Facile Synthesis of Island-like ZrO2-VO2 Composite Films with Enhanced Thermochromic Performance for Smart Windows

VO2-based film, as a very promising thermochromic material for smart windows, has attracted extensive attention but has not been widely applied because it is difficult to simultaneously improve in terms of both solar-modulation efficiency (ΔTsol) and visible transmittance (Tlum) when made using the magnetron-sputtering method, and it has poor durability when made using the wet chemical method. Herein, island-like ZrO2-VO2 composite films with improved thermochromic performance (ΔTsol: 12.6%, Tlum: 45.0%) were created using a simple approach combining a dual magnetron-sputtering and acid-solution procedure. The film’s ΔTsol and Tlum values were increased initially and subsequently declined as the sputtering power of the ZrO2 target was raised from 30 W to 120 W. ΔTsol achieved its maximum of 12.6% at 60 W, and Tlum reached its maximum of 51.1% at 90 W. This is likely the result of the interaction of two opposing effects: Some VO2 nanocrystals in the composite film were isolated by a few ZrO2 grains, and some pores could utilize their surface-plasmon-resonance effect at high temperature to absorb some near-infrared light for an enhanced ΔTsol and Tlum. More ZrO2 grains means fewer VO2 grains in the composite film and increased film thickness, which also results in a decrease in ΔTsol and Tlum. As a result, this work may offer a facile strategy to prepare VO2-based films with high thermochromic performance and promote their application in smart windows.


Introduction
Thermochromic windows are considered a promising way to reduce building energy consumption significantly due to their simple structure, good solar modulation, and zero-energy input characteristics [1]. Among these thermochromic films, VO 2 films have attracted much attention because they can modulate near-infrared (NIR) transmittance via their reversible and ultra-fast transition between monoclinic phase (M phase) and rutile phase (R phase) at about 68 • C [1]. In this sense, VO 2 thermochromic smart windows can respond to changes in the surrounding temperature and then automatically adjust the amount of solar radiation entering indoors. The smart windows can block NIR from entering the room when the temperature is high, thus reducing energy consumption caused by air conditioning, and they can allow NIR to enter indoors when the temperature is low in winter, raising the indoor temperature and thus reducing energy consumption caused by warming. Therefore, VO 2 films have emerged as a promising material for the upcoming generation of smart-window coatings due to the above advantages. In fact, VO 2 films are not widely applied in buildings, mainly because of undesirable modulation efficiency

Materials and Methods
Without additional purification, all chemicals were utilized after being acquired from Sinopharm Chemical Reagent Co., Ltd. In this experiment, Nanchang National Materials Technology Co., Ltd.'s V target (99.95% purity) and ZrO 2 target (99.99% purity) were employed. The magnetron-sputtering equipment used in the experiment was JPD-500. The size of the vacuum chamber was Φ500 × H420 mm and the size of rotating substrate table was Φ150 mm. Figure 1a displays magnetron-sputtering diagram. Before sputtering, cooling water was turned on. It was ensured that the vacuum chamber was evacuated to 3.0 × 10 −3 Pa. At this time, argon with a purity of 99.99% was introduced into the reaction chamber and the flow of argon was adjusted to 200 sccm. First, ZrO 2 -V composite films were deposited through direct-current magnetron sputtering of V targets at a power of 90 W and radio-frequency magnetron sputtering of ZrO 2 targets simultaneously at various power. The total continuous sputtering duration of the V target was 15 min, and the zirconia target was intermittently sputtered 9 times for 30 s each time. The sputtering pressure was 0.4 Pa. Figure 1b depicts the comprehensive manufacturing procedure of ZrO 2 -V composite films. Finally, ZrO 2 -VO 2 composite films were obtained through post-annealing the ZrO 2 -V composite films in a tube furnace. Specifically, the films were specifically placed in a tube furnace with an air pressure of 1000 Pa. the temperature was ramped up to 450 • C and V composite films. Finally, ZrO2-VO2 composite films were obtained through post-annealing the ZrO2-V composite films in a tube furnace. Specifically, the films were specifically placed in a tube furnace with an air pressure of 1000 Pa. the temperature was ramped up to 450 °C and held for 1 h at a rate of 5 °C/min, followed by a natural cooling process. For comparison, ZrO2-VO2 composite films with different sputtering power of ZrO2 were denoted as sample ZrO2-30 W, ZrO2-60 W, ZrO2-90 W, and ZrO2-120 W. ZrO2-30 W, ZrO2-60 W, ZrO2-90 W, and ZrO2-120 W were placed in a PTFE etching flower basket and each was firstly subjected to a 5 s treatment in hydrochloric acid at a molar concentration of 5.6 mol/L, noting that there was no obvious change in optical properties after the acid-solution process. The acid-solution-process time was therefore increased to 20 s. Immediately after corrosion, the films were taken out of the acid solution and ultrasonically cleaned for one minute in deionized water and anhydrous ethanol. The obtained VO2-based films were dried with N2 and labelled as ZrO2-30 W-acid, ZrO2-60 Wacid, ZrO2-90 W-acid, and ZrO2-120 W-acid. The preparation conditions of pure VO2 film and ZrO2-doped VO2 film are shown in Table 1.  ZrO 2 -30 W, ZrO 2 -60 W, ZrO 2 -90 W, and ZrO 2 -120 W were placed in a PTFE etching flower basket and each was firstly subjected to a 5 s treatment in hydrochloric acid at a molar concentration of 5.6 mol/L, noting that there was no obvious change in optical properties after the acid-solution process. The acid-solution-process time was therefore increased to 20 s. Immediately after corrosion, the films were taken out of the acid solution and ultrasonically cleaned for one minute in deionized water and anhydrous ethanol. The obtained VO 2 -based films were dried with N 2 and labelled as ZrO 2 -30 W-acid, ZrO 2 -60 W-acid, ZrO 2 -90 W-acid, and ZrO 2 -120 W-acid. The preparation conditions of pure VO 2 film and ZrO 2 -doped VO 2 film are shown in Table 1. The crystal structure of the film was characterized on an Empyrean diffractometer using grazing angle X-ray diffraction (GAXRD) measurement (Cu Kα, λ = 0.154178 nm, generated at 4 kW output power). The morphology and element distribution of the film were characterized by a field-emission scanning-electron microscope (SEM, JSM-5610LV, Tokyo, Japan). The valence and composition of elements of the film were determined by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi/ESCALAB 250Xi. Thermo Fisher, Waltham, MA, USA). To evaluate the optical properties of composite films, the solar transmittance of the film in the range of 300-2500 nm at 20 • C and 90 • C was analyzed by an ultraviolet-visible near-infrared spectrophotometer (UV-3600, Shimadzu Corporation, Kyoto, Japan). The integrals T lum and ∆T sol were calculated with the following Formulas (1) and (2) [1][2][3][19][20][21][22][23][24][25].
The ∆T sol was calculated with Formula (3).
where T(λ) is the transmittance, λ is the wavelength of the incident light, ν m (λ) denotes the spectral sensitivity of the light to the human eye, and ϕ sol (λ) represents the irradiance spectrum of the sunlight at an atmospheric mass of 1.5 (corresponding to the sun from the horizon 37 • ) [1-3,26-29].

Structures of the Films before and after Acid-Solution Treatment
The XRD patterns of the samples before (ZrO 2 -30 W, ZrO 2 -60 W, ZrO 2 -90 W, ZrO 2 -120 W) and after (ZrO 2 -30 W-acid, ZrO 2 -60 W-acid, ZrO 2 -90 W-acid, ZrO 2 -120 W-acid) the acid-solution procedure are shown in Figure 2. It can be seen that every sample had strong diffraction peaks at 2θ = 27.9 • , 37.1 • , 42.4 • , and 55.6 • , which could be matched to VO 2 (M) (JCPDS No.44-252, a = 5.753Å, b = 4.526Å, c = 5.383Å, space group: P2 1 /c, α = 90 • , β = 122.6 • ). No further noticeable peaks were observed, with the exception of the amorphous peak of the glass substrate at 2θ = 20 • .This indicates that VO 2 in the film only exists in the form of M phase. In the XRD diagram, we did not find an obvious diffraction peak of ZrO 2 , which may have been due to the short sputtering time of ZrO 2 , and the content of ZrO 2 prepared was lower than the XRD measurement limit. According to Goedicke's experimental analysis, ZrO 2 films prepared by magnetron sputtering are crystalline without annealing. There were obvious diffraction peaks at ZrO 2 (111), ZrO 2 (220), ZrO 2 (211), and ZrO 2 (220), and they grew preferentially with the increase in sputtering power. The successful preparation of ZrO 2 was proven in the subsequent XPS test analysis [30][31][32][33]. In order to further investigate the impact of the acid-solution process of the morphology evolution of the samples, SEM characterization was performed on samples ZrO2-60 W, ZrO2-60 W-acid, ZrO2-90 W, and ZrO2-90 W-acid, and the results are shown in Figure  3. It can be seen that the film's grain size was comparatively homogeneous and became small after the acid-solution process. As the power increased, the film became rougher and gradually thicker before the acid-solution process. After the acid-solution process, the intensity of the peak at 2θ = 37.1 • was progressively declined as sputtering power increased, showing the benefit of a proper ZrO 2 concentration for the preferred growth of VO 2 grains along the (020) crystal plane. Moreover, the film's diffraction-peak intensity was decreased when the sputtering power reached 120 W. This is probably because the excess ZrO 2 affected the crystallization of VO 2 . The intensity of the peak at 2θ = 37.1 • was gradually decreased as the ZrO 2 power increased after the acid-solution process, whereas the intensity of the peak at 2θ = 27.9 • gradually increased, as shown in Figure 2b. This indicates that hydrochloric acid preferentially corrodes (020) facets of VO 2 grains during the acid process [32][33][34][35].
In order to further investigate the impact of the acid-solution process of the morphology evolution of the samples, SEM characterization was performed on samples ZrO 2 -60 W, ZrO 2 -60 W-acid, ZrO 2 -90 W, and ZrO 2 -90 W-acid, and the results are shown in Figure 3. It can be seen that the film's grain size was comparatively homogeneous and became small after the acid-solution process. As the power increased, the film became rougher and gradually thicker before the acid-solution process. After the acid-solution process, voids gradually developed in the films to form numerous islands on the surface that resulted in a strong surface-plasmon-resonance (LSPR) effect and a higher ∆T sol , and the thickness of the films gradually decreased, which was extremely encouraging for achieving an increase in the films' T lum [36][37][38]. In order to further investigate the impact of the acid-solution process of the morphology evolution of the samples, SEM characterization was performed on samples ZrO2-60 W, ZrO2-60 W-acid, ZrO2-90 W, and ZrO2-90 W-acid, and the results are shown in Figure  3. It can be seen that the film's grain size was comparatively homogeneous and became small after the acid-solution process. As the power increased, the film became rougher and gradually thicker before the acid-solution process. After the acid-solution process, voids gradually developed in the films to form numerous islands on the surface that resulted in a strong surface-plasmon-resonance (LSPR) effect and a higher ΔTsol, and the thickness of the films gradually decreased, which was extremely encouraging for achieving an increase in the films' Tlum [36][37][38].  Figure 4a,b exhibit the solar transmittances of samples ZrO2-30 W, ZrO2-60 W, ZrO2-90 W, ZrO2-120 W, and samples ZrO2-30 W-acid, ZrO2-60 W-acid, ZrO2-90 W-acid, and ZrO2-120 W-acid, respectively, at 20 °C and 90 °C. Table 2 lists the films' solar-modulation efficiency (ΔTsol) and luminous transmittance (Tlum). When the sputtering power of the V target was 90 W, the pure VO2 film obtained by magnetron sputtering showed a ΔTsol of 12.4% and Tlum of 28.2%. The VO2-based films' Tlum increased dramatically when ZrO2 was introduced, reaching 46.9% when the ZrO2 target's sputtering power was 30 W (ZrO2-30 W). This is because the introduction of ZrO2 decreased the refractive index of VO2-based films. Furthermore, the introduction of ZrO2 improved the crystallinity of the film, resulting in an increase in ΔTsol to 13.3% at 30 W (ZrO2-30 W-acid) and 14.3% at 60 W. (ZrO2-60 W-acid). However, when the power exceeded 60 W (ZrO2-90 W and ZrO2-120 W), too much ZrO2 affected the oxidation and crystallization process of the V film, resulting in a   Table 2 lists the films' solar-modulation efficiency (∆T sol ) and luminous transmittance (T lum ). When the sputtering power of the V target was 90 W, the pure VO 2 film obtained by magnetron sputtering showed a ∆T sol of 12.4% and T lum of 28.2%. The VO 2 -based films' T lum increased dramatically when ZrO 2 was introduced, reaching 46.9% when the ZrO 2 target's sputtering power was 30 W (ZrO 2 -30 W). This is because the introduction of ZrO 2 decreased the refractive index of VO 2 -based films. Furthermore, the introduction of ZrO 2 improved the crystallinity of the film, resulting in an increase in ∆T sol to 13.3% at 30 W (ZrO 2 -30 W-acid) and 14.3% at 60 W. (ZrO 2 -60 W-acid). However, when the power exceeded 60 W (ZrO 2 -90 W and ZrO 2 -120 W), too much ZrO 2 affected the oxidation and crystallization process of the V film, resulting in a decrease in T lum and ∆T sol . After a 20 s treatment in hydrochloric acid at a molar concentration of 5.6 mol/L, the obtained films showed very good acid resistance. From Figure 5 and Table 2, the T lum was increased after the acid-solution process, probably attributable to the reduction in film thickness and the generation of few pores in the film. The reduction of VO 2 content in the film probably led to a decrease in ∆T sol . Consequently, it is unrealistic to enhance T lum by extending acid-solution-processing time, which is in good agreement with previous works [32]. In particular, the ZrO 2 -VO 2 obtained after the acid-solution process (ZrO 2 -90 W-acid) exhibited the highest T lum of 51.1% while keeping a good ∆T sol of 9.4%. decrease in Tlum and ΔTsol. After a 20 s treatment in hydrochloric acid at a molar concentration of 5.6 mol/L, the obtained films showed very good acid resistance. From Figure 5 and Table 2, the Tlum was increased after the acid-solution process, probably attributable to the reduction in film thickness and the generation of few pores in the film. The reduction of VO2 content in the film probably led to a decrease in ΔTsol. Consequently, it is unrealistic to enhance Tlum by extending acid-solution-processing time, which is in good agreement with previous works [32]. In particular, the ZrO2-VO2 obtained after the acidsolution process (ZrO2-90 W-acid) exhibited the highest Tlum of 51.1% while keeping a good ΔTsol of 9.4%.     decrease in Tlum and ΔTsol. After a 20 s treatment in hydrochloric acid at a molar concentration of 5.6 mol/L, the obtained films showed very good acid resistance. From Figure 5 and Table 2, the Tlum was increased after the acid-solution process, probably attributable to the reduction in film thickness and the generation of few pores in the film. The reduction of VO2 content in the film probably led to a decrease in ΔTsol. Consequently, it is unrealistic to enhance Tlum by extending acid-solution-processing time, which is in good agreement with previous works [32]. In particular, the ZrO2-VO2 obtained after the acidsolution process (ZrO2-90 W-acid) exhibited the highest Tlum of 51.1% while keeping a good ΔTsol of 9.4%.    In order to investigate the effect of the acid-solution process on the composition of island-like ZrO 2 -VO 2 composite film, XPS characterization was performed on samples ZrO 2 -90 W and ZrO 2 -90 W-acid, and the results are shown in Figure 6. It is evident that both films contained C, V, Zr, and O elements. The binding energies for O 1s, V 2p, Zr 3d, and C 1s, where the C contents came from adventitious carbon, were 530 eV, 515 eV, 180 eV, and 284.8 eV, respectively.

Thermochromic Properties of VO 2 -Based Films
In order to investigate the effect of the acid-solution process on the composition of island-like ZrO2-VO2 composite film, XPS characterization was performed on samples ZrO2-90 W and ZrO2-90 W-acid, and the results are shown in Figure 6. It is evident that both films contained C, V, Zr, and O elements. The binding energies for O 1s, V 2p, Zr 3d, and C 1s, where the C contents came from adventitious carbon, were 530 eV, 515 eV, 180 eV, and 284.8 eV, respectively. As seen in Figure 7a,b, the peak of Zr 3d could be fitted into Zr 3d5/2 and Zr 3d3/2, located at 182.1 eV and 184.5 eV, respectively, corresponding to Zr 4+ . After the acid-solution process, the peaks of Zr 3d5/2 and Zr 3d3/2 were located at 182.2 eV and 184.6 eV, respectively, which also corresponded to Zr 4+ in ZrO2, and the intensity of the peak was almost unchanged, indicating that the Zr content was nearly unchanged after the acidsolution process. This is probably due to the excellent chemical stability ZrO2 displayed in acidic solutions. After the acid-solution process, only a small amount of Zr element was probably reacted. This is confirmed by the XPS result in Figure 6. The results shows that the molar ratio of Zr/V increased from 0.02 to 0.05 during the acid-solution process, indicating that V was lost from the film at a faster rate than Zr. The 20 s of the acid-solution process did not entirely remove Zr [39]. If extending the processing time, the solarmodulation rate of the films would decrease significantly. The characteristic peak of V in the sample before the acid-solution process had two peaks corresponding to V 2p1/2 and V 2p3/2, as shown in Figure 7c. For V 2p3/2, there were two binding energy peaks located at 516.2 eV and 517.5 eV that could be assigned to +4 and +5 valences of V, respectively [5,40,41]. The primary source of V's +5 valence was V2O5, indicating that the film surface was partly oxidized. O 1s spectra could be fitted into two peaks. The peak at ~530 eV was attributed to the V-O bond in the crystal lattice, whereas the peak at ~532 eV belonged to the surface-chemisorbed oxygen species [30,31]. After the acid-solution process, the film's peak position was barely altered, whereas the content of the chemisorbed oxygen species was decreased, indicating that the island-like ZrO2-VO2 composite film exhibited good oxygen resistance, different from previous works [38]. The resistance was recorded at gradient temperatures ranging from 20 °C to 90 °C in order to determine the phase-transition temperature (Tc) of samples ZrO2-90 W and ZrO2-90 W-acid. Figure 8 depicts the electrical hysteresis loop of the film. After the introduction of ZrO2 to the film, it can be seen that the Tc of the film increased to 52 °C, which may be attributed to the interfacial stresses caused by the ZrO2 and VO2 grains. Additionally, the electrical hysteresis loop's breadth was about 8 °C. After the acid-solution process, some channels or voids were generated in the film and the stress between the crystal grains was released; hence, the Tc of the film rose to 58 °C and the width of the electrical hysteresis loop was expanded to 12 °C, which As seen in Figure 7a,b, the peak of Zr 3d could be fitted into Zr 3d 5/2 and Zr 3d 3/2 , located at 182.1 eV and 184.5 eV, respectively, corresponding to Zr 4+ . After the acidsolution process, the peaks of Zr 3d 5/2 and Zr 3d 3/2 were located at 182.2 eV and 184.6 eV, respectively, which also corresponded to Zr 4+ in ZrO 2 , and the intensity of the peak was almost unchanged, indicating that the Zr content was nearly unchanged after the acidsolution process. This is probably due to the excellent chemical stability ZrO 2 displayed in acidic solutions. After the acid-solution process, only a small amount of Zr element was probably reacted. This is confirmed by the XPS result in Figure 6. The results shows that the molar ratio of Zr/V increased from 0.02 to 0.05 during the acid-solution process, indicating that V was lost from the film at a faster rate than Zr. The 20 s of the acid-solution process did not entirely remove Zr [39]. If extending the processing time, the solar-modulation rate of the films would decrease significantly. The characteristic peak of V in the sample before the acid-solution process had two peaks corresponding to V 2p 1/2 and V 2p 3/2 , as shown in Figure 7c. For V 2p 3/2 , there were two binding energy peaks located at 516.2 eV and 517.5 eV that could be assigned to +4 and +5 valences of V, respectively [5,40,41]. The primary source of V's +5 valence was V 2 O 5 , indicating that the film surface was partly oxidized. O 1s spectra could be fitted into two peaks. The peak at~530 eV was attributed to the V-O bond in the crystal lattice, whereas the peak at~532 eV belonged to the surface-chemisorbed oxygen species [30,31]. After the acid-solution process, the film's peak position was barely altered, whereas the content of the chemisorbed oxygen species was decreased, indicating that the island-like ZrO 2 -VO 2 composite film exhibited good oxygen resistance, different from previous works [38]. The resistance was recorded at gradient temperatures ranging from 20 • C to 90 • C in order to determine the phase-transition temperature (T c ) of samples ZrO 2 -90 W and ZrO 2 -90 W-acid. Figure 8 depicts the electrical hysteresis loop of the film. After the introduction of ZrO 2 to the film, it can be seen that the T c of the film increased to 52 • C, which may be attributed to the interfacial stresses caused by the ZrO 2 and VO 2 grains. Additionally, the electrical hysteresis loop's breadth was about 8 • C. After the acid-solution process, some channels or voids were generated in the film and the stress between the crystal grains was released; hence, the T c of the film rose to 58 • C and the width of the electrical hysteresis loop was expanded to 12 • C, which can be mainly attributed to the hysteresis effect in the thermal transmission process caused by the channels or voids between the grains [37][38][39]. can be mainly attributed to the hysteresis effect in the thermal transmission process caused by the channels or voids between the grains [37][38][39]. Above all, the island-like ZrO2-VO2 composite films (ZrO2-90 W-acid), especially those prepared by magnetron sputtering, exhibited improved visible transmittance and superior solar modulation compared with pure VO2 films. As demonstrated in Table 3, the thermochromic performance attained was superior to that in the majority of earlier works.  can be mainly attributed to the hysteresis effect in the thermal transmission process caused by the channels or voids between the grains [37][38][39]. Above all, the island-like ZrO2-VO2 composite films (ZrO2-90 W-acid), especially those prepared by magnetron sputtering, exhibited improved visible transmittance and superior solar modulation compared with pure VO2 films. As demonstrated in Table 3, the thermochromic performance attained was superior to that in the majority of earlier works.   Above all, the island-like ZrO 2 -VO 2 composite films (ZrO 2 -90 W-acid), especially those prepared by magnetron sputtering, exhibited improved visible transmittance and superior solar modulation compared with pure VO 2 films. As demonstrated in Table 3, the thermochromic performance attained was superior to that in the majority of earlier works.

Conclusions
In this work, ZrO 2 -VO 2 composite films with enhanced thermochromic properties were prepared by a combination of magnetron sputtering and an acid-solution process. After the acid-solution treatment, the grains with poor crystallinity in the film were etched, and some channels or voids were generated at the same time, leading to an increase in the T lum of the film. Without affecting the oxidation of the V film, the ∆T sol of the film was increased with the increment of the ZrO 2 content because its introduction could improve the crystallinity of the film. Excessive ZrO 2 probably affected the crystallization and oxidation process of V in the film, leading to a decrease in the ∆T sol and T lum . When the sputtering power of ZrO 2 was 30 W and 60 W, the composite films prepared by the acid-solution process exhibited better thermochromic properties. Therefore, this work can provide a very facile and effective method to prepare VO 2 -based films with good thermochromic performance for smart windows.