Lab- and Large-Scale Hydrothermal Synthesis of Vanadium Dioxide Thermochromic Powder

Abstract

Vanadium dioxide (VO₂) is a phase-change material of great importance due to its thermochromic properties, which make it a potential candidate for energy-saving applications. In this work, a comparative study between VO₂ thermochromic films prepared from powders synthesized by either a lab-scale hydrothermal autoclave or a large-scale hydrothermal reactor is presented. In both cases, the as-obtained material, after the hydrothermal step, was subsequently annealed at 700 °C under a nitrogen atmosphere, in order to obtain the monoclinic VO₂(M) thermochromic phase. The VO₂ powder prepared in the large-scale hydrothermal reactor exhibited a critical transition temperature of 54 °C with a hysteresis width of 9 °C, while for the one prepared in the lab-scale autoclave, the respective values were 62 °C and 5 °C. Despite these differences, the prepared films showed similar thermochromic performance with the lab-scale material displaying a 17% IR (InfraRed), switching at 2000 nm upon heating, and a transmittance solar modulation of 11%, compared to 17% and 9%, respectively, for the large-scale material. Moreover, both films appeared to have similar luminous transmittance of 44% and 46%, respectively, at room temperature (25 °C). These results showcase the potential for scaling up the hydrothermal synthesis of VO₂, resulting in films with similar thermochromic performance to those from lab-scale fabrication.

1. Introduction

Vanadium dioxide (VO₂) is a well-known thermochromic material that undergoes a reversible semiconductor to metal transition (SMT) at a critical transition temperature of T_C = 68 °C, for the pristine material [1]. This electrical transition is accompanied by structural and optical ones. More specifically, below T_C, VO₂ forms a monoclinic (M) structure, behaves as a semiconductor and is highly IR-transmissive, while above T_C, its structure becomes tetragonal rutile (R), behaves as a metal and is IR-reflective. Despite having been studied in the literature, it is still unclear whether this transition is a Mott–Hubbard one involving electron–electron correlation or a Peierls one due to electron–phonon interaction [2]. Furthermore, the visible transmittance of VO₂ remains almost constant, independent of temperature [3]. Thus, it can be used in a number of applications, including IR-photodetectors [4], optical memories [5], dielectric metasurfaces [6,7], batteries [8], gas sensors [9,10] and thermochromic coatings on “smart” glazing systems, in order to regulate the internal temperature of buildings [11,12,13,14].

Among various techniques that can be employed to grow VO₂, such as sputtering, chemical vapor deposition, sol–gel, pulsed laser deposition, etc. [4,15,16,17,18,19,20], the hydrothermal synthesis has attracted much attention during the last decade due to its simplicity and cost-effectiveness [21,22,23,24,25,26]. In addition, a great advantage of this technique is that VO₂ is synthesized in powder form; therefore, it can be deposited as a film even on flexible substrates (e.g., plastics), which require fabrication at low temperatures (<100 °C). More specifically, Khaled et al. [27] hydrothermally synthesized thermochromic W-doped VO₂ films with a critical transition temperature of 40 °C, a solar modulation of 7% and a luminous transmittance of 44%, while Okada et al. [24] used microwave-assisted hydrothermal synthesis to grow thermochromic VO₂ powders with a critical transition temperature of about 50 °C, which were used to fabricate thermochromic films with a transmittance solar modulation of 20.73% and a luminous transmittance of 55.04%. In addition, Guo et al. [28] prepared flexible thermochromic films based on hydrothermally synthesized VO₂(M) nanoparticles, showing a transmittance solar modulation of 12.34% and a luminous transmittance of 54.26%, while the critical transition temperature was about 49 °C, as Saini et al. [29] synthesized thermochromic VO₂(M) powders with a critical transition temperature of 62 °C in a one-pot hydrothermal synthesis, at 230 °C for 24 h, by varying the molar ratio between vanadium pentoxide and the reducing agent citric acid.

For the wider implementation of this technology, issues regarding large-scale synthesis and deposition of VO₂ thermochromic films should be tackled. Until now, only a few works have reported on the large-scale deposition of thermochromic VO₂ by utilizing sputtering and roll-to-roll techniques. In the work of Xiang et al. [30], the authors proposed, as a large-scale method, the thermal oxidation at 470 °C of vanadium nitride films prepared by sputtering, which resulted in thermochromic VO₂ films with an IR switching of 69.3% at λ = 2500 nm. Moreover, Rezek et al. [31] applied the sputtering technique on a roll-to-roll manufacturing device in order to fabricate W-doped VO₂ thermochromic films on thin flexible (0.1 nm) glass, which was pre-coated with ZrO₂ using the same technique, resulting in a critical transition temperature of 22 °C, a luminous transmittance of 45% and a solar transmittance modulation of 10%. In this current work, a comparison of the thermochromic performance of films prepared from hydrothermally synthesized VO₂ powders using a lab-scale autoclave and a large-scale reactor is presented for the first time. The VO₂ powder that was synthesized hydrothermally in the large-scale reactor, followed by thermal annealing, exhibited a critical transition temperature of 54 °C with a hysteresis of 9 °C, while the respective films prepared had an IR switching of 17%, a luminous transmittance of 44% and a solar transmittance modulation of 9%.

2. Materials and Methods

VO₂ powders were hydrothermally synthesized in a lab-scale (125 mL stainless steel Parr Teflon-lined autoclave) and a large-scale (hydrothermal reactor, Reaction kettle-10, 13.6 L, Weihai New Era Chemical Machinary Co., Ltd., Weihai, China, Figure 1) system, using vanadium pentoxide as the vanadium source and oxalic acid as the reducing agent, followed by thermal annealing at 700 °C under a nitrogen atmosphere for 2 h.

The procedure is described in detail elsewhere [32]. Due to technical limitations with the large-scale reactor, the reaction duration was reduced to 6 h, instead of 12 h used with the lab-scale autoclave. The powder obtained with the lab-scale autoclave was about 0.315 g per batch, while the one prepared in the large-scale reactor was almost 39.4 g per batch, an increase of two orders of magnitude. Each VO₂ powder was mixed with polyvinylpyrrolidone (PVP), ethanol and ethylene diamine tetraacetic acid (EDTA), followed by sonication and ball milling (6 h). Finally, each slurry was spin-coated on a glass substrate, resulting in coatings with good adhesion.

The structure of each VO₂ powder was examined using an X-Ray Diffractometer (XRD) (Bruker AXS D8 Advance) (Karlsruhe, Germany) operating at 40 kV and 40 mA over the 2θ/θ collection range of 10–80° with a scan rate of 0.05°/1 s. From the XRD pattern, the crystallite size was calculated using Scherrer’s Equation (1) below:

D (nm) = (0.9 × λ)/[B × cos(θ_Β)]

(1)

where D is the crystallite size in the direction perpendicular to the lattice planes, λ = 0.154 nm (Cu Kα source), B is the full-width at half-maximum of the X-Ray diffraction peak in radians, and θ_B is the corresponding Bragg angle. Moreover, the distance between the (hkl) crystalline planes was calculated by Bragg’s Law (Equation (2)):

d (nm) = λ/[2 × sin(θ_Β)]

(2)

The morphology was studied with a Scanning Electron Microscope (SEM) (JEOL 7000) (Tokyo, Japan) operating at 15 keV. The thermochromic behavior of the VO₂ powders was verified by Differential Scanning Calorimetry (DSC) (PL-DSC system), at a temperature ranging from 20 °C to 170 °C with a step of 10 °C/min, under a nitrogen flow of 20 cc/min, while connected with a liquid nitrogen cooling system. From these, the critical transition temperature (T_C) and hysteresis width (ΔT_C) were derived using Equations (3) and (4):

T_C = (T₁ + T₂)/2

(3)

ΔT_C = T₁ − T₂

(4)

where T₁ and T₂ are the critical transition temperatures during heating and cooling procedures, respectively, as determined by the DSC measurements.

Finally, the thermochromic performance of the VO₂ films was investigated by optical transmittance measurements upon heating. Using an Ultraviolet/Visible/Near-InfraRed (UV/Vis/NIR) spectrophotometer (PerkinElmer Lambda 950) (Shelton, CO, USA), working at λ = 250–2500 nm and equipped with a home-made heated holder, the transmittance spectra at 25 °C and 90 °C were recorded. From these, the IR switching (ΔTr_IR) at 2000 nm, the transmittance solar modulation (ΔTr_sol), the visible transmittance (Tr_Vis) at 600 nm and the luminous transmittance (Tr_lum) were calculated by the equations presented below:

ΔTr_IR = Tr(25 °C) − Tr(90 °C)

(5)

$${\mathrm{T}\mathrm{r}}_{\mathrm{i}}\left(\mathsf{{\rm T}}\right)={\displaystyle \frac{\int {\mathsf{\phi }}_{\mathrm{i}}\left(\mathsf{\lambda }\right)·\mathrm{T}\mathrm{r}\left(\mathsf{\lambda },\mathrm{T}\right)·\mathrm{d}\mathsf{\lambda }}{\int {\mathsf{\phi }}_{\mathrm{i}}\left(\mathsf{\lambda }\right)\mathrm{d}\mathsf{\lambda }}}$$

(6)

ΔTr_sol (%) = Tr_sol(25 °C) (%) − Tr_sol(90 °C) (%)

(7)

where Tr (25 °C) and Tr (90 °C) are the transmittance (λ = 2000 nm) at 25 °C and 90 °C, respectively; Tr_sol and Tr_lum are the integrated solar and the luminous transmittance, respectively; Tr (λ,Τ) denotes the measured transmittance spectrum at a specific temperature; i denotes solar (sol) or luminous (lum); φ_sol is the solar irradiance spectrum for an air mass of 1.5 (corresponding to the sun standing 37° above the horizon) [33]; and φ_lum is the standard luminous efficiency function for photopic vision [34].

3. Results and Discussion

XRD patterns of the as-obtained and annealed VO₂ powders, prepared by both systems, are presented in Figure 2. Both of the as-obtained powders can be mainly identified as a mix of vanadium oxide phases such as the VO₂(B) (monoclinic, Joint Committee on Powder Diffraction Standard JCPDS Card No. 81-2392) and the VO₂(A) (tetragonal, JCPDS Card No. 42-0876) metastable polymorphs, as well as V₆O₁₃ (monoclinic, JCPDS Card No. 27-1318); however, the characteristic peak of VO₂(M) at 2θ = 27.8°, corresponding to the (011) crystallographic plane (monoclinic, JCPDS Card No. 44-0252), is present in the autoclave-prepared powders. Moreover, after thermal annealing, both powders mainly consist of polycrystalline VO₂(M) with different preferred orientations each. The autoclave-prepared sample has a preferred orientation of (011) (2θ = 27.8°), while the reactor one has a preferred orientation of (20$\overline{2}$) (2θ = 36.8°). The crystallite size was calculated for these crystallographic planes using Scherrer’s formulae and found to be 57 nm and 42 nm, respectively. The difference in the crystallite size and the XRD patterns can be attributed to the different duration of the hydrothermal synthesis. In

Table 1. The structural parameters of the hydrothermally synthesized powders in both the autoclave and reactor before and after annealing.

Sample	2θ (deg)	Phase/Crystalline Planes	dhkl (nm)
Autoclave As-obtained	14.85	VO2(A)/(110)	0.596
	25.35	V6O13/(110)	0.351
	27.80	VO2(M)/(011)	0.320
	29.95	VO2(A)/(220)	0.298
	33.30	VO2(A)/(212)	0.268
	37.05	VO2(B)/(-112)	0.242
	55.51	VO2(M)/(211)	0.165
Reactor As-obtained	23.95	VO2(B)/(201)	0.371
	30.45	V6O13/(400)	0.293
	36.90	VO2(M)/(20-2)	0.243
	42.00	VO2(M)/(21-2)	0.214
	55.16	VO2(M)/(21-3)	0.166
Autoclave Annealed	25.30	V6O13/(110)	0.351
	26.80	V6O13/(003)	0.332
	27.80	VO2(M)/(011)	0.320
Reactor Annealed	27.35	V6O13/(111)	0.325
	28.30	VO2(B)/(-202)	0.315
	36.90	VO2(M)/(20-2)	0.243

, the structural parameters extracted from the XRD patterns (Figure 2) are presented. The distance between the (hkl) crystalline planes was calculated using Equation (2).

The surface morphology of the samples was investigated by scanning electron microscopy, the images of which are presented in Figure 3. The as-obtained materials (Figure 3a,c) have a mainly rod-like morphology; however, in the case of the reactor, product particles (Figure 3c) are highly aggregated, forming flake-like structures. After annealing, the morphology of the powders prepared in the reactor (Figure 3d) remains similar with more well-shaped sintered features, while the one prepared in the autoclave consists of larger discrete rods, flakes and sphere-like particles. Finally, the lack of smoothness and uniformity can be attributed either to the duration of the annealing process or the existence of various vanadium oxide phases, as indicated by the XRD analysis.

The thermochromic behavior of all powdered samples was examined by DSC measurements, as presented in Figure 4. From these, the critical transition temperature and the hysteresis width are calculated. For the as-obtained material prepared in the lab-scale autoclave, the values were calculated as T_C = 56 °C and ΔT_C = 12 °C, while the one prepared in the large-scale reactor showed no thermochromicity, probably due to the shorter time of hydrothermal synthesis, which resulted in the absence of the VO₂(M) phase before annealing. It should be noted that this limitation in the hydrothermal synthesis duration for the large-scale method was imposed for safety reasons due to pressure build-up during operation of the reactor. Moreover, after thermal annealing, the autoclave sample is still thermochromic with T_C = 64 °C and ΔT_C = 5 °C, while the reactor sample displays thermochromic behavior with T_C = 59 °C and ΔT_C = 9 °C. The increase in T_C and the simultaneous decrease in ΔT_C of the material prepared in the autoclave can be attributed to its increased crystallinity resulting from thermal annealing, as confirmed by the XRD pattern (Figure 1) and SEM images (Figure 3b). In both cases, an enhancement of their thermochromic performance is achieved after the thermal annealing step.

The thermochromic performance of both VO₂ annealed samples was also investigated by optical measurements as films on glass. In Figure 5, the transmittance spectra of film samples are presented, as acquired at temperatures below and above the critical transition temperature (25 °C and 90 °C, respectively). In addition, both the solar irradiance spectrum and the standard luminous efficiency function for photopic vision are presented in the same figure. For both VO₂ films, a decrease in IR transmittance can be observed upon heating, confirming their thermochromic behavior. More specifically, the IR switching at 2000 nm was almost 17% for both films, indicating that the film fabricated with the VO₂ powder from the large-scale reactor has a thermochromic performance similar to that with the VO₂ prepared in the lab-scale autoclave. Moreover, the transmittance solar modulation is calculated as 9% and 11% for the VO₂ films with powders from the reactor and the autoclave, respectively. The difference between these values could be related to the difference in crystallinity of the two samples. Finally, the luminous transmittance was calculated as 44% and 46%, while the visible transmittance at 600 nm was 47% and 49%, for the films with powders from the reactor and the autoclave, respectively. These values are in agreement with those reported in the literature concerning the hydrothermal method [35] and other methods, such as sputtering [31].

The differences between thermochromic properties of the films can be related to the presence of various vanadium oxide phases both before and after the annealing process, and the difference in morphology between the lab- and large-scale hydrothermally synthesized powders, as indicated by XRD patterns and SEM images, respectively. On the other hand, differences in structural and morphological characteristics of powdered samples can be attributed to the heating profile and characteristics of each hydrothermal system. The autoclave vessel that was used for the lab-scale hydrothermal synthesis was heated by placing it inside a furnace, while the large-scale hydrothermal reactor is equipped with electric resistances (coils) attached to its main body, which are used to heat the vessel. In addition, the quantity of material in the reactor was increased compared to that of the lab-scale autoclave. Overall, the thermochromic performance of VO₂ prepared in the large-scale reactor seems to be similar to that of the VO₂ prepared in the lab-scale autoclave, despite the slight differences in morphology or chemical composition of the vanadium oxide phases, showcasing the applicability of large-scale systems to the synthesis of thermochromic VO₂, supporting its wider implementation in a wide range of applications.

Finally, a comparative table of the thermochromic performance of VO₂ films prepared by various lab- and large-scale techniques is presented in

Table 2. Thermochromic performance of VO₂ films prepared by various methods.

Method	Material	TC (°C)	ΔTrIR (%)	ΔTrsol (%)	Trlum (%)	Ref.
Thermal Oxidation (large-scale)	VO2	-	69.3 (@2000 nm)	-	-	[30]
Sputtering (large-scale)	W:VO2	22	-	10	46	[31]
Sputtering	VO2	-	-	7	30	[36]
Sputtering	VO2	62	74 (@2500 nm)	5.87	38.77	[37]
Pulsed Laser Deposition	VO2	60	66.26 (@2600 nm)	-	-	[38]
Sol–Gel	VO2	72	-	8.9	56.4	[16]
Hydrothermal	VO2	64		7.27	53.68	[39]
Microwave Hydrothermal	VO2	50		20.73	55.04	[24]
Hydrothermal Synthesis (large-scale)	VO2	54	17 (@2000 nm)	9	44	This work

. We can observe that the hydrothermally synthesized VO₂ films display similar thermochromic properties to those prepared by physical vapor deposition methods (thermal oxidation and sputtering). In addition, the critical transition temperature of the VO₂ films prepared by the large-scale reactor is lower than films prepared by other techniques (in lab- or large-scale) using undoped VO₂, while both luminous transmittance and solar transmittance modulation have values comparable to those of other works referenced (Table 2). In contrast, the IR switching is lower enough compared to those works. It should be noted that only the VO₂ films prepared by lab-scale microwave-assisted hydrothermal synthesis from the work of Okada et al. [24] exhibit far greater luminous transmittance and solar transmittance modulation while simultaneously exhibiting a lower critical transition temperature when compared to the large-scale VO₂ films of this work. This enhanced performance could be related to the differences between the microwave-assisted lab-scale hydrothermal system used to synthesize the VO₂ powders, or even to the different slurry, processing methods and deposition used to prepare the corresponding VO₂ films. Moreover, these VO₂ films were deposited on thinner PolyEthylene Terephthalate (PET) plastic substrates (50 μm), compared to the 1 mm thick glass substrates that were used in the current work. Both the heating procedure of the hydrothermal synthesis of VO₂ powders and the VO₂ slurry used to prepare the films can play a critical role in their final properties, and thus in their thermochromic performance. Taking into account that the hydrothermal synthesis can be used to obtain VO₂ in a powdered form, which can then be deposited as a film even at low temperatures (lower than 100 °C), compatible with plastic flexible substrates (e.g., PET), it can be concluded that the hydrothermal synthesis is an excellent technique for large-scale applications.

4. Conclusions

This comparative study presents, for the first time, the thermochromic properties of hydrothermally synthesized VO₂ powder, and the corresponding films, prepared via lab- and large-scale techniques. Both powders were polycrystalline, while the critical transition temperature was reduced in the case of the large-scale reactor sample. Furthermore, films prepared using either of the hydrothermally synthesized powders displayed a similar IR switching of 17% upon heating and a comparable solar transmittance modulation and luminous transmittance. Based on these results, hydrothermal synthesis is suggested as a promising technique for large-scale synthesis of VO₂ in powdered form that can be used for thermochromic applications.