1. Introduction

A sudden change in refractive index between two media will decrease the transmittance of a light beam propagating through these two media [1]. Micro/nano-structured surfaces with a period shorter than the light wavelengths can enhance the light absorption or transmission, while suppress reflection [2,3]. These structured surfaces can improve the performance of materials for various applications, such as solar cells [4,5], light emitting diodes [6,7], optical sensors [8] and laser desorption/ionization mass spectrometry [9]. Increasing the transparency of glass plays an important role in optical sensing, imaging, and other fields [10–13]. However, it is worth noting that glass transparency and light transmittance are two different concepts. Glass transparency means one can see through the glass better. The higher transparency the glass is, the more visible the objects behind it are (clear glass, CG). Light transmittance is the ability of light to pass through a medium (including diffuse transmission). Transmittance is the ratio of the transmitted flux to the incident flux. Glass with good light transmittance but poor transparency is frosted glass (FG) and has an excellent decorative effect. FG can increase the concealment of buildings. It is widely used in our daily lives and various industries.

Surfaces with nano/micro-cube [14], micro-cones [3,15], pyramids [16], and other porous structures [17] have been shown to increase light transmission. Usually, those structures are achieved by various methods, such as e-beam lithography [18,19], nanoimprint [20,21], sol-gel coating [22], reactive ion etching (RIE) [10], and inductively coupled plasma (ICP) etching [23]. However, most reported methods require complex processes and are of high cost. Hydrothermal synthesis, as a cheap and facile method to process nano-materials, has attracted tremendous interest [24–26]. It has been demonstrated that by using hydrothermal method, a superhydrophobic surface with good light transmission was fabricated on glass [24]. An anti-reflection glass surface with self-cleaning and anti-dust was also processed [25]. By using this method, only CG was studied but not FG [24,25].

In this paper, a facile hydrothermal method is used to enhance light transmittance of soda lime glass. By using this method, we can get not only CG, but also FG. The fabrication of FG or CG is highly dependent on the baking temperature and time. As a result, by changing the baking temperature and time, we process two types of structures on soda lime glass, making glass either frosted or more transparent. The transmittance of the FG can reach more than 95%, and the CG can reach 96.7% over a wide wavelength range. We show that the FG surfaces is covered with irregular micro-craters, while the CG surfaces is covered with extensive nanoflakes.

2. Experimental details

A soda lime glass (72.5% SiO₂, 13.7% Na₂O, 9.80% CaO, 3.5% MgO, 0.4% Al₂O₃, and 0.1% K₂O; Corning) is cleaned with ethanol and deionized water in an ultrasonic cleaner, 10 minutes each time. The glass substrate is placed in a 30 mL Teflon container with 20 mL ammonium hydroxide (NH₄OH) of 3 mol/L and then sealed by a stainless steel autoclave. The autoclave is placed in an oven (Quincy lab, Inc, America, model 10 lab oven) and baked at an appropriate temperature (100$\;\ ^\circ\textrm{C}$, 160 $^\circ\textrm{C}$) for a given time (4 h, 8 h, 12 h, 24 h). It is naturally cooled to room temperature, and then the glass substrate is taken out for further characterization. The optical transmittance is determined by a UV-visible scanning spectrophotometer (Lambda 365, PerkinElmer, America) from 350 to 1100 nm wavelengths. The surfaces morphology is observed by a scanning electron microscopy (SEM) (SEM, S-4100, Hitachi, Japan).

3. Results and discussion

Figure 1(a) shows two detection approaches of the transmittance used in our experiment. The integrating sphere can collect the scattered part of the transmitted light. When the integrating sphere was used, all the light through the sample will be collected by the detector. In our experiment, by using hydrothermal method, we obtained two types of structures on soda lime glass, achieving CG and FG. For normal soda lime glass (NG), the light transmittance can reach 91% in visible range, which means most light pass through it, as shown in Fig. 1(b). For CG, the transmittance is improved, due to suppression of the reflection (as shown in Fig. 1(b)). For FG, most of the light can still pass through the glass by forward scattering, only less light is reflected back (as shown in Fig. 1(b),). The fabrication of FG or CG is highly dependent on the treating temperature and time. The optical properties of both FG and CG are characterized by the two approaches shown in Fig. 1(a).

 

Fig. 1. (a) Two detection approaches of the transmittance used in our experiment. (b) The different transmission and reflection situation of different glass (NG, CG, FG). The yellow arrows represent the coming light and the light pass through the glass. The green arrows represent the reflected or scattered light. The size of the sun or the thickness of the arrows represents the strength of the light.

Download Full Size | PDF

After optimizing treatment parameters, the CG is obtained by baking bare glass sample in 3 mol/L NH₄OH solution at a temperature of 100 $^\circ{\textrm{C}}$ for 24 h. The optical properties and real photos of the samples are shown in Fig. 2. The transmittance in 350–1100 nm wavelength range increased from approximately 91% for the soda lime glass substrate (black line) to greater than 96% for the CG (purple line), as shown in Fig. 2(a). Furthermore, compared to the prepared CG (red rectangle), we can see clearly that the surfaces reflection on the NG (black rectangle) is much stronger, as shown in Fig. 2(c) and Fig. 2(e). Figure 2(d) shows that after treatment, the transparency of the glass did not decrease. We can see the logos and letters through the sample. More importantly, the logos and letters under the prepared CG are much clear, which means the transmittance of the CG is really increased. By comparing Fig. 2(a) and Fig. 2(b) that are obtained without using integrating sphere and using integrating sphere, respectively, we can find that no matter using the integrating sphere or not, the treated sample can increase the transmittance without affecting its transparency under this experiment condition. This result means the light is straight pass through the glass, with negligible scattering. The effect of treating time on the glass transmittance is shown in Fig. 2(a) and Fig. 2(b). As the treatment time increasing, the transmittance of the glass increased except for 12 h. Furthermore, compared to the long-wavelength band, the transmittance in short wavelength is reduced with increasing treatment time.

 

Fig. 2. The optical properties of the CG. Transmittance of the soda lime glass before and after treated at 100 $^\circ{\textrm{C}}$ (a) without the integrating sphere. (b) with the integrating sphere. (c) a photograph of the NG and prepared CG (Baked at 100$\;\ ^\circ{\textrm{C}}$ for 24 h) under the incandescent lamp. A strong surface reflection on the NG is showed obviously. (d) A comparison between the NG and treated CG. The logos look much clear through the treated CG. The black rectangle in (c) and (d) represent the NG. The red rectangle in (c) and (d) represent the CG. (e) The reflectance of the soda lime glass before and after treated at 100 $^\circ{\textrm{C}}$ measured with the integrating sphere.

Download Full Size | PDF

Figure 3 shows the SEM images of the glass substrates prepared at 100 $^\circ{\textrm{C}}$ with different treatment time. It can be seen clearly that the glass surface is relatively smooth before treatment, as shown in Fig. 3(a). After treated for 4 hours, many small nanoflakes emerged on the surface, but the structure size is relatively small, as shown in Fig. 3(b). It is well known that when the period of the micro/nano-structure covered on the surface is shorter than the wavelengths of the light, it will enhance the light absorption or transmission of the surface [2,3]. As a result, we can see that for samples treated 4 h or 8 h, the short-wavelength have a bigger enhancement compared to the long-wavelength. As the treating time increasing, the size of the nanoflakes increased, resulting in a decreased transmittance in short wavelength, while increased transmittance in long-wavelength.

 

Fig. 3. The top-view SEM images of CG prepared at 100 $^\circ{\textrm{C}}$ with different treating time: (a) bare glass; (b) 4 h; (c) 8 h; (d) 24 h.

Download Full Size | PDF

Researchers found that the formation of the rough glass surfaces during the hydrothermal processing may undergo three steps, including etching, phase separation, and leaching [25]. Figure 4 shows the SEM images of glass substrate prepared at 100 $^\circ{\textrm{C}}$ for 12 h. Some soluble and insoluble network will occur during etching process, as shown in Fig. 4(b). Once phase separation occurs, the insoluble structure shown in the Fig. 4(c) will clear away, showing a uniform structure in Fig. 4(d), where those structures will continue to leaching to form the final structure shown in Fig. 3(d). The insoluble network cannot peel off from the glass when the treating time is not length enough. As the time increasing, the density of the insoluble network will increase, which will block the coming light. This may be the reason of the decrease in the transmittance of the sample that was baked for 12 hours (compared to the sample baked for 4 h or 8 h). We also have done EDX test, the result is similar to Ref. [25], as shown in Fig. 4(e) and Fig. 4(f). It can be clearly seen that after hydrothermal treatment at 100 $^\circ{\textrm{C}}\;\ $for 12 h, the amount of Na is sharply decreased, while the Ca, Al, and Mg is slight increased. As a result, the soluble component is some particle that contain element Na, while the insoluble component is those particle that contain element Ca, Al, and Mg.

 

Fig. 4. SEM images of glass substrate prepared at 100 $^\circ{\textrm{C}}$ for 12 h (a). (b) The enlarged images of selected areas of (a). Panels (c) and (d) show selected areas of (b) registered at a higher magnification. (c) Part1 in (b), (d) Part 2 in (b). EDX spectra of the glass substrate prepared at 100 $^\circ{\textrm{C}}$ for 12 h (e) and without treatment (f)

Download Full Size | PDF

By changing the baking temperature and time, we can also get the FG sample. From the Fig. 5(a) we can see that by using the integrating sphere, the prepared FG (baked at 160$\;\ ^\circ{\textrm{C}}$ for 4 h, 8 h, 12 h) has a better transmittance compared to the NG. Without using the integrating sphere, we will get a converse result, as shown in Fig. 5(b). It means that most of the light can pass through FG, but the propagating direction of the light has changed, just as shown in Fig. 1(b). In addition, as the treated time increasing, the percentage of the scattered light also increased, as shown in Fig. 5(a). By comparing the photograph of the NG and FG (160$\;\ ^\circ{\textrm{C}}$, 12 h) under the incandescent lamp, it can be clearly seen that there is less reflection on the FG (as shown in Fig. 5(c) and Fig. 5(e)). However, it is difficult to see the letters and the logos on the paper through the treated FG shown in Fig. 5(d). This is a property of the FG, the light can pass through the glass, but the image cannot.

 

Fig. 5. The optical properties of the FG. Transmittance of the soda lime glass before and after treated at 160 $^\circ{\textrm{C}}$ (a) without using the integrating sphere, (b) using the integrating sphere. (c) A photograph of the NG and prepared FG (Baked at 160 $^\circ{\textrm{C}}$ for 12 h) under the incandescent lamp. (d) A comparison between the NG and treated FG. Through the prepared FG, we cannot see the logos and letters. The black rectangle in (c) and (d) represent the NG. The red rectangle in (c) and (d) represent the FG. (e) The reflectance of the soda lime glass before and after treated at 160 $^\circ{\textrm{C}}$ measured with the integrating sphere.

Download Full Size | PDF

Figure 6 shows the SEM images of the of glass substrate prepared at 160 $^\circ{\textrm{C}}$ for different treating time. Some micro-craters can be observed on the glass surface, as shown in Fig. 6(a). As a result, when the light reaches the surfaces, the micro-craters will act as a concave mirror or a plane-concave lens scattering the light. As the treated time increasing, the density of the micro-crater increased, which means more light will be scattered, as shown in Fig. 6(c). But when the treating time reaches to 24 h, micro-craters would connect to each other, resulting less forward scattering light. So the transmittance of the surfaces treated for 24 h decreased, as shown in Fig. 5(b). Compared to the sample treated for 4 h, the scattered light through the sample treated for 8 h also increased (Fig. 5(a)). However, it is speculated that due to most of the scattered light is in the backward direction, the transmittance is reduced (Fig. 5(b)).

 

Fig. 6. The surface SEM images of FG prepared at 160 $^\circ{\textrm{C}}$ for different treating time. (a) 4 h; (b) 8 h; (b) 12 h; (b) 24 h.

Download Full Size | PDF

We also have done the experiment with increasing temperature from 100 $^\circ{\textrm{C}}$ to 160 $^\circ{\textrm{C}}$ with 10 $^\circ{\textrm{C}}$ increments. After treatment, the surfaces structures were observed by SEM. The result shows that when the temperature is 140$^\circ{\textrm{C}}$, the processed surfaces is covered with some micro-crater and numerous nanoflakes. And then with the temperature increased or decreased, only micro-crater or nanoflakes can be found on the surface.

4. Conclusion

In conclusion, by using hydrothermal process, we process two different types of structures on soda lime glass, achieving CG and FG. The CG surface is covered with non-uniform nanoflakes, which can improve the transmittance regardless of considering the scatter. The transmittance of the CG improved from 91% to 96.7% at a broadband wavelength. The FG surface is covered with irregular micro-craters, which will acts as a concave or plano-concave lens scattering the light. The transmittance of the FG can reach more than 95%. The fabrication of FG or CG is highly dependent on the temperature and time during hydrothermal process. The facility and economy of this method to fabricate the CG and FG glass are expected to find extensive applications in many fields.