An abnormal phase transition behavior in VO₂ nanoparticles induced by an M1–M2–R process: two anomalous high (>68 °C) transition temperatures

Abstract

Vanadium dioxide (VO₂) has a reversible metal–insulator transition (MIT) at 68 °C and can be used to develop thermally and electrically sensitive devices. In this study, an abnormal phase transition behavior of VO₂ nanoparticles was discovered during the comparison of pristine nanoparticles without and with high temperature thermal treatment. The single phase transition temperature at 65.1 °C for the pristine VO₂ nanoparticles split into two temperatures at approximately 74 °C (T₁) and 84 °C (T₂) after thermal treatment at 400 °C for 6 h. Both temperatures are much larger than 68 °C. Through characterization by Raman and transmission electron microscopy (TEM), the two higher transition temperatures could be well explained by the formation of VO₂ (M2). Grain boundaries were observed during the merger and fusion processes of VO₂ nanoparticles at high temperatures. The grain boundaries and interfacial defects resulted in the dislocation of the lattice structure and produced stress and strain in the VO₂ nanoparticles. Consequently, VO₂ (M2) with a higher temperature formed in the heating process and the initial MIT (M1–R) became an M1–M2–R transition. Moreover, the thermal treatment improves the phase transition enthalpy (ΔH) of VO₂, which promotes the increase in the solar modulation ability (ΔT_sol) of VO₂–PET composite film from 12.8% to 15.2–17.0% without loss in the luminous transmittance. These findings are of great significance to the deep understanding of MIT and the development of VO₂ smart windows.

---

1 Introduction

Vanadium dioxide, a strongly correlated electron material, is extremely interesting for exploration of its metal–insulator transition (MIT), as well as its various applications. The MIT occurs from a monoclinic (M1), insulating phase at low temperatures to a rutile (R), metallic phase at high temperatures at a critical temperature of 68 °C (341 K), accompanied by remarkable changes in electronic and optical properties.¹ Large numbers of applications have been proposed and employed such as smart windows,^2,3 Mott transistors,⁴ strain,⁵ gas⁶ and temperature⁷ sensors, and thermal actuators.⁸

For a series of applications, it is necessary to tailor the phase transition temperature (T_MIT) to meet different requirements. Thus various methods have been used to tune the T_MIT. Doping is widely employed to decrease the T_MIT by substitution of V⁴⁺ with W⁶⁺, Mo⁶⁺, Mg²⁺, F⁻ and Nb⁵⁺.^9–12 Moreover, the size effect¹³ via controlling the size of VO₂ nanoparticles could regulate the MIT of VO₂ and the strain along the c_R-axis could increase the T_MIT due to formation of the M2 phase from the M1 phase of VO₂.^14,15 Recent study also found that oxygen vacancies can be employed to tune the T_MIT over a wide range.¹⁶ In addition, other impacting factors, including crystallinity and stoichiometry, have been investigated to regulate the T_MIT.^17,18 Thus, the MIT and the corresponding T_MIT of VO₂ are critically influenced by the above factors.

Because VO₂ nanoparticles have different morphologies due to different hydrothermal conditions, it is complicated to distinguish the effects of those factors on the MIT of VO₂ nanoparticles. Chen et al.¹⁹ prepared fine crystal-quality VO₂ nanoparticles via the hydrothermal reaction at high temperatures (300–390 °C) and the obtained VO₂ nanoparticles showing an asymmetrical phase transition and increased insulator–metal transition temperatures. Two endothermic peaks located at 68 °C (low MIT peak) and 91 °C (high MIT peak) were observed unexpectedly during the DSC test. This phenomenon was ascribed to the size effect and the shape-dependence; however, it was still hard to clearly distinguish the function of the above factors. Li et al.²⁰ took advantage of the transformation from VO₂ (D) to VO₂ (M1) to obtain VO₂ (M1) nanoparticles via the combination of the hydrothermal synthesis and a subsequent mild thermal treatment at 300–450 °C, and found that interfacial defects and the size effect played essential roles in regulating the MIT of VO₂.

In this study, we found an abnormal phase transition process of VO₂ nanoparticles. The single peak at 65.1 °C in the phase transition of pristine VO₂ nanoparticles splits into two peaks after thermal treatment at high temperatures (400–440 °C) for a certain time. Two peaks of samples obtained from annealing at 400 °C for 6 hours are located approximately at 74 °C (the low temperature, T₁) and 84 °C (the high temperature, T₂) which are both higher than 68 °C for bulk VO₂. The reason for the increased MIT temperatures could be attributed to the formation of VO₂ (M2) and will be explained in detail in this study. On the other hand, the thermal treatment improves the ΔH of VO₂ and the ΔT_sol of VO₂–PET composite film. The ΔT_sol increases from 12.8% to 15.2–17.0% without loss of luminous transmittance. This study could enhance the understanding of the MIT of VO₂ nanoparticles and promote the development of VO₂–PET flexible foil.

2 Experimental methods

2.1 The preparation of VO₂ (M1) powders

VO₂ nanoparticles were prepared according to a previously reported method.² All reagents were purchased from the Sinopharm Chemical Reagent Co., Ltd. and used without further purification. Vanadium pentoxide (V₂O₅, analytically pure) and diamide hydrochloride (N₂H₄·HCl, analytically pure) were employed as starting materials to prepare a VO²⁺ solution. Concentrated HCl (6 mL, 38%) and a solution containing 1 g of N₂H₄·HCl were added to an aqueous suspension (20 mL) containing 3.5 g of V₂O₅. The solution was treated with a small amount of V₂O₅ or N₂H₄·HCl until it contained no VO²⁺ or V³⁺ and was then filtered to form a clear VO²⁺ solution (pH = 1). The solution was stirred for 10 min and then transferred to a 50 mL stainless steel autoclave. The hydrothermal reaction was carried out at 260 °C for 6 h. The final black product was separated by centrifugation, washed with water and ethanol and then placed in a vacuum drying oven at 60 °C for 24 h. The obtained VO₂ (M1) nanoparticles were heated in a tube furnace under a flow of nitrogen gas at 400–440 °C for a predefined time to study the effects of annealing time and annealing temperature.

2.2 Characterization

The morphologies of the resulting powders were analysed via transmission electron microscopy (TEM, JEM2010, JEOL, Japan) and scanning electron microscopy (SEM, Magellan 400). The crystalline phases of the nanoparticles were determined by X-ray diffraction (XRD, Model D/Max 2550 V, Rigaku, Japan). The phase transition temperatures of the products were measured via differential scanning calorimetry (DSC, DSC200F3, NETZSCH) in a nitrogen flow at a heating rate of 10 °C min⁻¹. The Raman spectra of the composite samples were obtained using a Raman microscope (Renishaw inVia) with a 514 nm laser source at an input power of 1 mW. The thermochromic properties were evaluated by the VO₂–PET composite films. For measurements, the VO₂ powders were uniformly dispersed in polyurethane after surface modification by poly(vinylpyrrolidone) (PVP) and then deposited on polyethylene terephthalate (PET) via knife-coating and finally dried at 70 °C. The optical transmittance characteristics were monitored using a Hitachi U-4100 UV-visible-near-IR spectrophotometer equipped with a film heating unit in the wavelength range of 350–2600 nm. For all samples, the integral visible transmittance (T_lum, 350–750 nm) and solar transmittance (T_sol, 240–2600 nm) were obtained based on the measured spectra using the following equations:

ΔT_sol = T_sol(T < T_c) − T_sol(T > T_c) (2)

where T(λ) denotes the transmittance at wavelength λ, i denotes ‘lum’ or ‘sol’ for the calculations, T and ΔT_sol are the temperature and solar modulation ability, respectively; φ_lum is the standard luminous efficiency function for the photopic vision according to CIE 1931 standards, and φ_sol is the solar irradiance spectrum for the air mass 1.5 (corresponding to the sun standing 37° above the horizon) according to ASTM G173-03 Reference Spectra.

3 Results and discussion

3.1 The effect of annealing time and temperature on the MIT of VO₂

In Fig. 1a, the pristine VO₂ nanoparticles prepared through a hydrothermal method at 260 °C for 6 h were indexed to VO₂ (M1) (JCPDS no. 72-0514) without other impure phases. The phase transition temperature is 65.1 °C (Fig. 1b), which is slightly lower than the corresponding transition temperature for bulk VO₂ (68 °C). To the best of our knowledge, the post-annealing has been widely used to transform other phases (A, B, D) of VO₂ into VO₂ (M1).^20–22 In addition, a series of phase transition behaviors have been discovered for the VO₂ (A, B, D) nanoparticles, depending on the post-annealing conditions. However, thermal treatment for only VO₂ (M1) nanoparticles has not been studied yet. Herein, we focused on the thermal treatment for pristine VO₂ (M1) nanoparticles. The novel phase transition behaviors will be discussed in the following.

Fig. 1 (a) XRD patterns and (b) DSC curves of VO₂ nanoparticles after annealing at 400 °C for different times; the dashed lines were obtained via the Gaussian fitting method to analyze the different peaks.

The thermal treatment was carried out in a tube furnace under a flow of nitrogen gas at 400 °C to investigate the effect of annealing time. After annealing for 6 h, 12 h, 24 h, 48 h, VO₂ (M1) maintains the monoclinic phase without obvious change and only some weak peaks of V₃O₇ appear in Fig. 1a. It indicates that some of VO₂ (M1) is oxidized to V₃O₇ rather than V₂O₅ at high temperatures in an N₂ atmosphere. As is well known, thermal treatment can effectively improve the crystallinity of metal oxides, by the accompanying grain growth and reduction of point defects. Thus the (011) peak of VO₂ (M1) gradually increases when heating from 0 h to 24 h, but abnormally decreases when heating for 48 h. It can be well explained by the serious oxidation of VO₂, which is verified by the obvious increase in the peaks of V₃O₇ in Fig. 1a.

The phase transition behaviors of annealed VO₂ nanoparticles were characterized by DSC in Fig. 1b and the DSC curves were operated with peak-fit processing via the Gaussian fitting method. It is amazing that the DSC curve emerges with two peaks and the latent heat of metal–insulator transition (ΔH) is noticeably enhanced after annealing in Table 1. A single T_MIT of pristine VO₂ (M1) nanoparticles splits into two T_MIT's and the two T_MIT's both exceed 65.1 °C. T₁ is located around 74 °C without obvious fluctuation when the annealing time ranged from 6 to 48 h, whereas the ΔH₁ (corresponding to the peak at low temperatures) gradually decreases from 23.9 to 11.2 J g⁻¹. Compared to T₁, T₂ is located around 84 °C. The ΔH₂ (corresponding to the peak at high temperatures) increases from 16.2 to 19.7 J g⁻¹, opposite to the change in ΔH₁. This is an extraordinary and interesting phenomenon in the MIT behaviors of VO₂.

Table 1 The properties of T_MIT and ΔH of the pristine VO₂ and the VO₂ after annealing at 400 °C for different times

Sample	T1 (°C)	T2 (°C)	ΔH1 (J g^−1)	ΔH2 (J g^−1)	ΔH1 + ΔH2	ΔH1/ΔH2
Pristine VO2 (M)	TMIT = 65.1 °C		—	—	21.7	—
400 °C – 6 h	74.0	84.0	23.9	16.2	40.1	1.474
400 °C – 12 h	74.8	84.6	21.3	18.1	39.5	1.174
400 °C – 24 h	74.7	84.2	18.2	18.5	36.7	0.984
400 °C – 48 h	74.9	84.2	11.2	19.7	30.9	0.565

---

The effect of annealing temperatures on MIT of VO₂ was also then investigated. Similar phenomena were also found when annealing for 6 h in the range of 400–440 °C. In Fig. 2a, as the annealing temperature increases from 400 to 440 °C, the samples remain M1 phase and the (011) peaks of VO₂ (M1) become more intense due to the increase of the crystallinity of the VO₂ nanoparticles. In Fig. 2b, a single T_MIT (65.1 °C) also splits into two T_MIT, which are both above 65.1 °C. T₁ increases from 74.0 to 76.2 °C and T₂ decreases from 84.0 to 82.4 °C when the annealing temperature increases from 400 to 440 °C. Moreover, the ΔH₁ decreases from 23.9 to 11.4 J g⁻¹ and the ΔH₂ increases from 16.2 to 23.3 J g⁻¹, as shown in Table 2. We can also find a tendency that T₁ and T₂ peaks gradually merge with the improvement of annealing temperature in Fig. 2b. It seems like the two T_MIT peaks combine into one T_MIT peak (82.0 °C). In summary, the MIT temperatures of VO₂ would divide into two transition temperatures higher than 65.1 °C when the VO₂ samples are annealed at high temperatures.

Fig. 2 (a) XRD patterns and (b) DSC curves of VO₂ nanoparticles after annealing at different temperatures for 6 h; the dashed lines were obtained via the Gaussian fitting method to analyze the different peaks.

Table 2 The properties of T_MIT and ΔH of the pristine VO₂ and the VO₂ after annealing at different temperatures for 6 h

Sample	T1 (°C)	T2 (°C)	ΔH1 (J g^−1)	ΔH2 (J g^−1)	ΔH1 + ΔH2	ΔH1/ΔH2
Pristine VO2 (M)	TMIT = 65.1 °C		—	—	21.7	—
400 °C – 6 h	74.8	84.6	23.9	16.2	40.1	1.474
420 °C – 6 h	76.5	84.1	18.5	15.8	34.4	1.175
440 °C – 6 h	76.2	82.4	11.4	23.3	34.7	0.488

---

3.2 The mechanism for the novel transition behavior of VO₂

According to the previous studies, the influence of phase transition temperature could be attributed to a series of effects, including defects, size effects and crystallinity.^13,18,20,23 According to the abovementioned findings, when the annealing time and temperature increase, the small nanoparticles merge and grow into large ones. Thereupon, the particle size enlarges, the point defect concentration decreases and the VO₂ crystallinity improves. In terms of a heterogeneous nucleation process for VO₂, the phase transition depends on the availability of a suitable nucleating defect in the sample space.²⁴ Thus a higher T_MIT induced by a reduced number of defects in the pristine VO₂ nanoparticles will be realized via the thermal treatment. However, the transition temperature of single crystal bulk VO₂ is 68 °C. In any event, the increase of T_MIT of VO₂ cannot exceed 68 °C, let alone reach 74 °C and 84 °C. Therefore, it is hard to explain our observed phenomenon of two increasing transition temperatures by these factors.

However, the formation and properties of another monoclinic phase (M2) of VO₂ in VO₂-based strain sensors and actuators were uncovered by Wang et al. and Wu et al., respectively.^14,15,25 The M2 phase with a high transition temperature (>68 °C) under the effect of stress and strain could be interpreted as an intermediate and transitional phase between the M1 and R phases.²⁶ Under tensile strain, the high transition temperature is derived from an insulator (M1)–insulator (M2) phase transition before the insulator (M1)–metal (R) phase transition. For VO₂ nanoparticles, however, the elusive M2 phase is almost impossible to measure and characterize in a single VO₂ particle of few tens of nanometers. Therefore, the Raman spectra for VO₂ samples were measured to record the phase transition changes of VO₂ nanoparticles. For pristine VO₂ samples, the feature peaks at 190, 222, 310, 387, and 610 cm⁻¹ denoted by blue dashed lines can be ascribed to the vibration modes in VO₂ (M1)⁵ in Fig. 3a. For VO₂ nanoparticles prepared by annealing at 400 °C for 6 h, except for feature peaks of the M1 phase, the peak at 645 cm⁻¹ corresponding to the M2 phase is found in Fig. 3b. In other words, the annealed VO₂ nanoparticles have an M1–M2–R transition during the heating process. The formation of M2 phase is always caused by the generation of stress and strain during annealing at high temperatures, resulting in the abnormal increase of transition temperatures. Also, in the heating range from 70 to 90 °C, the M1 and M2 phases coexist in Fig. 3b. It is obvious that the T₁ peak around 74 °C corresponds to the temperature region from 70 to 80 °C and the T₂ peak around 84 °C corresponds to the temperature region from 80 to 90 °C in Fig. 3b. Therefore, when heating from 70 to 80 °C, one part of VO₂ has the M1–M2–R phase transition and the other part remains unchanged; heating above 80 °C, the rest of the VO₂ (M1) also experiences the M1–M2–R phase transition, indicating the thermally treated VO₂ (M1) sample holds two different states. Consequently, two high and different phase transition temperatures are observed in the VO₂ (M1) samples.

Fig. 3 Characteristic Raman spectra for (a) pristine VO₂ nanoparticles and (b) VO₂ nanoparticles after annealing at 400 °C for 24 h.

3.3 The morphologies of the annealed VO₂ nanoparticles

Fig. 4 shows the SEM images of pristine and annealed VO₂ nanoparticles. The pristine samples (Fig. 4a) are agglomerative VO₂ nanoparticles with the size of approximately 35 nm and the vague surface seems to be surrounded by amorphous substances with quantities of defects. After annealing at 400 °C for different periods (Fig. 4b–d), the amorphous substances gradually disappear from the surface and the nanoparticles show sharp outlines. This is possibly caused by the recrystallization of VO₂ nanoparticles, the merging of small particles and further growth into large ones. Thus the particle size gradually increases from 35 nm to 100 nm with the annealing time changing from 0 to 48 h. In Fig. 4d, the fusion of small particles can be clearly observed in the red circle. Naturally, it will produce many grain boundaries and dislocation defects during the fusion process. Likewise, a similar merger and growth of small nanoparticles occurs at a higher annealing temperature (440 °C) and the shape boundaries are also obvious in Fig. 5.

Fig. 4 The SEM images of VO₂ nanoparticles after annealing at 400 °C for (a) 0 h; (b) 6 h; (c) 12 h; and (d) 48 h.

Fig. 5 The SEM images of VO₂ nanoparticles after annealing at 440 °C for 6 h.

The TEM analyses also clearly reveal crystallographic merger and fusion for VO₂ nanoparticles. Fig. 6a shows the pristine VO₂ nanoparticles around 35 nm, the large ones are possibly the reunion of small particles. The VO₂ nanoparticles then grow larger to 100 nm (Fig. 6b–d) after annealing. However, the fusion process of grain boundaries between two particles can be found in the blue circles in Fig. 6b–d. In specific, the agglomerative VO₂ nanoparticles (Fig. 4a) surrounded with amorphous substances connect together and arrange more closely. Annealing at high temperatures, the amorphous substances recrystallize, leading to the merger of two adjacent particles. If the interface between the two particles has a different crystallographic orientation or a lattice misfit, a grain boundary will be produced naturally. Accompanied by the generation of interfacial defects, the particles will produce stress and strain due to the dislocation of the lattice between the two particles. We can find a large VO₂ particle combined with several small particles in Fig. 6e and two particles arranged closely and combined together in Fig. 6f.

Fig. 6 The TEM images of VO₂ nanoparticles after annealing at 400 °C for (a) 0 h; (b) 6 h; (c) 12 h; (d–f) 48 h.

To further reveal the detailed structural changes of grain boundaries, the high-resolution TEM images and the selected area electron diffraction (SAED) patterns of the annealed VO₂ nanoparticles are shown in Fig. 7. The lattice dislocations between two particles can be clearly observed in Fig. 7a–d. Fig. 7a shows a twin boundary with the lattice misfit in the (200) lattice plane and a misfit angle of 44.8°. The yellow part (above the boundary) and the blue part (below the boundary) have the same [011] zone axis in a monoclinic phase and grow along the (011) preferential direction. The interplanar spacing is calculated to be 0.312 nm, which is indexed as the (011) facet of VO₂ (M1). The corresponding diffraction pattern (the inset in Fig. 7a) obtained by Fourier transform of the HRTEM lattice including the twin boundary presents two lattices. They both have the same structural orientation. The interplanar spacings characterized in Fig. 7a–c are 0.312, 0.311 and 0.309 nm, respectively. They are slightly smaller than the (011) interplanar spacing (3.2 nm) of normal VO₂ (M1). This means that there is an extra stress and strain along the tetragonal a-axis of the M1 phase. The strain values determined by ε = (d_standard − d)/d_standard are 2.5%, 2.28% and 3.13%, which are large enough to drive the M1–M2–R transition behavior and to further raise the transition temperature. Therefore a higher transition temperature will be presented compared to pristine nanoparticles.

Fig. 7 Lattice-resolved HRTEM images of the boundaries of annealed VO₂ nanoparticles at 400 °C for (a) 6 h, the insets are the SAED pattern and the selected TEM image respectively; (b) 12 h; (c) 48 h; and (d) 48 h.

Through the abovementioned characterizations, including DSC, Raman spectra, SEM and TEM, two increased transition temperatures (T₁ and T₂) can be further explained clearly as follows. During the grain growth of VO₂ nanoparticles, grain boundaries will generate spontaneously and also can be eliminated by dislocation movement and lattice rearrangement. Therefore, stress of different magnitudes exists on the VO₂ nanoparticles. For T₂ (corresponding to the incompletely grown VO₂ nanoparticles), the stress and strain caused by grain boundaries lead to a M1–M2–R behavior and greatly increases the transition temperature to around 84 °C, but for T₁ (corresponding to the completely grown VO₂ nanoparticles), although the grain boundaries disappear via lattice rearrangement, residual stress still exists in the VO₂ nanoparticles and is smaller than the stress induced by grain boundaries. Therefore T₁ (74 °C) is lower than T₂ (84 °C) but slightly larger than 68 °C for bulk VO₂.

Moreover, because the ΔH₁ and ΔH₂ correlate to VO₂ nanoparticles without and with grain boundaries, the ration ΔH₁/ΔH₂ reveals the proportions of each part of the VO₂ nanoparticles. The data for ΔH₁ and ΔH₂ in Tables 1 and 2 are summarized in Fig. 8a and b. We find that the value of ΔH₁/ΔH₂ decreases from 1.474 to 0.565 with annealing from 6 to 48 h at 400 °C and decreases from 1.474 to 0.488 with increasing annealing temperature from 400 to 440 °C for 6 h. This indicates that the smaller ΔH₁/ΔH₂ means the production of more VO₂ nanoparticles with grain boundaries. Likewise, from the SEM and TEM images of Fig. 4a–d and 5a–d, the number of VO₂ nanoparticles with grain boundaries in blue circles obviously increases with increasing annealing time. It reveals that the merger of VO₂ nanoparticles with grain boundaries dominates with increased annealing temperature or increased annealing time.

Fig. 8 Correlations of ΔH₁ and ΔH₂ for the annealed VO₂ nanoparticles.

3.4 The optical property of the annealed VO₂ nanoparticles

Fig. 9a and Table 3 show the variation in the optical and thermochromic properties of the VO₂–PET composite films for different annealing conditions. The VO₂ solid content and the film thickness of these foils are fixed. The composite film prepared by pristine VO₂ nanoparticles exhibits solar modulation ability (ΔT_sol) of 12.8% and luminous transmittances (T_lum) of 50.5% at 20 °C and 48.4% at 100 °C. For the annealed VO₂ nanoparticles, however, the ΔT_sol significantly improves without loss of the T_lum value. The T_lum value slightly increases to 51.0–52.2% at 20 °C and 48.5–49.5% at 25 °C after annealing. The transmittance of annealed VO₂ is much lower than the transmittance of pristine VO₂ at 900–2600 nm in Fig. 9a. The solar transmittance (T_sol) at 100 °C decreases from 45.9% to 40.9–41.7%, whereas the T_sol at 20 °C remains fairly stable. Therefore, the ΔT_sol significantly increases from 12.8% to 15.2–17.0%. The phase transition enthalpy (ΔH) increased from 21.7 J g⁻¹ to 36.7–40.1 J g⁻¹ after thermal treatment (see Table 1). Thus, the ΔT_sol and the ΔH of VO₂ have a positive correlation in Fig. 9b. Therefore, thermal treatment can improve the solar modulation ability without loss of luminous transmittance.

Fig. 9 (a) The transmittance spectra of VO₂–PET composite films at 20 and 90 °C for different annealing conditions; (b) the relation between the solar modulation ability (ΔT_sol) and the phase transition enthalpy (ΔH).

Table 3 The solar energy control properties of VO₂–PET composite films for different annealing conditions

Annealing condition	Tlum (%)		Tsol (%)		ΔTsol (%)
	20 °C	100 °C	20 °C	100 °C
Pristine VO2	50.5	48.4	58.6	45.9	12.8
400 °C – 6 h	51.7	48.8	57.8	40.9	17.0
400 °C – 12 h	51.0	49.5	56.6	41.4	15.2
400 °C – 24 h	52.2	48.5	57.4	41.7	15.7

---

4 Conclusions

An abnormal phase transition process of VO₂ nanoparticles was discovered during the comparison of pristine nanoparticles without and with high temperature thermal treatment. One transition temperature (T_c = 65.1 °C) of pristine VO₂ nanoparticles splits into two transition temperatures (T₁ around 74 °C and T₂ around 84 °C). T₂ corresponds to one part of the VO₂ nanoparticles containing grain boundaries. The grain boundaries, induced by the merger and fusion of VO₂ nanoparticles at high temperature, lead to the generation of stress and strain for the VO₂ nanoparticles, resulting in an M1–M2–R transition behavior for VO₂. Conversely, T₁ corresponds to the other part of completely grown VO₂ nanoparticles wherein the residual stress still remains after merger and fusion. Thus, the two transition temperatures increase notably and T₁ is lower than T₂. Moreover, the solar modulation ability (ΔT_sol) of VO₂–PET composite film increases from 12.8% to 15.2–17.0% and the luminous transmittance (T_lum) at 20 °C slightly increases from 50.5% to 51.0–52.2%. This study could enhance the understanding of the MIT of VO₂ nanoparticles and promote the development of VO₂–PET flexible foil.

Acknowledgements

This study was supported in part by funds from MOST (2014AA032802), NSFC (State Outstanding Young Scholars, 51325203) and Shanghai Municipal Science and Technology Commission (15XD1501700, 13NM1402200, 13521102100).

References

1. F. J. Morin, Phys. Rev. Lett., 1959, 34–36, 34–36.
2. Y. Gao, C. Cao, L. Dai, H. Luo, M. Kanehira, Y. Ding and Z. L. Wang, Energy Environ. Sci., 2012, 5, 8708.
3. Y. F. Gao, H. J. Luo, Z. T. Zhang, L. T. Kang, Z. Chen, J. Du, M. Kanehira and C. X. Cao, Nano Energy, 2012, 1, 221–246.
4. S. Hormoz and S. Ramanathan, Solid-State Electron., 2010, 54, 654–659.
5. B. Hu, Y. Ding, W. Chen, D. Kulkarni, Y. Shen, V. V. Tsukruk and Z. L. Wang, Adv. Mater., 2010, 22, 5134–5139.
6. E. Strelcov, Y. Lilach and A. Kolmakov, Nano Lett., 2009, 9, 2322–2326.
7. S. Lee, C. Cheng, H. Guo, K. Hippalgaonkar, K. Wang, J. Suh, K. Liu and J. Wu, J. Am. Chem. Soc., 2013, 135, 4850–4855.
8. J. Cao, W. Fan, Q. Zhou, E. Sheu, A. Liu, C. Barrett and J. Wu, J. Appl. Phys., 2010, 108, 083538.
9. X. G. Tan, T. Yao, R. Long, Z. H. Sun, Y. J. Feng, H. Cheng, X. Yuan, W. Q. Zhang, Q. H. Liu, C. Z. Wu, Y. Xie and S. Q. Wei, Sci. Rep., 2012, 2, 466.
10. T. J. Hanlon, J. A. Coath and M. A. Richardson, Thin Solid Films, 2003, 436, 269–272.
11. J. Zhou, Y. Gao, X. Liu, Z. Chen, L. Dai, C. Cao, H. Luo, M. Kanahira, C. Sun and L. Yan, Phys. Chem. Chem. Phys., 2013, 15, 7505–7511.
12. L. Dai, S. Chen, J. Liu, Y. Gao, J. Zhou, Z. Chen, C. Cao, H. Luo and M. Kanehira, Phys. Chem. Chem. Phys., 2013, 15, 11723.
13. L. Dai, C. Cao, Y. Gao and H. Luo, Sol. Energy Mater. Sol. Cells, 2011, 95, 712–715.
14. H. Guo, K. Chen, Y. Oh, K. Wang, C. Dejoie, S. A. Syed Asif, O. L. Warren, Z. W. Shan, J. Wu and A. M. Minor, Nano Lett., 2011, 11, 3207–3213.
15. K. Wang, C. Cheng, E. Cardona, J. Y. Guan, K. Liu and J. Q. Wu, ACS Nano, 2013, 7, 2266–2272.
16. J. Jeong, N. Aetukuri, T. Graf, T. D. Schladt, M. G. Samant and S. S. Parkin, Science, 2013, 339, 1402–1405.
17. M. H. Lee and M. G. Kim, Thin Solid Films, 1996, 286, 219–222.
18. Z. Lu, C. Li and Y. Yin, J. Mater. Chem., 2011, 21, 14776.
19. Z. Chen, Y. Gao, L. Kang, C. Cao, S. Chen and H. Luo, J. Mater. Chem. A, 2014, 2, 2718.
20. M. Li, X. Wu, L. Li, Y. Wang, D. Li, J. Pan, S. Li, L. Sun and G. Li, J. Mater. Chem. A, 2014, 2, 4520.
21. Y. Zhang, M. Fan, F. Niu, Y. Zhong, C. Huang, X. Liu, B. Wang and H. Li, Micro Nano Lett., 2011, 6, 888.
22. R. Li, S. Ji, Y. Li, Y. Gao, H. Luo and P. Jin, Mater. Lett., 2013, 110, 241–244.
23. J. Cao, E. Ertekin, V. Srinivasan, W. Fan, S. Huang, H. Zheng, J. W. L. Yim, D. R. Khanal, D. F. Ogletree, J. C. Grossmanan and J. Wu, Nat. Nanotechnol., 2009, 4, 732–737.
24. R. Lopez, T. E. Haynes, L. A. Boatner, L. C. Feldman and R. F. Haglund, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 65, 224113.
25. J. Cao, W. Fan, K. Chen, N. Tamura, M. Kunz, V. Eyert and J. Wu, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 82, 241101.
26. J. I. Sohn, H. J. Joo, D. Ahn, H. H. Lee, A. E. Porter, K. Kim, D. J. M. Kang and E. Welland, Nano Lett., 2009, 9, 3392–3397.

---