Facile Solution Process of VO2 Film with Mesh Morphology for Enhanced Thermochromic Performance

The fabrication and applications of VO2 film continue to be of considerable interest due to their good thermochromic performance for smart windows. However, low visible transmittance (Tlum) and solar modulation efficiency (∆Tsol) impede the application of VO2 film, and they are difficult to improve simultaneously. Here, a facile zinc solution process was employed to control the surface structure of dense VO2 film and the processed VO2 film showed enhanced visible transmittance and solar modulation efficiency, which were increased by 7.5% and 9.5%, respectively, compared with unprocessed VO2 film. This process facilitated the growth of layered basic zinc acetate (LBZA) nanosheets to form mesh morphology on the surface of VO2 film, where LBZA nanosheets enhance the visible transmittance as an anti-reflection film. The mesh morphology also strengthened the solar modulation efficiency with small caves between nanosheets by multiplying the times of reflection. By increasing the zinc concentration from 0.05 mol/L to 0.20 mol/L, there were more LBZA nanosheets on the surface of the VO2 film, leading to an increase in the solar/near-infrared modulation efficiency. Therefore, this work revealed the relationship between the solution process, surface structure, and optical properties, and thus can provide a new method to prepare VO2 composite film with desirable performance for applications in smart windows.


Introduction
Energy saving in buildings has attracted much attention since there is a strong drive to minimize the use of non-renewable fuels, specifically fossil fuels [1]. Among various ways to save energy, smart windows based on thermochromic material have been widely explored, which can control the transmittance of sunlight in response to ambient temperature, intelligently, to keep the indoor temperature suitable for human beings, thus greatly reducing building energy consumption [2][3][4][5][6][7]. VO 2 is one of the best thermochromic materials and is promising for the reversible transition between rutile structure and monoclinic structure [8][9][10][11], accompanied by an abrupt change in transmittance of near-infrared (NIR) light at 68 • C, which is somewhat higher than room temperature [12]. Ideally, VO 2 film allows NIR light to enter to make the room warm in cold winter while it also keeps NIR light from going through in hot summer to sustain the cool indoor temperature [13]. However, low solar modulation efficiency (∆T sol ), poor luminous transmittance (T lum ), and the high phase transition temperature (T c ) at 68 • C obstruct the application of VO 2 films in smart windows [14].
To overcome these issues, many methods, including doping, composites with other materials, changing film structures, and surface engineering were employed [15][16][17][18]. Surface engineering is supposed to be an efficient method to promote the optical properties of VO 2 film. Liu et al. prepared a highly ordered honeycomb-like structure, which enhanced T lum significantly [15], and similarly, Ning et al. prepared various nanostructures through the sol-gel method in pure VO 2 , which increased T lum from 20% to 36% [16]. Gao et al. fabricated nanoporous VO 2 (M) films to improve T lum [17], and Xie et al. introduced a similar strategy of introducing a periodic porous structure into VO 2 (M) film to block infrared rather than visible light, which showed excellent T lum and ∆T sol [18]. In this sense, the surface structure has an important effect on the thermochromic properties of VO 2 films, indicating that surface engineering is a possible way to enhance T lum and ∆T sol of VO 2 films.
To control the surface structure of VO 2 film, various methods were employed. The addition of protective layers is always used. These layers include TiO 2 [19,20], V 2 O 5 [21], SiO 2 [22], SiN x [23], HfO 2 [24], ZnO [25], and so on. There is some strain between these layers and the VO 2 film, which will have an important effect on the band structure of VO 2 and thus control the phase transition temperature [26][27][28][29][30][31][32][33]. However, the additional protective layer always leads to the deteriorated optical performance of VO 2 film with reduced solar modulation efficiency. Recently, ascorbic acid was used to process the VO 2 surface to enhance its antioxidation and anti-acid properties [34]. An acid solution was utilized to etch pristine VO 2 film prepared with magnetron sputtering to produce a karst landforms-like structure so that the solar modulation efficiency and visible transmittance were improved simultaneously [35]. Unfortunately, the karst landforms-like structure was not controlled easily because VO 2 reacted with acid rather easily. Thus, it is of great significance to develop a facile method for the surface engineering of VO 2 -based film.
Inspired by these above works, and different from acid etching, the zinc solution process was employed to control the surface structure of VO 2 films in this work. Here, zinc acetate was used to easily produce an LBZA compound on the surface of VO 2 film at a low-temperature solution and likely had important effects on thermochromic performance because many Zn compounds such as ZnO or ZnS can help VO 2 film enhance its thermochromic properties [25,36]. At first, VO 2 films were synthesized by the magnetron sputtering method and then processed in the zinc solution. The obtained film exhibited a mesh morphology on the surface, and enhanced T lum and ∆T sol compared with unprocessed films. The effects of the zinc concentration on surface structures and thermochromic properties of the obtained films were investigated in detail. In addition, the relationship between surface structures and the thermochromic performance was revealed to explain the enhancement in thermochromic properties of the films.

VO 2 Film Preparation with Magnetron Sputtering
VO 2 films were deposited by magnetron sputtering on quartz glass substrates (2 cm × 2 cm). A V (99.95%) target was used for deposition. We evacuated the chamber to 3.0 × 10 −3 Pa, then introduced Ar (99.99%) into the chamber and fixed the gas flow at 200 sccm without O 2 in the atmosphere. The duration of the sputtering process was 15 min with a sputtering power of 75 W, and the temperature of the substrate was 298 K. This kind of VO 2 film usually shows a compact structure and uniform particle distribution on the substrate with low visible transmittance. VO 2 films obtained by magnetron sputtering were processed in a Zn 2+ solution to manufacture multilayer composite films.

Post-Synthetic Solution Processing of VO 2 Films
First, 1.3172 g (0.06 mol) of zinc acetate dihydrate (Zn (CH 3 COO) 2 ·H 2 O) was dissolved in 60 mL of methanol (CH 3 OH) in a conical flask (100 mL) under vigorous stirring at room temperature to form a zinc solution, then subjected to ultrasonic radiation for 10 min before adding 15 mL of distilled water to the above solution. The solution was transformed from a transparent state to a white suspension during the process. The VO 2 film with the substrate was dipped into the solution and the flask was covered with plastic film and processed in a water bath at 60 • C for 6 h. Finally, the VO 2 film was taken out and washed with distilled water before heating on a hot plate at 60 • C.

Characterization
X-ray diffraction (XRD, D8Advance, CuKα, λ = 0.154178 nm under an output power of 3 kW) was employed to determine the phase structure in the film. The surface morphology, particle size, and composition of the films were examined with a scanning electron microscope (SEM, JSM-5610LV, JEOL, Tokyo, Japan) equipped with energy dispersed X-ray spectra (EDX). The surface structure was also examined with a scanning probe microscope (AFM, Nanoscope, IV/Nanoscope IV, VEECO, New York, NY, USA). The optical properties of the films (300-2500 nm) were measured using an ultraviolet-visible-near infrared spectrophotometer (UV-Vis-NIR, UV-3600, Shimadzu, Kyoto, Japan) at 20 • C and 90 • C, respectively. The integrated T lum (380-780 nm) and ∆T sol (380-2500 nm) were obtained from the following equation: Here, T(λ) is the transmittance at wavelength λ of the film, ϕ lum (λ) is the standard luminous efficiency function for the photopic vision of human eyes, and ϕ sol (λ) denotes the AM 1.5 solar irradiance spectrum. Moreover, ∆T sol/NIR is obtained from the following equation:

Results and Discussion
3.1. Structure and Thermochromic Performance of Solution-Processed VO 2 -Based Film VO 2 film manufactured by magnetron sputtering was processed in a 0.10 mol/L zinc solution. Figure 1 presents the XRD patterns of processed and unprocessed samples. Three major diffraction peaks in the XRD patterns match VO 2 (M) (JPCDS No. 65-2358) well, suggesting no phase change in the sample. However, XRD failed to detect any phase related to zinc-containing compounds, possibly due to the poor crystallinity or small contents. In order to characterize the surface structure of the VO2 film after the solution process, SEM and EDX were employed. Figure 2 shows the typical morphology of VO2 film before (a) and after (b) solution processing and the element mapping images of Zn (c), V (d), and O (e) on the film. In Figure 2b, the obtained film exhibits a mesh In order to characterize the surface structure of the VO 2 film after the solution process, SEM and EDX were employed. Figure 2 shows the typical morphology of VO 2 film before (a) and after (b) solution processing and the element mapping images of Zn (c), V (d), and O (e) on the film. In Figure 2b, the obtained film exhibits a mesh morphology, and the uniform nanoplates are distributed on the film surface, indicating that the nanoplates have grown on the film surface. Moreover, the three elements including Zn, V, and O are distributed uniformly on the film as shown in Figure 2c-e. The mesh morphology is similar to that of LBZA in the previous reports [37]. The related reaction for the formation of the LBZA nanostructure on the VO 2 surface can be expressed as follows:  In order to characterize the surface structure of the VO2 film after the solution process, SEM and EDX were employed. Figure 2 shows the typical morphology of VO2 film before (a) and after (b) solution processing and the element mapping images of Zn (c), V (d), and O (e) on the film. In Figure 2b, the obtained film exhibits a mesh morphology, and the uniform nanoplates are distributed on the film surface, indicating that the nanoplates have grown on the film surface. Moreover, the three elements including Zn, V, and O are distributed uniformly on the film as shown in Figure 2c-e. The mesh morphology is similar to that of LBZA in the previous reports [37]. The related reaction for the formation of the LBZA nanostructure on the VO2 surface can be expressed as follows:  Figure 3 shows the roughness of the VO2 film before (a) and after (b) the solution process, which exhibits a great difference. Figure 3a shows the morphology of pure VO2 film prepared by magnetron sputtering and the film surface is nearly plane. In Figure 3b, the lower section of the film is supposed to be VO2 while the higher section has the typical morphology of nanoplates. The structures in the figure seem to be nanorods, but, in fact, they are nanoplates because the unit of height is of a nanometer scale while the width and length are of micrometer scale. As a result, nanoplates have grown on the VO2 film, which matches the above SEM results well. In addition, the solution-processed VO2 film exhibits a smaller height (6.3 nm) than that before the solution process (32 nm), likely because some VO2 grains in the film are dissolved in the solution [38].  Figure 3 shows the roughness of the VO 2 film before (a) and after (b) the solution process, which exhibits a great difference. Figure 3a shows the morphology of pure VO 2 film prepared by magnetron sputtering and the film surface is nearly plane. In Figure 3b, the lower section of the film is supposed to be VO 2 while the higher section has the typical morphology of nanoplates. The structures in the figure seem to be nanorods, but, in fact, they are nanoplates because the unit of height is of a nanometer scale while the width and length are of micrometer scale. As a result, nanoplates have grown on the VO 2 film, which matches the above SEM results well. In addition, the solution-processed VO 2 film exhibits a smaller height (6.3 nm) than that before the solution process (32 nm), likely because some VO 2 grains in the film are dissolved in the solution [38].  Figure 4 shows the XPS spectra of the sample after the solution process. It can be seen that Zn, O, and C elements are present in the sample (Figure 4a). In Figure 4b, the XPS spectrum of Zn 2p3/2 can be divided into three peaks at 1021.9 eV, 1022.6 eV, and 1025.3 eV, respectively, indicating that the Zn element in the obtained sample exhibits three states in different chemical environments. The peak at 1021.9 eV matches crystalline ZnO well, whose binding energy value of Zn 2p3/2 is 1021.9 ± 0.1 eV [37], suggesting that the sample includes some ZnO. The peak at 1025.3 eV corresponds to Zn 2+ in Zn(CH3COO)2·2H2O [39], indicating that Zn(CH3COO)2·2H2O likely exists in the sample. The peak at 1022.6 eV is assigned to the acetate zinc group [40], suggesting that there is LBZA with an acetate zinc group in the sample. It can be seen that the sample mainly includes LBZA, as well as a few ZnO and Zn(CH3COO)2·2H2O on its surface. This suggests the nanoplates on the film surface are mainly LBZA. In Figure 4c, the C 1s region includes three peaks detected to be 288.0 eV, 285.7 eV, and 284.9 eV, which correspond to O-C=O [41], C-O [42], and -CH3 [43]. In Figure 4d, the O 1s region can be also divided into three peaks at 534.0 eV, 532.0 eV, and 531.5 eV. The peak at 532.0 eV is ascribed to Zn-O-C=O [37] in LBZA while the 531.5 eV peak is ascribed to ZnO [44]. The 534.0 eV peak is related to O in the H2O of Zn(CH3COO)2·2H2O [45]. However, XPS failed to detect a V signal because the VO2 film is covered by LBZA and the detection depth of XPS is limited.  Figure 4 shows the XPS spectra of the sample after the solution process. It can be seen that Zn, O, and C elements are present in the sample (Figure 4a). In Figure 4b, the XPS spectrum of Zn 2p 3/2 can be divided into three peaks at 1021.9 eV, 1022.6 eV, and 1025.3 eV, respectively, indicating that the Zn element in the obtained sample exhibits three states in different chemical environments. The peak at 1021.9 eV matches crystalline ZnO well, whose binding energy value of Zn 2p 3/2 is 1021.9 ± 0.1 eV [37], suggesting that the sample includes some ZnO. The peak at 1025.3 eV corresponds to Zn 2+ in Zn(CH 3 COO) 2 ·2H 2 O [39], indicating that Zn(CH 3 COO) 2 ·2H 2 O likely exists in the sample. The peak at 1022.6 eV is assigned to the acetate zinc group [40], suggesting that there is LBZA with an acetate zinc group in the sample. It can be seen that the sample mainly includes LBZA, as well as a few ZnO and Zn(CH 3 COO) 2 ·2H 2 O on its surface. This suggests the nanoplates on the film surface are mainly LBZA. In Figure 4c, the C 1s region includes three peaks detected to be 288.0 eV, 285.7 eV, and 284.9 eV, which correspond to O-C=O [41], C-O [42], and -CH 3 [43]. In Figure 4d, the O 1s region can be also divided into three peaks at 534.0 eV, 532.0 eV, and 531.5 eV. The peak at 532.0 eV is ascribed to Zn-O-C=O [37] in LBZA while the 531.5 eV peak is ascribed to ZnO [44]. The 534.0 eV peak is related to O in the H 2 O of Zn(CH 3 COO) 2 ·2H 2 O [45]. However, XPS failed to detect a V signal because the VO 2 film is covered by LBZA and the detection depth of XPS is limited. Figure 5 exhibits the solar transmittance spectra of VO 2 film before and after the solution process. At 20 • C, the transmittance of processed VO 2 film shows notable promotion, especially in the short-wavelength area. After calculation, visible transmittance of the processed film at 20 • C is 40.1%, while unprocessed VO 2 film is 37.3%. In addition, solar modulation efficiency is also promoted from 9.5% to 10.4%. This is possibly ascribed to the LBZA nanoplates on the surface of VO 2 films. To investigate the enhancement in the thermochromic performance of solution-processed VO 2 -based film, the zinc concentration will be investigated as an important factor later.  Figure 5 exhibits the solar transmittance spectra of VO2 film before and after the solution process. At 20 °C, the transmittance of processed VO2 film shows notable promotion, especially in the short-wavelength area. After calculation, visible transmittance of the processed film at 20 °C is 40.1%, while unprocessed VO2 film is 37.3%. In addition, solar modulation efficiency is also promoted from 9.5% to 10.4%. This is possibly ascribed to the LBZA nanoplates on the surface of VO2 films. To investigate the enhancement in the thermochromic performance of solution-processed VO2-based film, the zinc concentration will be investigated as an important factor later.    Figure 5 exhibits the solar transmittance spectra of VO2 film before an lution process. At 20 °C, the transmittance of processed VO2 film shows no tion, especially in the short-wavelength area. After calculation, visible tra the processed film at 20 °C is 40.1%, while unprocessed VO2 film is 37.3% solar modulation efficiency is also promoted from 9.5% to 10.4%. This is pos to the LBZA nanoplates on the surface of VO2 films. To investigate the en the thermochromic performance of solution-processed VO2-based film, the tration will be investigated as an important factor later.

The Effect of Zinc Concentration on Structure and Thermochromic Performance
To explore the effects of zinc concentrations on the thermochromic properties of VO 2 film, samples were processed in four different zinc concentrations (0.05 mol/L, 0.10 mol/L, 0.15 mol/L, and 0.20 mol/L) for 6 h, respectively. Figure 6 presents the XRD patterns of the obtained films. The green, blue, and red lines denote the films obtained at zinc concentrations of 0.15 mol/L, 0.10 mol/L, and 0.05 mol/L, respectively. All the diffraction peaks show great agreement with VO 2 (M) (JPCDS No. 65-2358), and no other peaks indexed to other compounds, indicating that obtained films are M-phase VO 2 and the solution process does not change the phase structure of VO 2 film obtained by magnetron sputtering. By increasing zinc concentration, the peak intensity decreased and the noncrystalline peak becomes weaker, which is attributed to the film thickness.
Materials 2022, 15, x FOR PEER REVIEW

The Effect of Zinc Concentration on Structure and Thermochromic Performance
To explore the effects of zinc concentrations on the thermochromic properties film, samples were processed in four different zinc concentrations (0.05 mol/L, 0.10 0.15 mol/L, and 0.20 mol/L) for 6 h, respectively. Figure 6 presents the XRD pattern obtained films. The green, blue, and red lines denote the films obtained at zinc con tions of 0.15 mol/L, 0.10 mol/L, and 0.05 mol/L, respectively. All the diffraction peak great agreement with VO2(M) (JPCDS No. 65-2358), and no other peaks indexed t compounds, indicating that obtained films are M-phase VO2 and the solution proce not change the phase structure of VO2 film obtained by magnetron sputtering. By i ing zinc concentration, the peak intensity decreased and the noncrystalline peak b weaker, which is attributed to the film thickness. The morphology of the obtained films was investigated further by SEM an characterizations and the results are schematically illustrated in Figures 7-9. Fi shows the morphology of the film by zinc solution process at 0.05 mol/L. The nan are thought to be layered basic zinc acetates, and the little grains below are the morphology of VO2. The structure corresponds well with Figure 9a. Although Fi appears to display nanorods, they are, in fact, nanoplates because the unit of heig nanometer scale while the width and length are of micrometer scale. At 0.05 mo nanoplates exhibit a scattered distribution on the surface. In Figure 7b-d, the nan increase and finally cover the surface to form a mesh-like surface structure with creasing zinc concentration from 0.10 to 0.20 mol/L. This is in good agreement w AFM results in Figure 3 that the roughness of film increases with the increase in t concentration. The film thickness was also characterized by SEM and the result is in Figure 8. It can be seen that the thickness of the VO2 film before the solution was roughly 103.4 nm as shown in Figure 8a. After the solution process, the film th was also 103.4 nm. This indicates that the film thickness was not decreased. The morphology of the obtained films was investigated further by SEM and AFM characterizations and the results are schematically illustrated in Figures 7-9. Figure 7a shows the morphology of the film by zinc solution process at 0.05 mol/L. The nanoplates are thought to be layered basic zinc acetates, and the little grains below are the typical morphology of VO 2 . The structure corresponds well with Figure 9a. Although Figure 9a appears to display nanorods, they are, in fact, nanoplates because the unit of height is of nanometer scale while the width and length are of micrometer scale. At 0.05 mol/L, the nanoplates exhibit a scattered distribution on the surface. In Figure 7b-d, the nanoplates increase and finally cover the surface to form a mesh-like surface structure with an increasing zinc concentration from 0.10 to 0.20 mol/L. This is in good agreement with the AFM results in Figure 3 that the roughness of film increases with the increase in the zinc concentration. The film thickness was also characterized by SEM and the result is shown in Figure 8. It can be seen that the thickness of the VO 2 film before the solution process was roughly 103.4 nm as shown in Figure 8a. After the solution process, the film thickness was also 103.4 nm. This indicates that the film thickness was not decreased.       Based on the above results, the structure change of the film surface is clearly shown in Figure 10. The orange layer represents the VO2 film while the blue layer below represents the glass substrate. The violet plates above represent the nanoplates of LBZA, which multiply with the increasing zinc concentration. To characterize the thermochromic properties of the obtained samples after the solution process, their solar transmittance (300-2500 nm) was measured with a UV-vis-NIR spectrophotometer at 20 °C as well as 90 °C, and the transmittance is shown in Figure 11. These four curves are almost the same, indicating that zinc concentration has no obvious Based on the above results, the structure change of the film surface is clearly shown in Figure 10. The orange layer represents the VO 2 film while the blue layer below represents the glass substrate. The violet plates above represent the nanoplates of LBZA, which multiply with the increasing zinc concentration. Based on the above results, the structure change of the film surface is clearly shown in Figure 10. The orange layer represents the VO2 film while the blue layer below represents the glass substrate. The violet plates above represent the nanoplates of LBZA, which multiply with the increasing zinc concentration. To characterize the thermochromic properties of the obtained samples after the solution process, their solar transmittance (300-2500 nm) was measured with a UV-vis-NIR spectrophotometer at 20 °C as well as 90 °C, and the transmittance is shown in Figure 11. These four curves are almost the same, indicating that zinc concentration has no obvious To characterize the thermochromic properties of the obtained samples after the solution process, their solar transmittance (300-2500 nm) was measured with a UV-vis-NIR spectrophotometer at 20 • C as well as 90 • C, and the transmittance is shown in Figure 11. These four curves are almost the same, indicating that zinc concentration has no obvious effect on the solar transmittance of the obtained films. From the data in Figure 11, the visible transmittance, solar modulation efficiency, and near-infrared modulation efficiency were calculated and are exhibited in Table 1. With the increase in the zinc concentration in the solution, the visible transmittance increased at first and then decreased. When the zinc concentration is 0.10 mol/L, the visible transmittance of the VO 2 film reached the maximum value (40.1%). As the zinc concentration increased, the solar modulation efficiency and infrared modulation efficiency also increased. When the zinc concentration was 0.20 mol/L, the solar/NIR modulation efficiency reached the maximum. However, the difference decreases when the zinc concentration is larger than 0.15 mol/L, so this can be ignored. It is obvious that, at lower zinc concentrations, the transmittance from 300 nm to 2500 nm increased at both 20 • C and 90 • C, which causes higher visible transmittance. When the zinc concentration in the solution increases, the transmittance from 300 nm to 2500 nm at 20 • C remains high while the transmittance from 300 nm to 2500 nm at 90 • C decreases, leading to an enhanced solar modulation efficiency. Specifically, when the zinc concentration is 0.15 mol/L, the film shows a visible transmittance of 38.2% at 20 • C, a solar modulation efficiency of 11.3%, and an infrared modulation efficiency of 20.3%, which is better than that of most VO 2 films in previous works [15][16][17]46] as shown in Table 2.
effect on the solar transmittance of the obtained films. From the data in Figure 11, the visible transmittance, solar modulation efficiency, and near-infrared modulation efficiency were calculated and are exhibited in Table 1. With the increase in the zinc concentration in the solution, the visible transmittance increased at first and then decreased. When the zinc concentration is 0.10 mol/L, the visible transmittance of the VO2 film reached the maximum value (40.1%). As the zinc concentration increased, the solar modulation efficiency and infrared modulation efficiency also increased. When the zinc concentration was 0.20 mol/L, the solar/NIR modulation efficiency reached the maximum. However, the difference decreases when the zinc concentration is larger than 0.15mol/L, so this can be ignored. It is obvious that, at lower zinc concentrations, the transmittance from 300 nm to 2500 nm increased at both 20 °C and 90 °C, which causes higher visible transmittance. When the zinc concentration in the solution increases, the transmittance from 300 nm to 2500 nm at 20 °C remains high while the transmittance from 300 nm to 2500 nm at 90 °C decreases, leading to an enhanced solar modulation efficiency. Specifically, when the zinc concentration is 0.15 mol/L, the film shows a visible transmittance of 38.2% at 20 °C, a solar modulation efficiency of 11.3%, and an infrared modulation efficiency of 20.3%, which is better than that of most VO2 films in previous works [15][16][17]46] as shown in Table 2.    The enhancement in visible transmittance and solar modulation is related to the nanostructure of LBZA. There are many small cavities on the surface of LBZA nanoplates in Figure 12. Light reflected by VO 2 is more likely to be reflected by LBZA again. After the second reflection, some light is reflected onto VO 2 again, and the third reflection on VO 2 will also occur in the same way. Compared with pure VO 2 film, the VO 2 /LBZA film can reflect the light more times, and the reflected light had a second chance to go through the film. In this sense, more light can transmit through the composite film, leading to an increased visible transmittance, which is in good agreement with the results in Figure 5. With the occurrences of light reflection increased, more near-infrared light is absorbed by the VO 2 film, resulting in an increase in the solar modulation efficiency. It seems that some light is trapped in the small cavities and continues reflection until escaping, leading to an enhanced solar modulation efficiency. Here, we assume the reflectivity of LBZA is R, then the light reflected by LBZA back to VO 2 is R * (1 − T), in which ∆T = R * (1 − T) * T transmits the film with a second chance. Since R is constant, the additional transmittance depends on T (transmittance of VO 2 ). Obviously, the additional transmittance obtains the maximum with T at 50%. As shown in Figure 5, the transmittance at 20 • C is always closer to 50% than that at 90 • C, so the increase is always larger at 20 • C. Since ∆T sol = T sol (20 • C) − T sol (90 • C) and T sol (20 • C) increases more, ∆T sol also increases at the same time. As a result, the enhancement in both T lum and ∆T sol can be attributed to the nanostructure of LBZA on the surface of the VO 2 film. In addition, the thermochromic performances of the obtained VO 2 composite film were similar after 100 thermal cycles, leading to good stability. This is likely because the LBZA/VO 2 composite film is very stable at a temperature below 100 • C.  The enhancement in visible transmittance and solar modulation is related to the nanostructure of LBZA. There are many small cavities on the surface of LBZA nanoplates in Figure 12. Light reflected by VO2 is more likely to be reflected by LBZA again. After the second reflection, some light is reflected onto VO2 again, and the third reflection on VO2 will also occur in the same way. Compared with pure VO2 film, the VO2/LBZA film can reflect the light more times, and the reflected light had a second chance to go through the film. In this sense, more light can transmit through the composite film, leading to an increased visible transmittance, which is in good agreement with the results in Figure 5. With the occurrences of light reflection increased, more near-infrared light is absorbed by the VO2 film, resulting in an increase in the solar modulation efficiency. It seems that some light is trapped in the small cavities and continues reflection until escaping, leading to an enhanced solar modulation efficiency. Here, we assume the reflectivity of LBZA is R, then the light reflected by LBZA back to VO2 is R * (1 − T), in which ΔT = R * (1 − T) * T transmits the film with a second chance. Since R is constant, the additional transmittance depends on T (transmittance of VO2). Obviously, the additional transmittance obtains the maximum with T at 50%. As shown in Figure 5, the transmittance at 20 °C is always closer to 50% than that at 90 °C, so the increase is always larger at 20 °C. Since ∆Tsol = Tsol(20 °C) − Tsol(90 °C) and Tsol(20 °C) increases more, ∆Tsol also increases at the same time. As a result, the enhancement in both Tlum and ∆Tsol can be attributed to the nanostructure of LBZA on the surface of the VO2 film. In addition, the thermochromic performances of the obtained VO2 composite film were similar after 100 thermal cycles, leading to good stability. This is likely because the LBZA/VO2 composite film is very stable at a temperature below 100 °C.

Conclusions
In this work, the zinc solution process was used to prepare LBZA nanoplates on the surface of VO2 film. The obtained VO2 film exhibited enhanced thermochromic performance compared with unprocessed VO2 film: The visible transmit-

Conclusions
In this work, the zinc solution process was used to prepare LBZA nanoplates on the surface of VO 2 film. The obtained VO 2 film exhibited enhanced thermochromic performance compared with unprocessed VO 2 film: The visible transmittance was increased from 37.3% up to 40.1%, solar modulation efficiency was promoted from 9.5% to 10.4%, and the infrared modulation efficiency was improved from 17.4% to 18.8%. The enhancement was likely attributed to the formed LBZA nanoplates, which can increase the occurrences of solar light reflection. In addition, by increasing the zinc concentration, the visible transmittance was increased at first and then decreased, while the solar modulation efficiency and infrared modulation efficiency were increased. Therefore, zinc solution processing is a facile method to control the surface structure of VO 2 film and thus has important effects on the thermochromic performance, which can shed light on the preparation of VO 2 films with high performance for smart windows.