, these structures can lead to the power conversion efficiency enhancement of encapsulated solar cells due to the improved light harvesting in the absorption layer of the solar cells caused by the combined effects of strong light scattering and antireflection coating

In order to verify the effect of the periodic upright nanopyramid patterns, the patterned glass and the bare glass substrate were employed as a cover encapsulation glass on the monocrystalline Si solar cell. **Figure 7(a)** shows the current density-voltage characteristics of the encapsulated monocrystalline Si solar cell with and without UNP patterned cover glass. The

There was no significant change in the open circuit voltage (VOC), but a significant enhancement in the short-circuit current density (JSC) was observed as expected. The fill factor (FF) of monocrystalline Si solar cell was slightly enhanced from 55.23 to 59.76% with UNP patterned cover glass. Such improvement in FF could be attributed to the enhanced density of free carriers [39] induced by the increased number of photons entering the active layer of the solar cell and reducing the effective series resistance. The value of JSC for the monocrystalline Si solar cell without cover glass was 34.38 mAcm−2. This value was decreased to 31.60 mAcm−2 with a bare cover glass. This reduction in JSC indicates that the cover glass reduces the number of photons entering the active layer of the solar cell through reflection and absorption processes. However, by replacing the bare glass with UNP patterned cover glass, JSC value was increased to 32.39 mAcm−2 for encapsulated monocrystalline Si solar cell. Hence, the use of the upright nanopyramid patterned glass as a cover glass is an effective way to improve the power conversion efficiency (PCE). The encapsulated monocrystalline Si solar cell with patterned glass efficiency has increased by 10.97% compared to the encapsulated monocrystalline Si solar cell with bare cover glass. The experiment was repeated with commercially manufactured polycrystalline solar cell and similar trend was observed. This enhancement is mainly due to the strong light scattering effect via the

**Figure 7.** (a) Current density-voltage characteristics and (b) external quantum efficiency (EQE) spectra of encapsulated monocrystalline Si solar cell with and without the upright nanopyramid patterned cover glass and bare cover glass.

monocrystalline Si solar cell performances are summarized in **Table 1**.

[37, 38].

32 Emerging Solar Energy Materials

upright nanopyramid structures.

In real outdoor environments, the cover glass layer of the solar cell can easily be contaminated by dust particles which interfere with the incident light directed into the solar cell active layer and thus reducing the solar cells' performance. Therefore, the nanopyramids covered glass encapsulation has the advantages of acting as an antireflection layer and as self-cleaning surface and will maintain the solar cell performance under real outdoor environment condition [40, 41]. The water wetting behaviors of the samples with different morphologies were investigated. **Figure 8** shows (a) the photographs of a water droplet on (I) the bare glass, (II) UNP structured glass and (III) SAM-coated UNP structured glass and (b) sequential photographs of water droplet self-cleaning process for (I) the bare cover glass and (II) UNP structured glass.

As shown in **Figure 8(a)**, the bare glass exhibited a hydrophilic surface with a water contact angle (θCA) of ~36° while UNP patterned glass showed a hydrophobic behavior with a θCA value of ~112°. This hydrophobic behavior is associated to the enhanced surface roughness of the UNP patterned glass, which can be demonstrated by the Cassie-Baxter Equation [42]. Moreover, F13- TCS-based SAM was coated onto the UNP patterned glass in order to enhance its hydrophobic surfaces. In this case, the contact angle of the SAM-coated UNP glass was increased to 132° as shown in **Figure 8(a)**. These contact angle (θCA) values are comparatively lower than those reported with superhydrophobicity (i.e., θCA > 150°) in previous studies [43, 44]. However, it can be observed that the black dust particles on the surface of UNP patterned glass were cleared away by the rolling down water droplets without any remaining dust particles or water droplets at the surface, as shown in **Figure 8(b)**. In contrast, the dust particles remained on the bare glass even with rolling down water droplets. Thus, the dust particles partially remained especially at the edge of the bare glass. The conclusion is that the UNP patterned cover glass has dual functionality of light-harvesting and self-cleaning properties and would enhance the practicability of solar cells in real outdoor environments.

**Figure 8.** Photographs of (a) a water droplet on (I) bare glass, (II) UNP glass and (II) SAM-coated UNP glass and (b) sequential photographs of a self-cleaning process for (I) the bare glass and (II) UNP glass. θCA is the water contact angle.
