**3. Optical properties of textured stainless steel substrate**

#### **3.1 Measurements of optical properties of textured stainless steel substrate**

The total reflection (TR) and diffuse reflection (DR) rates of incident light from the textured substrate were carefully studied by using a Perkin Elmer Lambda 750S spectrometer. It was known that the specula reflection takes place on a smooth surface, and the angle of reflection is the same as the angle of incidence. DR is a phenomenon where an incident beam of light strikes an uneven or granular surface and then scatters in all directions. In Fig. 6, the 6 cm

Fig. 4. The experimental flow charts of lift-off process.

In this study, lift-off and etching processes were used to fabricate the different textures of the 304BA SS substrates. The striped texture was created on the 304BA SS substrate using the lift-off process. After the hard-baking process, a silver (Ag) thin film was deposited on the substrate by e-beam evaporation. An acetone solution was used to remove the residual photo resistor (PR). The depth of the striped texture was controlled by the thickness of the Ag thin film deposited. Four different striped textures were created on the 304BA SS substrates, including period/height: 6/0.1, 6/0.3, 12/0.1 and 12/0.3 μm. Two other types of textured 304BA SS substrate, the ridged-stripe and pyramid texture with 22.5 μm width were created by the etching process. After hard-baking, the 304BA SS substrate was etched by aqua regia (HNO3 : HCl : DI water=1 : 3 : 4). The etching temperature was 28-35℃ with an etching time of 7-12 min. to control the etching depth of the textured 304BA SS substrate. The detail experimental flow charts of lift-off and etching processes are shown in Fig. 4 and

**3. Optical properties of textured stainless steel substrate** 

Fig. 4. The experimental flow charts of lift-off process.

3. Softbake & exposing

1. Substrate cleaning

2. Coating PR

**3.1 Measurements of optical properties of textured stainless steel substrate**

UV light mask

AZ4620

The total reflection (TR) and diffuse reflection (DR) rates of incident light from the textured substrate were carefully studied by using a Perkin Elmer Lambda 750S spectrometer. It was known that the specula reflection takes place on a smooth surface, and the angle of reflection is the same as the angle of incidence. DR is a phenomenon where an incident beam of light strikes an uneven or granular surface and then scatters in all directions. In Fig. 6, the 6 cm

Developer SS 304BA

4. Development

5. Hardbake & metal deposition

Aceton

6. Removing PR

7. Metal deposition

Ag film

Ag film

**2.4 Lift-off and etching process** 

Fig. 5, respectively.

Integrating Sphere is used for diffuse reflectance measurements. Reflectance measurements include total and diffuse reflectance at an incident angle of 8 degrees. Specular reflectance can be calculated from the total and diffuse reflectance measurements. The TR and DR rate of a textured substrate are important indexes when increasing the light trapping efficiency of thin-films solar cells.

Fig. 5. The experimental flow charts of etching process.

Fig. 6. The total reflection and diffuse reflection measured by integrating Sphere.

Enhanced Diffuse Reflection of Light by

Total reflection rate (%)

200 400 600 800 1000 1200

different intervals of textured 430BA SS substrate.

random distribution of light by reflection from a textured surface.

improvement over the uncoated textured 430BA SS substrate.

Wavelength (nm)

Using a Periodically Textured Stainless Steel Substrate 45

0

10

20

Diffuse reflection rate (%)

Fig. 9. The TR and DR rates versus the wavelength curves for raw 430BA SS substrate at

be, resulting in a better diffuse reflection rate. We also investigated the TR and DR rates of the textured 430BA SS substrate with a different interval for samples with a fixed diameter of 4μm. In Fig. 9 we found that the DR rate at the wavelength of 600 nm decreased from 23.1% for a diameter/gap of 4/3 μm textured 430BA SS substrate to 18.9% and 16.1% respectively for a diameter/gap of 4/5μm and 4/7μm textured 430BA SS substrates. The decrease of the DR rate is related to the increase in the interval of the concave substrate. The textured surface of the 430BA SS substrate leads to a lower TR rate compared to a specular surface of raw 430BA SS substrate. The lowering of the TR rate for the textured surface of the 430BA SS substrate can be understood on the basis of (a) the multiple scattering as a result of the multiple reflections from the textured surface of the 430BA SS substrate and a concomitant reduction in light intensity at each reflection due to the finite value of reflectance for 430BA SS, (b) light trapping in the indentations of a highly textured surface. Therefore, the results show that the textured 430BA SS substrate can generate a

It is known that the incident light is reflected back into the cell for a second pass and subsequent passes. This phenomenon results in enhanced absorption in the cell. Thus, a back reflector should possess high reflectance in the solar part of the spectrum, which makes Ag a good candidate. Thus, we also performed the Ag coating on a textured 430BA SS substrate to study the TR and DR rates of incident light. The TR and DR rates versus the wavelength of textured 430BA SS with a silver film thickness of 300 nm are shown in Fig. 10. The peak at around 325 nm can be attributed to the diffuse reflectance spectrum of the deposited Ag film on the surface of the textured 430BA SS substrate (Xiong et al. 2003). In Fig. 10, the DR rate at the 600 nm wavelength are 40.6%, 47.2% and 64.6%, respectively for a diameter/gap of 2/3, 4/3 and 6/3 μm Ag film coated/textured 430BA SS substrate. The DR rate of Ag film coated/textured 430BA SS substrate increased about 2 times in comparison with the uncoated textured 430BA SS substrate (see Fig. 8). Similar results were also observed in Fig. 11 for the Ag film coated/textured 430BA SS substrate with a different interval for samples with a fixed diameter of 4μm. In addition, the TR rate increased to more than 90% for the Ag film coated/textured 430BA SS substrate which was an 80%

30

40

50 (b)

200 400 600 800 1000 1200

Wavelength (nm)

 430BA D4G5 D4G3 D4G7

 430BA D4G5 D4G3 D4G7

### **3.2 Optical properties on periodically textured 430BA stainless steel substrate**

Lately the light trapping properties of textured substrates have attracted substantial interest because of their potential to reduce the thickness of solar cell material. In this study, the different kinds of textured patterns formed on 430BA SS substrate have been proposed for the purpose of trapping light in the application of thin film solar cells. Figure 7 shows the surface morphology of the 430BA SS substrate etched by using aqua regia. It should be noted that the dark and light regions of the OM images indicate the concave structure and the flat surface on the textured 430BA SS substrate, respectively.

In order to understand the optical reflection of a textured 430BA SS substrate, the Perkin Elmer Lambda 900 spectrometer was used to analyze both the TR and DR rates of incident light. The TR and DR rates versus the wavelength curves for the raw and textured 430BA SS substrates are shown in Fig. 8 and Fig. 9, respectively. The "D" and "G" indicate the diameter and the gap for these periodically textured 430BA SS substrates, respectively. It must be noted that the discontinued data line in the wavelength of 850 to 950nm was due to the change in detector, from a PMT to a PbS detector. First, we compared the textured 430BA SS substrates with different diameters of 2, 4 and 6μm and with the same interval of 3μm. In Fig. 8, it was found that the DR rate at the wavelength of 600nm increases substantially, from 4.5% of a raw 430BA SS substrate to 19.7 %, 23.1% and 31.8% for textured 430BA SS substrates with a diameter of 2, 4 and 6μm, respectively. It was evident that for the same areas of analysis, the larger the size of the concave shape, the worse the TR rate would

Fig. 7. The OM images of a concave periodically textured 430BA SS substrate with a diameter/gap of (a) 4/5 μm (b) 4/7 μm (c) 4/3 μm (d) 6/3 μm.

Fig. 8. The TR and DR rates versus the wavelength curves for raw 430BA SS substrate at different diameter of textured 430BA SS substrate.

Lately the light trapping properties of textured substrates have attracted substantial interest because of their potential to reduce the thickness of solar cell material. In this study, the different kinds of textured patterns formed on 430BA SS substrate have been proposed for the purpose of trapping light in the application of thin film solar cells. Figure 7 shows the surface morphology of the 430BA SS substrate etched by using aqua regia. It should be noted that the dark and light regions of the OM images indicate the concave structure and

In order to understand the optical reflection of a textured 430BA SS substrate, the Perkin Elmer Lambda 900 spectrometer was used to analyze both the TR and DR rates of incident light. The TR and DR rates versus the wavelength curves for the raw and textured 430BA SS substrates are shown in Fig. 8 and Fig. 9, respectively. The "D" and "G" indicate the diameter and the gap for these periodically textured 430BA SS substrates, respectively. It must be noted that the discontinued data line in the wavelength of 850 to 950nm was due to the change in detector, from a PMT to a PbS detector. First, we compared the textured 430BA SS substrates with different diameters of 2, 4 and 6μm and with the same interval of 3μm. In Fig. 8, it was found that the DR rate at the wavelength of 600nm increases substantially, from 4.5% of a raw 430BA SS substrate to 19.7 %, 23.1% and 31.8% for textured 430BA SS substrates with a diameter of 2, 4 and 6μm, respectively. It was evident that for the same areas of analysis, the larger the size of the concave shape, the worse the TR rate would

Fig. 7. The OM images of a concave periodically textured 430BA SS substrate with a

200 400 600 800 1000 1200

Wavelength (nm)

 430BA D4G3 D2G3 D6G3

Diffuse reflection rate (%)

Fig. 8. The TR and DR rates versus the wavelength curves for raw 430BA SS substrate at

diameter/gap of (a) 4/5 μm (b) 4/7 μm (c) 4/3 μm (d) 6/3 μm.

 430BA D4G3 D2G3 D6G3

(a) (b) (c) (d)

200 400 600 800 1000 1200

different diameter of textured 430BA SS substrate.

Wavelength (nm)

Total reflection rate (%)

**3.2 Optical properties on periodically textured 430BA stainless steel substrate** 

the flat surface on the textured 430BA SS substrate, respectively.

Fig. 9. The TR and DR rates versus the wavelength curves for raw 430BA SS substrate at different intervals of textured 430BA SS substrate.

be, resulting in a better diffuse reflection rate. We also investigated the TR and DR rates of the textured 430BA SS substrate with a different interval for samples with a fixed diameter of 4μm. In Fig. 9 we found that the DR rate at the wavelength of 600 nm decreased from 23.1% for a diameter/gap of 4/3 μm textured 430BA SS substrate to 18.9% and 16.1% respectively for a diameter/gap of 4/5μm and 4/7μm textured 430BA SS substrates. The decrease of the DR rate is related to the increase in the interval of the concave substrate.

The textured surface of the 430BA SS substrate leads to a lower TR rate compared to a specular surface of raw 430BA SS substrate. The lowering of the TR rate for the textured surface of the 430BA SS substrate can be understood on the basis of (a) the multiple scattering as a result of the multiple reflections from the textured surface of the 430BA SS

substrate and a concomitant reduction in light intensity at each reflection due to the finite value of reflectance for 430BA SS, (b) light trapping in the indentations of a highly textured surface. Therefore, the results show that the textured 430BA SS substrate can generate a random distribution of light by reflection from a textured surface.

It is known that the incident light is reflected back into the cell for a second pass and subsequent passes. This phenomenon results in enhanced absorption in the cell. Thus, a back reflector should possess high reflectance in the solar part of the spectrum, which makes Ag a good candidate. Thus, we also performed the Ag coating on a textured 430BA SS substrate to study the TR and DR rates of incident light. The TR and DR rates versus the wavelength of textured 430BA SS with a silver film thickness of 300 nm are shown in Fig. 10. The peak at around 325 nm can be attributed to the diffuse reflectance spectrum of the deposited Ag film on the surface of the textured 430BA SS substrate (Xiong et al. 2003). In Fig. 10, the DR rate at the 600 nm wavelength are 40.6%, 47.2% and 64.6%, respectively for a diameter/gap of 2/3, 4/3 and 6/3 μm Ag film coated/textured 430BA SS substrate. The DR rate of Ag film coated/textured 430BA SS substrate increased about 2 times in comparison with the uncoated textured 430BA SS substrate (see Fig. 8). Similar results were also observed in Fig. 11 for the Ag film coated/textured 430BA SS substrate with a different interval for samples with a fixed diameter of 4μm. In addition, the TR rate increased to more than 90% for the Ag film coated/textured 430BA SS substrate which was an 80% improvement over the uncoated textured 430BA SS substrate.

Enhanced Diffuse Reflection of Light by

total surface area was 50%.

film and must be carefully investigated in future study.

Fig. 13. The SEM image of Ag film coated/textured 430BA SS substrate.

**3.3 Optical properties on periodically textured 304BA stainless steel substrate** 

Fig. 14 shows the OM images of the striped texture on the 304BA SS substrate. There are four patterns (i.e. the period/depth of 12/0.1, 12/0.3, 6/0.1, and 6/0.3 μm) which are designed and

Using a Periodically Textured Stainless Steel Substrate 47

From Fig. 8, it is evident that the TR and DR rates are not only dependent on the size of the concave shape but also depend on the interval of the concave substrate. Fig. 12(a) and (b) show the TR rate and DR rates as a function of the ratio of the etch pit area to the total surface area for the textured 430BA SS and the Ag film coated/textured 430BA SS, respectively. The ratio of the etch pit area to the total surface area is calculated by the number of pit in the total surface area multiplied by the single pit area divided by the total surface area. The total surface area is the analysis area in the spectrometer, measuring 1 cm2. It is evident that the DR rate increased with the increase effectiveness of the pit regions compared to the smooth regions for both the textured 430BA SS and the Ag film coated/textured 430BA SS. However, the TR rate showed the opposite trend compared with the DR rate and decreased with the increase of the ratio of the etch pit area to total surface area. It is worth noting that once again the TR rate for the the Ag film coated/textured 430BA SS was more than 90% even the ratio of the etch pit area to the

As shown in Figs. 8-11, we found that the increase in TR and DR rates as increase in light wavelength differed in the infrared range. The TR and DR rates clearly increased with the increasing light wavelength of the textured 430BA SS substrate. However, the TR and DR rates didn't increase with the increasing light wavelength of the Ag film coated/textured 430BA SS substrate. Huang et al. indicated that a metal with a lower work function can enhance the Raman signal of diamond films, which is referred to as surface-enhanced Raman scattering (SERS) (Huangbr et al. 2000). A very similar effect, surface-enhanced infrared absorption (SEIRA) was reported to occur with thin metal films (Hartstein et al. 1980, Hatta et al. 1982, Osawa 1997). Moreover, it was reported that the enhancement depends greatly on the morphology of the metal surface (Nishikawa et al. 1993). Fig. 13 shows the SEM image of an Ag film coated/textured 430BA SS substrate. It was found that the surface was covered with Ag particles ranging in size from tens to hundreds of nanometers. Thus, we believe that the difference in increase of the TR and DR rates in the infrared range for the textured 430BA SS and the Ag film coated/textured 430BA SS reflectors are due to the absorption in the infrared range by the Ag films. Further, the surface morphology was related to the thickness of the Ag

Fig. 10. The TR and DR rates versus wavelength curves for Ag film deposited at different diameter of textured 430BA SS substrate.

Fig. 11. The TR and DR rates versus wavelength curves for Ag film deposited at different intervals of textured 430BA SS substrate.

Fig. 12. The TR and DR rates as a function of the ratio of the etch pit area to the total surface area for (a) the textured 430BA SS and (b) the Ag film coated/textured 430BA SS.

Diffuse reflection rate (%)

Diffuse reflection rate (%)

200 400 600 800 1000 1200

200 400 600 800 1000 1200

0 10 20 30 40 50 <sup>0</sup>

Etch pit area/total surface area (%)

Wavelength (nm)

<sup>100</sup> (b)

Wavelength (nm)

 D2G3-Ag D4G3-Ag D6G3-Ag

 D4G3-Ag D4G5-Ag D4G7-Ag

0

20

40

60

Total reflection rate (%)

80

200 400 600 800 1000 1200

200 400 600 800 1000 1200

intervals of textured 430BA SS substrate.

0 10 20 30 40 50 <sup>0</sup>

Etch pit area/total surface area (%)

Wavelength (nm)

diameter of textured 430BA SS substrate.

Wavelength (nm)

 D2G3-Ag D4G3-Ag D6G3-Ag

 D4G3-Ag D4G5-Ag D4G7-Ag

Fig. 11. The TR and DR rates versus wavelength curves for Ag film deposited at different

Total reflection rate (%)

Fig. 12. The TR and DR rates as a function of the ratio of the etch pit area to the total surface

20

40

Diffuse reflection rate (%)

60

80

100

area for (a) the textured 430BA SS and (b) the Ag film coated/textured 430BA SS.

Fig. 10. The TR and DR rates versus wavelength curves for Ag film deposited at different

50

0

(a)

Diffuse reflection rate (%)

20

40

60

Total reflection rate (%)

80

<sup>100</sup> (a)

60

70

80

Total reflection rate (%)

90

<sup>100</sup> (a)

From Fig. 8, it is evident that the TR and DR rates are not only dependent on the size of the concave shape but also depend on the interval of the concave substrate. Fig. 12(a) and (b) show the TR rate and DR rates as a function of the ratio of the etch pit area to the total surface area for the textured 430BA SS and the Ag film coated/textured 430BA SS, respectively. The ratio of the etch pit area to the total surface area is calculated by the number of pit in the total surface area multiplied by the single pit area divided by the total surface area. The total surface area is the analysis area in the spectrometer, measuring 1 cm2. It is evident that the DR rate increased with the increase effectiveness of the pit regions compared to the smooth regions for both the textured 430BA SS and the Ag film coated/textured 430BA SS. However, the TR rate showed the opposite trend compared with the DR rate and decreased with the increase of the ratio of the etch pit area to total surface area. It is worth noting that once again the TR rate for the the Ag film coated/textured 430BA SS was more than 90% even the ratio of the etch pit area to the total surface area was 50%.

As shown in Figs. 8-11, we found that the increase in TR and DR rates as increase in light wavelength differed in the infrared range. The TR and DR rates clearly increased with the increasing light wavelength of the textured 430BA SS substrate. However, the TR and DR rates didn't increase with the increasing light wavelength of the Ag film coated/textured 430BA SS substrate. Huang et al. indicated that a metal with a lower work function can enhance the Raman signal of diamond films, which is referred to as surface-enhanced Raman scattering (SERS) (Huangbr et al. 2000). A very similar effect, surface-enhanced infrared absorption (SEIRA) was reported to occur with thin metal films (Hartstein et al. 1980, Hatta et al. 1982, Osawa 1997). Moreover, it was reported that the enhancement depends greatly on the morphology of the metal surface (Nishikawa et al. 1993). Fig. 13 shows the SEM image of an Ag film coated/textured 430BA SS substrate. It was found that the surface was covered with Ag particles ranging in size from tens to hundreds of nanometers. Thus, we believe that the difference in increase of the TR and DR rates in the infrared range for the textured 430BA SS and the Ag film coated/textured 430BA SS reflectors are due to the absorption in the infrared range by the Ag films. Further, the surface morphology was related to the thickness of the Ag film and must be carefully investigated in future study.

Fig. 13. The SEM image of Ag film coated/textured 430BA SS substrate.
