Design, Fabrication, and Characterization of Thermal and Optical Properties of Nano-Composite Self-Cleaning Smart Window

Polymer-based smart windows have recently received attention due to their capabilities in energy consumption reduction. A smart window provides desired optical properties when heated/cooled by using solar energy when the ambient temperature requires regulation. The main issue here is the design and fabrication of such a smart element which is the main axis of the current research. The window in the proposed design operates in such a way that the percentage of light transmission depends on the presence of nanofluid between the two walls and refractive index conformity between the fluid and the polymeric walls; Therefore the percentage of light transmission will be at its minimum value (45%) in the absence of fluid and it will be at the maximum value (80%) at the presence of fluid. The fundamental steps of the present design includes design, fabrication, and characterization of the materials. In this regard experiments to determine the mechanical, physical, structural, optical, and thermal properties of components have been performed after considering, designing, and manufacturing various samples. The results show that the proposed smart offers acceptable performance with a fast switching rate and even more than other similar smart glasses due to the usage of discharge/injection mechanism. In overall, the product can be used as a smart transparent element in various structures such as buildings and even vehicles to regulate energy consumption and/or block the view for security purposes.


Introduction
Buildings now account for 40% of energy consumption and consequently 36% of carbon dioxide emissions according to previous research (Allen, Connelly, Rutherford, & Wu, 2017). This indicates the major importance of the study of buildings in terms of environment and energy consumption.
Currently, standard windows typically waste a third of all energy used for heating and ventilation.
Thus efforts have been focused solely on windows, and the use of appropriate technology to reduce this energy loss (Donaldson, 2018). A switchable glass (transparent/opaque) is a logical solution to transmit/reflex solar light based on requirements. These devices are designed to reduce air conditioning costs via sunlight blocking in summer and improve light harvesting during winter. Their fundamental mechanism is based on the tunable optical transmittance of incorporated switchable devices, generally stimulated as a response of an applied energy or changes in environmental conditions (Sala, Gonçalves, Camargo, & Leite, 2018). The main issue of such devices is to utilize newer technologies and solutions to improve thermal and optical properties of them to compromise fabrication and production costs. Polymer-based smart windows are one of the reasonable elements that can be used in buildings to have the required features in this regard. The economical design and manufacturing capabilities, flexibility, optical/thermal features, and nano-technological compatibilities of polymeric products, are the main reasons of the focus of this study on polymerbased smart windows. These types of smart-windows are of the most controllable products in terms of design, fabrication, and application. Polymer-based smart windows can be comfortably embedded in buildings and vehicles to improve energy consumption, automation, and even safety of them in terms of blocking the sight while required.
Previous studies about smart windows design and application are found to be in different categories ranging from energy consumption management purposes to advanced fabrication methods from which the more important cases are described here in a classified manner.

Energy consumption optimization
States. It was found that compared to the new standard low-transmisson windows, a low-temperature switching thermochromic window consumes less energy by 10 to 17 percent in the south and west.
These results have been verified only in the specific geographical regions which have been examined, however Warwick et al. (Warwick, Ridley, & Binions, 2014) have analyzed this type of windows under a broader range of conditions in and have proven a 50 percent reduction in energy consumption compared to standard windows.
Thermochromic performance of copolymerized micro-hydrogel particles has also been investigated by adjusting the particle size and the structure (Li, Liu, Feng, & Fang, 2019) to increase the energy consumption reduction in these type of windows however subtle improvement is obtained and fabrication process of such window is complicated. Regardless of production cost, thermochromic smart windows must be investigated deeper due to their potential application for building automation purposes (Piccolo & Simone, 2015). Allen et al. (Allen et al., 2017) gained more control over the absorption of solar energy by considering the variable nature of smart windows, between clear and opaque modes, compared to the fixed behavior of the low emission (LE) window. For example, hydroxypropyl cellulose smart window transmits 74% of the sun's radiation in the transparent state and only 11% in the opaque state. This value for the LE window is a fixed value of 53%. Therefore, during the cold periods of the year, the low percentage of passing through the LE window loses useful solar energy and causes the additional energy required for heating to increase. In contrast, smart windows reduce the amount of energy required by high solar energy (74%) during the cold season. Although this advantage seems reasonable and practical due to inherent properties of the materials used in smart windows, the exact extent of its superiority over LE windows must be tested in a broader manner.

Electrochromic smart windows
Electrochromic switchable devices (SW) are usually composed of multi-layers of components in which switching process is obtained by oxidation/reduction processes and diffusion of externally activated ions (Baetens, Jelle, & Gustavsen, 2010). Then some features as rapid switching by external voltage (C.-C. Wu, Liou, & Diao, 2015) and reflection control can be achieved for these devices.
Although this type of device can be valuable, it requires several steps for materials development and layers construction, which might consequently lead to reduced transmittance and limited modulation level (Chen, Cao, Chen, Luo, & Gao, 2014). Additionally, since they change their optical properties by switching between oxidized and reduced form (Baetens et al., 2010), these electrochemical reactions can have a reduced performance overtime due to unwanted side reactions between the electrolyte and active materials (J. , such as decomposition of electrolytes and efficiency reduction of charge transfer reaction between the electrolyte and active materials (Hu et al., 2016).
1.3. Other smart windows (Photochromic, gasochromic, etc.) The efforts on manufacturing photochromic and gasochromic windows are almost considerable as photochromism has already been utilized in fabrication of smart glasses for years and gasochromism has appeared to be a more reasonable approch than electrochromism (Sala et al., 2018). In this regard, Wu et al. (L. Y. L. Wu, Zhao, Huang, & Lim, 2017), have used transparent photochromic films for smart window applications by embedding organic photochromic dyes in a sol-gel-based matrix. The degree of connection of the coating matrix and the types of organic groups have had effects on the transmission of visible light and the rate of bleaching. By changing the amount of color concentration in the coating, as well as the thickness of the coating, the amount of light transmission reduction has been adjusted between 30 to 60%. According to the author, this product will significantly reduce the passage of energy in the tropics. It should be noted that the high time of state change, lack of high percentage of transparency, and complexity of the structure are among the disadvantages of this design.
In another study conducted by Feng et al. (Feng et al., 2016), intelligent gasochromic windows were tested for energy efficiency in terms of optical and thermal properties. This was done in more depth by simulating energy consumption in a building. This study showed that the best areas to use this type of window are areas with cold winters and hot summers. The use of these areas reduces energy consumption in terms of ventilation, cooling, and heating. Nevertheless, the limitation in geographical location and relatively complex equipment required to produce gas should not be overlooked in this plan.

Simulations
Numerical simulation of the performance of smart windows have not been very common among the researchers yet, however there have been some attempts to perform this task. A numerical study on the performance of a multilayer smart window embedded in water-cooled third-generation solar cells has been conducted by Sabri et al. (Sabry, Eames, Singh, & Wu, 2014). System optimization parameters such as light concentration ratio and cooling flow (of water) are required to prevent system performance loss due to thermal stresses and high cell temperatures. In this study, detailed modeling of thermal properties of window system was performed using fluid analysis software; Finally, by considering the conductive, convective, and radiative heat transfer mechanisms in the proposed smart window, numerical solution results are presented.
In another study, the effect of passive ventilation system and smart windows has been studied by Khalesi et al. (Javad & Navid, 2019) in a building adaptable to climatic conditions. The distribution of temperature and air conditions have been investigated for two heat sources, smart windows ,and two types of ventilation systems in fluid analysis software. Thermal comfort criteria have also been set by the validated model for smart windows. The results of the analysis show that smart windows are superior in terms of fulfillment of comfort criteria.
It was also shown that the temperature difference between floor and ceiling can be reduced by up to 50% with the help of electrochromic windows. Here an innovative design of smart windows is provided by harnessing light refraction index conformity and taking the advantage of simple draining/injection system. Thereby, the optical transmittance of the smart windows can be changed between extreme values at any time without the need to reach a switching threshold in response to intense solar heating. The window in the proposed design operates in such a way that the percentage of light passing through the window depends on the presence or absence of fluid (nanofluid) inside the polymeric panel together with the refractive index conformity between these components according to the percentage of nanoparticles in the fluid. Therefore the percentage of light transmission will be at its minimum value (45%) in the absence of fluid and it will be at the maximum value (80%) at the presence of fluid. Characterizing experiments, fabrications steps, and obtained results are explained and discussed in the following sections in detail.

Materials
Transparent polymers usually exhibit various optical properties such as tailored emission/absorption properties and/or low/high refractive index. These facts about such polymers attract great interest because of the potential optoelectronic applications (Demir et al., 2007). Hence the polymer used is PMMA. Another reason for choosing this polymer is its availability. PMMA as a transparent polymeric material offers excellent transparency in the visible and near-infrared range (Zettl, Mayer, Klampaftis, & Richards, 2017). It is often used as an alternative to glass. PMMA has a transmission of more light and much lighter than silica glass. In addition, it is almost easy to find a fluid with a similar light refractive index (n), which can be modified with the desired coating. Nevertheless, PMMA transmits UV light; thus, manufacturers usually apply UV coatings on PMMA to overcome this deficiency (Hammani, Barhoum, & Bechelany, 2018).
The nanocomposite coating for this purpose is ZnO-PMMA.ZnO NP is a well-known multifunctional inorganic filler that has outstanding properties such as high refractive index, high thermal conductivity, self-cleaning behavior as well as photo-catalytic, antibacterial, and UV-protection properties (Sun, Miyatake, & Sue, 2007).
The target fluid is methyl salicylate, which turns to nanofluid after the addition of zinc oxide nanoparticles. Among the important properties of the desired fluid can be mentioned the following: • The refractive index close to the PMMA; • Medium and low viscosity, which is suitable for the state change mechanism, • High transparency, • Large boiling point and ignition which makes it stable in the desired temperature range, • Environmental adaptation (skin contact).

Fabrication
The experimental steps including fabrication process, and characterization tests (Fig. 1)  After analyzing different geometries, rotated cube appeared to be a better choice. In this pattern, the rotated cubes are considered as holes inside the plate. Fig. 3 shows the details of this design. Using this design will cause the light to refract successively, when the fluid is not inside the window.
Consequently, the percentage of transmission is significantly reduced. In contrast, the presence of fluid increases the transparency considerably due to refraction index conformity between fluid and polymer.

Nano-composite film
The desired values of the percentage of nanoparticles are listed in Table 1. It should be noted that according to research, the maximum achievable concentration of nanoparticles in the polymer matrix is about 10 wt%, which is due to the high surface energy and low mixability of nanoparticles.
Therefore, values lower than 10 wt% are considered here for the production of nanocomposites. ), was obtained as a relatively thick white liquid. Since zinc oxide powder was purchased in the form of nanoparticles, there was no need for initial synthesis to prepare the oxide powder. Therefore, the powder was used directly. Then different amounts of zinc oxide nanoparticle powder were added to the solution. In the next step, the sample was mixed for 60 min at a temperature range of 60 °C to 65 °C, and a semiwhite product was obtained. At next step the gel state product was placed on a glass plate to prepare the film. Finally the produced film was placed at 80 °C for 11 min to remove the solvent from the sample. The sample thickness was measured in different spots and the desired thickness was considered to be 30 μm.

Nano-fluid
To prepare the desired nanofluid, zinc oxide nanoparticles were added and mixed with concentrations of 0.1, 0.5, and 1 wt% according to the mass of the desired fluid ( Table 2). The mixture was then placed on a magnetic stirrer before performing the ultrasonic process to ensure re-dispersion of nanoparticles in the base fluid. In order to simplify the production and assembly process of sample, the dimensions of the sample are considered to be currently 0.1 × 0.1 m 2 . The final assembly was obtained by attaching the panels and installing holes required for fluid injection/drain (Fig. 7). Contact angle testing was performed for polymer sheets and nanocomposite films versus water, fluid, and nano-fluid with two replicates. The technical characteristics of this device are presented in Table   3, and the polymeric sample is shown in Fig. 8. Tensile test was performed for polymer panels with two replicates on the samples (Fig. 9). The Tinus-Olsen ST-100 was used to perform this test.   Table 4 has been used. Heat conductivity coefficient of polymer and nanocomposite film has been measured to recognize the heat transfer rate of external and internal surfaces of the product. SDK TCC 001 (Fig. 11), with the technical specifications in Table 5 has been used to perform this test.   The nanocomposite and fluid structural characterizations were each performed separately to determine the existing structure. This was done for nanoparticles by X-ray diffraction based on ASTM D5357 by Bruker D8-Advance with the technical specifications in Table 7  (1)

= ⋅ cos
Where is the crystal dimension, is the dimensionless shape factor with a value close to the unit value, is the beam wavelength, is the radiation angle, and is the width of the half of the maximum value at the peak of the scatter.  Table 8 with two replications.
Thus the morphology and chemical composition of nanocomposite films were examined with a 10 kV SEM with low suction.  Table 9 with two replications.
Nano particle size analysis was performed in the desired solution with a laser beam with a wavelength of 632 nm.  Fig. 12. The average value obtained for different parameters is presented in Table 10.

Flexural test
The results of the flexure test performed on the polymer plates is mentioned in Fig. 13. The average value obtained for different parameters is presented in Table 11. According to the results obtained in Fig. 13

Extension (mm)
The results of contact angle test between fluid and solid (polymer) are presented in Table 12. An example of the image presented in this test is depicted in Fig. 14.   Fig. 14 Photograph of a contact angle test between polymeric sample and water. Nano-composte film -Water

76
Nano-composte film -Fluid 79 Nano-composte film -Nano-fluid 81 According to the results obtained in Table 12, the results related to polymeric panels and water has been normal according to other studies (Ma, Cao, Feng, Ma, & Zou, 2007). The value has increased In the case of polymer and fluid. This is even more for panels and nanofluid which indicates higher fluid phobicity of the PMMA and nanofluids.
As the results show the contact angle for nanocomposite film and fluid has increased in all three cases indicating that the film is more fluid-phobic. The maximum value is obtained for nanocomposite film and nanofluid. Although fluid phobicity alone does not reflect the property of self-cleaning, the point here is that this property will be less effective if the surface is not exposed to environmental pollution.
Because when the film is exposed to contamination, there is a possibility of non-fluid particles adhering to it, which will require surface washing due to its subtle fluid-philicity property. Therefore, the fluid-phobicity of the inner surfaces of the window will cause self-cleaning inside the window.  Fig. 16. The result shows the common granularity of zinc oxide nanoparticles in nanocomposite. According to the results obtained in Fig. 15, zinc oxide particles size have been acceptable and it is within the standard range for nanocomposites and nanofluids. In addition, as seen in Fig. 16, the surface obtained from the nanocomposite film is in appropriate condition according to the results of similar studies (Hammani et al., 2018).

DLS test
The results of DLS test on nanofluid in the range of 0 − 100 nm shows the acceptable dispersion of nanoparticles in the fluid. The results show that the particle dimensions have been between 3.5 nm and 12 nm.

DSC test
The DSC test result is shown in Fig. 17 on pure polymer. The rest of PMMA results with different zinc oxide percentages are shown in Table 13. As data show in nanocomposite films are higher than pure polymer, this value has increased with increasing percentage of ZnO nanoparticles. of pure polymer is about 90.5 ° C.

97.8
According to the results obtained in Fig. 17 and Table 13 the range of vitrification and melting of the polymer is suitable for its manufacture. An important point to be seen in these results is that has increased with increasing the percentage of ZnO, which could be due to the lack of complete solubility of the polymer in the solvent and ruduction in molecular weight distribution. This is due to the direct relationship of value to the mobility of the polymer chain (Kim et al., 2012). Although does not necessarily affect the performance of the window, knowing its thermal characteristics can ensure a safe temperature range for its working conditions.

Thermal conductivity test
The results of the thermal conductivity coefficient test are listed in Table 14.

0.220
Nano-composite film (NCS 3 ) * 0.291 * Refer to Table 1 The thermal conductivity coefficient has increased slightly with increasing the percentage of zinc oxide according to the results obtained in Table 14. Although this parameter does not have considerable influence on the desired performance of the window, it will be considered in terms of heat loss, or the input of unwanted heat from outside. In general, it can be said that the panel of the window play a more insulating role than conductive.
3.5. Optical test Fig. 18 and Table 15 show the results of the UV-Vis test for polymers, nanocomposites, and nanofluids in visible and ultraviolet range in the room temperature. The results include plain and patterned polymeric panels, as well as windows containing fluid with different percentages of ZnO nanoparticles are shown in Table 15. According to the results, the variation in light transmittance for the opaque and transparent state is between 45% and 80%. The opaque state occurs when the windows does not include the fluid and the transparent state occurs when complete set is used. It should be noted that one of the features of smart glass is the time required to switch between the states. According to previous studies, this period is relatively time consuming for thermochromic, electrochromic, and gas chromic glasses due to the chemical processes required to change their state. This is also the case for other smart glasses. Glasses containing hydrogels, for example, take about 40 minutes to change state due to rising and falling temperatures (Gyenes, Szilágyi, Lohonyai, & Zrínyi, 2003). The more advanced version of this window (containing microgels) has just reduced this time to 4 minutes (M. . In glasses containing hydroxypropyl methyl cellulose, the time required to change state during the temperature reduction process is about 6 minutes (Kiruthika & Kulkarni, 2017).
Although some efforts are seen in previous studies to reduce the switching time, must of them operate during a period longer than 1 or 2 minutes, while this period has reduced to a duration as short as 1 minute due to harnessing injection/drain mechanism in the proposed window. As mentioned previously, the mechanism operates in such a way that the presence or absence of fluid (nanofluid) inside the window and the matching of the refractive index between the fluid (nanofluid) and the panels, respectively, make the window transparent or opaque. However, it should be noted that the area of the window and the inlet-outlet flow of the fluid will also affect this period of time.

Window performance
After making the sample, temperature variations during the daylong for a testing chamber were compared in constant climatic conditions. The qualitative values of this analysis are shown in Table   16. In overall, In comparison to other switchable windows, the versatile and efficient design proposed here has provided temperature independent system which is stimulated by small input power, requiring less energy consumption and just a drain/injection process to provide a reversible optical process.

Conclusion
The main benefits of using smart windows are reducing energy consumption and improving the building automation. The use of this technology is more necessary in summer, when energy consumption peaks for cooling systems as well as winter when these glasses play an effective role in reducing energy consumption by passing a high percentage of sunlight into the building since the sunlight is less intense due to the angles of the sunlight with the ground.
Although the technology of most smart glasses has had a positive impact on energy management, the significant cost of making these glasses is a major obstacle. Current smart windows which are switched by changes in the intensity of sunlight, are not only cost-effective to produce, but also do not work well in the long run. In addition, toxic substances are usually used in the process of making these types of windows, which can pose significant risks to nature and human health.
The proposed smart window offers a cheaper and more cost-effective idea for making these types of windows. In this window, the mentioned problems have been considerably resolved to an acceptable extent by using nontoxic materials and simple switching mechanism while having a stable performance in the long run. In addition, the time period for switching between the states is relatively short, while its fabrication cost is reasonable compared to similar types. The major advantages of this product can be summarized as follows: • The efficiency of the proposed window compared to the other available types such as electrochromic and low emission windows has been considerable in terms of energy consumption as expected. Light transmission in both states (opaque and transparent) for the proposed window causes to reduce heating and cooling costs compared to other smart windows. In fact, a higher percentage of sunlight passing in cold weather and a lower percentage of sunlight passing in hot weather is a clear proof for this.
• The selection of PMMA and proposed manufacturing method has reduced the complexity and the cost of fabricating a smart window as predicted. Although the cost used materials are acceptable, manufacturing costs can be reduced even more by using a turbid fluid. It can be adapted to the simple switching mechanism in reverse mode (Opaque when the fluid is injected, and transparent when the fluid is drained), making the design more simple and inexpensive. This way even the issue of refractive index compliance is no more a considered parameter, while the only issue will be self-cleaning. Therefore, the use of nanocomposite film is still inevitable. This mechanism require more research in the next studies.
• The proposed smart window has had a higher percentage of transparency than the existing products as predicted due to the matching properties of the polymer and nanofluid used here.
It has been concluded that the light transmittance range for both opaque and transparent conditions is comparable to other smart windows according to the results obtained in Table   15.
• The research method has shown that the switching time is reduced compared to the existing smart windows. According to the other studies, the minimum time required to switch between states is about 4 minutes, while in the proposed window, this time is reduced to less than 1 minute, due to the simple injection/drain mechanism however the area of the window and the flow rate are two important controlling factors.

Ethics declarations
The research was funded by Iran National Science Foundation (INSF).