High-throughput roll-to-roll fabrication of flexible thermochromic coatings for smart windows with VO2 nanoparticles

Youngkwang Kim a, Sangbae Yu a, Jaeseoung Park a, Daseob Yoon a, Amir Masoud Dayaghi a, Kun Joong Kim a, Jin Soo Ahn b and Junwoo Son *a
aDepartment of Materials Science and Engineering, Pohang University of Science and Technology (POSTECH), Pohang 37673, Republic of Korea. E-mail: jwson@postech.ac.kr
bResearch Institute of Industrial Science and Technology (RIST), Pohang 37679, Republic of Korea

Received 22nd December 2017 , Accepted 26th February 2018

First published on 26th February 2018

VO2-based ‘nanothermochromics’ that utilize the dispersion of VO2 nanoparticles in a passive host matrix have been evaluated as an economic strategy of “smart” windows to reduce energy consumption for heating and air conditioning in buildings. Here, we demonstrate a high-throughput roll-to-roll fabrication of thermochromic coatings for smart windows that can adapt their optical properties in accordance with external temperature. A large quantity (250 g) of VO2 nanoparticles (NPs) was synthesized at one time by controlled thermal treatment of bead-milled V2O5 NPs as a fast and inexpensive method. The amorphous nature of bead-milled V2O5 NPs combined with their nanometer size kinetically facilitates uniform synthesis of high-quality VO2 NPs even under less-reducing conditions than those used to obtain bulk VO2. This mass production of VO2 NPs could be used to fabricate the largest thermochromic coatings with VO2/PVP composites (12 cm × 600 cm) yet produced with excellent infrared modulation ability (∼45%). This scalable and continuous production of large coatings with thermochromic NPs will accelerate the commercialization of thermochromic coatings for smart windows, which will contribute to a large reduction in energy consumption to heat or cool buildings.


Buildings lose a large amount of energy through their windows. One way to reduce this loss is to develop energy-saving “smart” windows that can adapt their optical transmittances in response to environmental changes.1,2 Smart windows with thermochromic coatings dynamically control the solar heat transmittance into buildings by reversibly switching between an infrared-transparent state (when cold) and an infrared-reflective state (when hot) in response to the changes in the environmental temperature. This behavior could significantly alleviate the need for heating and air-conditioning, and thereby significantly reduce energy consumption to maintain the internal temperature.3,4 Vanadium dioxide (VO2) with a first-order phase transition has been regarded as a promising candidate material for use in themochromic smart windows, because of its large and reversible modulation of transmittance in the near-infrared region at metal–insulator (MI) transition temperatures (∼68 °C);5,6 solar heat radiation induced by infrared light can be regulated self-consistently depending on the external temperature without the loss of visible transmittance.7–9

Many strategies have been developed for the preparation of VO2-based thermochromic coatings. Most previous studies have focused on simple deposition of VO2 solid thin films on glass by pulsed laser deposition, sputtering, or chemical vapor deposition,10–12 but the high cost of large-scale production using vacuum processes hinders their practical application for thin-film-based thermochromic coatings. An alternative method is to use so-called ‘nanothermochromics’ that exploit the dispersion of VO2 nanoparticles (NPs) in a passive host matrix, and their application as coatings on glass.13–16 Compared to thin-film-based coatings, nanothermochromic coatings can be produced using low-cost solution processing, and may increase the visible transmittance and infrared modulation efficiency of thermochromic windows if the shape of VO2 NPs is controlled.17 In addition, various dopants and core–shell nanostructures in VO2 NPs could be utilized to enhance their optical properties and their chemical stabilities.14,18,19 However, several drawbacks of the current methods to synthesize VO2 NPs and to fabricate thermochromic coatings impede practical application of the nanothermochromic technology. Several methods to synthesize VO2 NPs have been reported,16,20–24 but all have small NP yields,8 long reaction times,16,20,21 complicated reactions and processes,24,25 or expensive starting materials.16 Moreover, because large-area production of VO2 NP-based thermochromic coating is not yet possible, the size of thermochromic coating has been limited (≤∼400 cm2).13,16,25 Thus, thermochromic coating on a large area has not been demonstrated using roll-to-roll (R2R) processing for mass production of commercial smart windows.

In this study, we report the high-throughput fabrication of thermochromic coating for smart windows using the mass production of VO2 NPs synthesized by simple and inexpensive methods (Fig. 1). A large quantity of high-quality VO2 NPs was synthesized by delicately controlled thermal treatment of bead-milled V2O5 NPs, which were obtained by grinding micro-sized V2O5 particles. The amorphous nature of bead-milled V2O5 NPs combined with size effects significantly reduces the activation energy for the phase transformation to VO2, and thereby kinetically facilitates the uniform synthesis of high-quality VO2 NPs even under less-reducing conditions than are used to obtain bulk VO2 from V2O5. This process to mass-produce VO2 NPs can be used in R2R methods to demonstrate the production of large sheets (12 cm × 600 cm) of flexible thermochromic coatings, which show an excellent infrared modulation ability (∼45%) over their entire area. This new strategy to synthesize oxide NPs and fabricate large-area thermochromic coatings suggests a simple and inexpensive approach for the commercialization of VO2-based thermochromic coatings for energy-efficient smart windows.

image file: c7tc05876d-f1.tif
Fig. 1 Schematic illustration of procedures to produce large-area VO2 nanothermochromic coatings. (a) Top-down synthesis of VO2 nanoparticles from commercial V2O5. (b) Photograph of castable slurry for tape casting. (c) Schematic illustration of tape-cast processing for flexible VO2–PVP composite coatings. (d) A photograph of flexible VO2–PVP composite coatings.

Results and discussion

High-quality VO2 NPs for thermochromic applications were synthesized in a top-down process by the reduction of abundant V2O5 powder. Vanadium ions tend to be stabilized as a +5 valence state, so V2O5 phases are thermodynamically the most stable vanadium oxides under atmospheric conditions, which makes V2O5 powders much cheaper than VO2 powders. However, direct reduction of micro-sized V2O5 powders to synthesize pure VO2, which is required for excellent switching of infrared transmittance for thermochromic applications, was very challenging because the process requires high temperatures (>1200 °C) and long reaction times (several days) (Fig. 4b).26–28 Furthermore, V2O5 and other vanadium sub-oxides co-exist after reduction of V2O5 micro-sized powders, probably because oxygen tends to be lost more easily from near the surfaces of the particles than from their cores (Fig. S1, ESI). This non-uniform reduction leads to an uneven distribution of the amount of oxygen, and that of valence states of vanadium ions in the particles. Due to the complex stoichiometry of vanadium oxides, the thermodynamic condition for the uniform phase transformation (V2O5 to VO2) appears to be very difficult simply by controlling the partial pressure of oxygen during thermal treatment when micron-sized powders are used.

To achieve uniform reduction of V2O5 particles, the size of commercially available V2O5 particles (250 g) can be effectively decreased using ethanol-based bead-milling.29 The positions of X-ray diffraction (XRD) peaks for the V2O5 phase did not shift (JCPDS #41-1426, λ (CuKα1) = 1.5406 nm), but their width increased and their intensity decreased as bead-milling time increased; this trend indicates that milling decreased the size of V2O5 particles at the expense of their amorphization (Fig. 2a and Fig. S2, ESI). Because the beads used for bead-milling had diameters <200 μm, the size of milled particles can be expected to decrease to ∼30 nm after bead-milling.29 However, physical collisions between the beads and micro-sized particles induce structural disorders and amorphization during milling; these processes lead to the reduction in intensity and the broadening of X-ray Bragg peaks.

image file: c7tc05876d-f2.tif
Fig. 2 Structural and compositional characterization of synthesized VO2 nanoparticles. (a) Change of X-ray diffraction after bead-milling: the increase of width and the decrease of intensity in V2O5 X-ray peaks (JCPDS #41-1426) indicate that the size of V2O5 particles decreased with the milling time at the expense of amorphization by milling damage. (b) Change of X-ray diffraction in bead-milled nanoparticles after the reduction process: the process yielded pure and uniform VO2 (JCPDS #40-1051) without other vanadium oxides. (c) Raman spectrum and (d) XPS spectrum of synthesized VO2 (M) nanoparicles.

To synthesize high-quality VO2 NPs, these amorphous V2O5 NPs were thermally treated at 500 °C either for <10 min under N2 gas or for 1 h under 130 Torr of air. Remarkably, even though both thermal treatments are thermodynamically insufficient to reduce V2O5 to VO2 in the bulk form, prominent new peaks appeared at 27.79°, 36.97°, 42.07°, 55.27°, and 57.42° in the XRD scan in the NPs (Fig. 2b); these peaks correspond to those of the VO2 (M) phase with lattice parameters a = 5.7517 Å, b = 4.5378 Å, and c = 5.3825 Å (JCPDS #43-1051). The absence of peaks related to other vanadium oxides with different valence states or polymorphs of VO2 indicates that this simple heat treatment spontaneously synthesized pure and uniform VO2 phases. Furthermore, unlike V2O5 nanoparticles before heat treatment, the peaks show high intensities, which is evidence that the transformed VO2 NPs exhibit high crystallinity. This result indicates that sufficient thermal energy during heat treatment not only switches the valence states of vanadium dioxides (+5 to +4), but also heals the structural disorder in the amorphous phase by rearranging the atoms into the most stable sites in the crystal lattices.

The structural properties of the synthesized VO2 NPs were further confirmed by Raman spectroscopy (Fig. 2c). The Raman peaks of the VO2 NPs occurred at 190, 222, 330, 336, 438 and 614 cm−1. The bands in the data were attributed to the vibration modes in monoclinic VO2; no other phases of VO2 were observed.30,31 The valence state of the synthesized VO2 NPs was confirmed by X-ray photoemission spectroscopy (XPS) (Fig. 2d). The V2p core level peaks of the NPs were measured to be ∼516.27 eV for 2p3/2, and ∼523.51 eV for 2p1/2; both can be attributed to V4+.32

The size distribution and localized crystal structures of the synthesized VO2 NPs were examined using transmission electron microscopy (TEM). The particles were spherical with an average diameter of 38 nm and narrow size distribution (30–60 nm) (Fig. 3b), which is close to the result shown in the field emission scanning electron microscopy (FESEM) image (Fig. 3a). Furthermore, the Williamson–Hall plot based on the Scherrer equation was also utilized to determine the average size of VO2 NPs,33 which was estimated to be 45.1 nm (Fig. S3, ESI). This size was comparable to the average value of VO2 NPs determined by TEM (∼38 nm). The detailed examination of the individual NPs using high-resolution TEM showed obvious lattice fringe spacings of 0.480, 0.458, and 0.327 nm, which match well with the (100), (010), and (110) interplanar distances of monoclinic VO2, respectively (Fig. 3c). In the selected area electron diffraction (SAED) pattern (inset of Fig. 3c) of individual particles, the independent and bright diffraction spots, which are indexed the (100), (010), and (110) planes of VO2 (M), indicate that the individual NPs have good crystallinity. These structural characteristics of our particles synthesized by this simple top-down method demonstrate that we achieved small, uniformly sized high-quality crystalline VO2 NPs, which have quality equivalent to that of VO2 (M) particles synthesized by complicated bottom-up processes, such as the hydrothermal method.25

image file: c7tc05876d-f3.tif
Fig. 3 Size and local crystal structure of VO2 (M) nanoparticles. (a) SEM image and (b) TEM image of synthesized VO2 (M) nanoparticles (NPs) (inset: histogram of size distribution of the NPs). (c) HRTEM of individual VO2 (M) NPs (inset: SAED image of synthesized VO2 (M)).

The facile transformation of amorphous V2O5 NPs to crystalline VO2 (M) NPs was characterized in situ by real-time X-ray diffraction techniques. The results were obtained by exposing synchrotron radiation (λ = 0.15401 nm, 3D beamline at Pohang Light Source II) on the sample holder at 500 °C under 1 atm of N2 (Fig. 4a). Heating of the amorphous V2O5 particles to 500 °C at 50 °C min−1 in an N2 atmosphere stimulated the formation of the new peaks of rutile VO2 (R); the rapid reactivity of the amorphous V2O5 NP phase demonstrates that the amorphous V2O5 transformed instantly to infrared-reflective VO2 (R) NPs as a high-temperature phase by the loss of oxygen at 500 °C. This infrared-reflective VO2 (R) phase could be reversibly converted to the infrared-transparent VO2 (M) phase by simply decreasing the temperature to room temperature; this change confirmed the temperature-induced metal–insulator (MI) transition of our synthesized VO2 NPs, which is a crucial requirement for the infrared-transmittance-tunable thermochromic functionality.

image file: c7tc05876d-f4.tif
Fig. 4 Synchrotron in situ XRD and DSC characterization. (a) In situ monitoring of the phase transformation from amorphous V2O5 nanoparticles (NPs) to crystalline VO2 (M) NPs by using synchrotron radiation. When amorphous V2O5 particles were heated to 500 °C in an N2 atmosphere, new peaks of rutile VO2 (R) appear instantly, which indicate the rapid reactivity of the amorphous phase for the phase transformation to the VO2 phase. (b) Phase diagram of vanadium oxide between VO2 and V2O5.28,41 The direct reduction of bulk V2O5 powders to the VO2 phase thermodynamically requires high-temperature reactions (>1200 °C) and long reaction times (several days). (c) Kinetically accelerated phase transformation from amorphous V2O5 NPs to VO2 NPs. ΔGa1, ΔGa2, and ΔGa3 represent activation energies of phase transformation from amorphous V2O5 nanoparticles, V2O5 nanoparticles and bulk V2O5, respectively, to VO2 nanoparticles. Latent heat measurement of synthesized VO2 NPs with metal–insulator transitions (d) during heating (e) and during cooling (c).

The temperature-induced reversible MI transition of the synthesized VO2 NPs between VO2 (M) and VO2 (R) was further revealed by differential scanning calorimetry (DSC) curves (Fig. 4d and e). The MI transition as a first-order phase transition is inevitably accompanied by the release of latent heat, so an endothermic peak occurs in DSC curves during the temperature-induced insulator-to-metal (I-to-M) transition and an exothermic peak occurs during the metal-to-insulator (M-to-I) transition.16,24,34 Because the position of the peaks corresponds to the transition temperature and their areas measure the latent heat of the transition, DSC curves can provide the quality of synthesized VO2 NPs that undergo the MI transition, and thus indirectly predict the thermochromic properties of coatings with our NPs.

In our VO2 NPs, endothermic (and exothermic) peaks appeared at TIM ∼ 87 °C during heating and TIM ∼ 59 °C during cooling. The first of these transitions corresponds to the I-to-M transition temperature of VO2 from the monoclinic phase to the high-temperature rutile phase; the second corresponds to the M-to-I transition. Interestingly, the I-to-M transition temperature of our VO2 NPs is 13 °C higher than that of bulk VO2 (∼74 °C); this difference widens the hysteresis window of the MI transition temperature (ΔT ∼ 28 °C). This ΔT increase can be observed in VO2 NPs embedded in the matrix, and can be ascribed to an increase in the energy barrier for nucleation of transformed phases, because of a decreased probability of finding potential defects on which to form nuclei in the VO2 NPs.16,35–38 The latent heats of phase transition were estimated from the area of the peaks to be 27 J g−1 for the I-to-M transition and 22 J g−1 for the M-to-I transition.

Our approach to synthesize VO2 NPs achieves good crystal quality and is compatible with mass production, and therefore has advantages for the development of thermochromic coatings. The decreased size of the particles after milling extremely decreases the transformation time (<10 min) compared to the previously reported batch hydrothermal technique16,20,21 (12–48 h), probably due to significantly accelerated kinetics of mass transport for phase transformation, i.e., oxygen out-diffusion. Furthermore, much lower temperatures and reducing pressures were required for our V2O5 to VO2 transformation with small particle sizes than were required for the previously reported transformation (Fig. 4b);26–28 these changes could be attributed to the amorphous nature of the as-milled V2O5 NPs: non-bonding states increase in amorphous states, and therefore increase the enthalpy of the as-milled amorphous V2O5 over that of crystalline V2O5. Both the amorphous nature and size reduction of the V2O5 reactants would decrease the activation energy for the V2O5 → VO2 transformation, and as a result facilitate the transformation to new ground states of VO2 by a fast, one-step reduction process (Fig. 4c).39

Thus, our top-down approach to synthesize VO2 NPs from micro-sized V2O5 particles is advantageous for the commercialization of smart windows that use a thermochromic coating based on VO2 NPs. By using our approach, we succeeded in producing 250 g of thermochromic VO2 NPs from V2O5 particles at a time; this yield cannot be achieved using other bottom-up synthesis approaches.23,25,34 The single reaction or one step reaction includes short reaction times (for N2 gas, <10 min), low reaction temperatures, use of a harmless low-cost vanadium source (V2O5), and no additives (i.e., polymers; acid or alkali to adjust the pH).

Using our mass-producible VO2 NPs, we demonstrated large-area thermochromic coatings with VO2/polyvinylpyrrolidone (PVP) composites (i.e., VO2 NPs embedded in the PVP matrix) based on the R2R tape-cast method. To achieve the uniform thermochromic coatings with excellent infrared-blocking ability, the VO2 NPs must be uniformly distributed in the castable slurry. After mixing PVP and VO2 NPs in ethanol as a solvent, the solvent was partially evaporated while stirring at 70 °C; the remaining castable slurry retains excellent dispersion and stability (Fig. 1b), and can therefore be used in a tape-casting process. The combination of this R2R process with a large amount of thermochromic NPs yielded homogeneous VO2/PVP composite coatings (12 cm long × 600 cm wide) on a flexible Si-coated PET substrate (Fig. 5a and b). Additional supply of VO2 NPs further enlarged the size of the coating to more than 12 cm × 600 cm by the continuous R2R process. This method can be easily scaled to increase productivity, and may enable continuous production of large-area coatings (Fig. 5e) that fulfill the requirements of practical VO2-based thermochromic technology.

image file: c7tc05876d-f5.tif
Fig. 5 Thermochromic characterization of large-area VO2/PVP composite coatings. (a) Photograph of homogeneous VO2/PVP composite coatings prepared by tape casting with a size of 12 cm × 600 cm on the flexible Si-coated PET substrate. (b) Visible-transmittance-tuned thermochromic coatings (TS1 > TS2 > TS3). (c) UV/Vis/NIR transmittance measurements of S1 (black line), S2 (red line) and S3 (blue line) during heating (solid line) and cooling (dotted line). (d) Temperature-dependent luminous transmittance of S3 at a fixed wavelength of 1766.5 nm. (e) Comparison of our work with previous approaches in terms of NP powder production mass, coating area production, and infrared-blocking ability.

This large-area thermochromic coating showed excellent infrared modulation across the MI transition temperature of VO2 NPs. The visible transmittances of thermochromic coatings were tuned by adjusting either the ratio of the PVP matrix to VO2 NPs, or the thickness of composites; as a result the samples with three different visible transmittances (TS1 > TS2 > TS3) were produced (Fig. 5b). All films had excellent flexibility, uniformity and acceptable visible transmittance. However, as the concentration of VO2 NPs or thickness of composites increased, the infrared-modulation ability increased at the expense of a decrease in visible transmittance (Fig. 5c). In all samples, the thermally-induced MI transition caused no change in the visible transmittance (380 nm ∼ 780 nm), but did cause modulation in the infrared transmittance of 42.5%, 45% and 42.8% at 1500 nm, 1766.5 nm and 2000 nm, respectively, in S3 (Table 1), which is comparable to the best infrared-modulation ability reported previously in thermochromic coatings that use VO2 NPs (Fig. 5e).13,31,34,40

Table 1 Integral luminous transmittance Tlum and solar energy modulation ability ΔTsol of the thermochromic coatings achieved using R2R processing
Sample T lum (%) T sol (%) ΔTsol (%)
30 °C 100 °C 30 °C 100 °C
Sample 1 68.7 69.1 72.2 69.9 2.3
Sample 2 41.4 44.5 49.9 42.8 7.1
Sample 3 22.4 22.9 33.8 21.6 12.2

A hysteresis loop of the infrared transmittance (at 1776.5 nm wavelength) of S3 as a function of temperature appears with a modulation of up to ∼45% during the MI transition heating–cooling cycles (Fig. 5d). For S3, the transmittance at 1776.5 nm gradually decreased from 66.51% at 27 °C to 21.26% at 110 °C in a heating cycle; this change corresponds to a transition from the VO2 (M) to VO2 (R) phase. During a cooling cycle, the infrared transmittance of the coating recovered to 66% at 30 °C; this change demonstrates the reversible modulation of the infrared transmittance by the MI transition.


We reported a novel process to synthesize VO2 NPs and used them in thermochromic composite coatings on a flexible substrate. A relatively large quantity of VO2 NPs (∼250 g) could be produced by delicately controlled thermal treatment of bead-milled amorphous V2O5 NPs. The amorphous nature of V2O5 reactants combined with the reduced size decreased the activation energy for the phase transformation from V2O5 to VO2, and decreased the out-diffusion length for oxygen atoms; as a result, the kinetics of this reaction were accelerated and enabled the synthesis of high-quality VO2 NPs by a fast, one-step heat treatment. This mass production of VO2 NPs allowed fabrication of the largest thermochromic coatings (12 cm × 600 cm) yet reported in the literature. Using the well-developed R2R tape-cast method, these coatings showed excellent infrared modulation ability (∼45%) and are suitable for commercial applications. This simple and efficient strategy to synthesize oxide NPs and fabricate thermochromic films will accelerate the commercialization of VO2-based thermochromic coatings for smart windows, which will contribute to a significant reduction in energy consumption by air conditioning or heating in buildings.


Synthesis of mass-producible VO2 nanoparticles

Bead-milling was applied to decrease the size of micron-sized V2O5 particles. Commercial V2O5 powder (99.6%, Sigma Aldrich, 250 g) was dispersed in pure ethanol (3 L) with a dispersing agent, polyvinylpyrrolidone (PVP, Sigma Aldrich, 5 g). The obtained dispersion was bead-milled using 0.2 mm-sized zirconia beads at 3000 rpm for 18 h (Nano Particle Mill, Nanointech), and then allowed to settle in an ambient environment for several hours. The bead-milled amorphous V2O5 NPs were gathered by centrifugation, and then air-dried at 70 °C for several hours to remove the residual solvent.

From these amorphous V2O5 NPs, high-quality VO2 nanoparticles were synthesized by thermal treatment at 500 °C. The amorphous V2O5 nanoparticles were placed in a quartz crucible, and then transferred into a tube furnace. The furnace was heated at a constant rate to 500 °C. The reaction was conducted either for <10 min under N2 gas, or for 1 h under 130 Torr of air. After the reaction, the furnace was allowed to cool passively to room temperature.

VO2/PVP composite coatings prepared by the roll-to-roll tape-cast process

The VO2 (M)/PVP composite coatings were prepared by tape-casting (Fig. 1c). First, the synthesized VO2 (M) NPs (∼0.3 g), PVP (∼3 g) and pure ethanol were mixed by vigorous stirring at 70 °C; this process leads to partial evaporation of the solvent and yields a viscous castable slurry with excellent dispersion and stability (Fig. 1b). A limited amount of slurry was cast on a slowly moving substrate composed of Si-coated polyethylene terephthalate (PET), and then heated at 70 °C during the tape-casting process. The thickness of the coating was controlled by adjusting the distance between a stationary doctor blade and the substrate in continuous tape-cast equipment. The casting was conducted at 250 rpm. The visible transmittances of these thermochromic composite coatings were tuned by either adjusting the ratio of the PVP matrix to VO2 NPs (1[thin space (1/6-em)]:[thin space (1/6-em)]10 or 1[thin space (1/6-em)]:[thin space (1/6-em)]20) or by adjusting the thickness of the composites.

Structural and optical characterization

Crystal structures of the particles were determined using an X-ray diffractometer (XRD, Model D/Max 2500, Rigaku, Japan) with Cu Kα radiation (λ = 1.5418 Å) at a voltage of 40 kV and a current of 200 mA. The size and local crystal structure of the particles were analyzed using a field-emission scanning electron microscope (FE-SEM, Philips electron optics B.V., XL30S FEG, Netherlands) and a high-resolution transmission electron microscope (HR-TEM, JEOL, JOEL JEM-2200FS (with image Cs-corrector), Tokyo, Japan). The latent heat of the synthesized VO2 particles during the phase transition was measured using a differential scanning calorimeter (DTA/DSC, SETARAM) at heating and cooling rates of 10 °C min−1 with temperatures from 20 to 110 °C. Transmission measurements were conducted at wavelengths ranging from 280 to 2500 nm at 30 and 100 °C using a UV-vis-near-IR spectrophotometer (Perkin Elmer Lambda 750S) by inserting the thermochromic coatings into a temperature-controlled holder. For all samples, the integrated luminous transmittance (Tlum, 380–780 nm) and solar transmittance (Tsol, 280–2800 nm) were obtained as
image file: c7tc05876d-t1.tif

ΔTsol = Tsol(T < Tc) − Tsol(T > Tc)
where T(λ) is the transmittance at wavelength λ, i denotes ‘lum’ or ‘sol’ for the calculation, T is the temperature, φlum is the standard luminous efficiency function for the photopic vision, φsol is the solar irradiance spectrum for an air mass of 1.5 (corresponding to the Sun at 37° above the horizon), and ΔTsol is the solar energy modulation ability.

Conflicts of interest

There are no conflicts of interest to declare.


We thank H. Park for TEM. We acknowledge support for this work by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2017R1A2B2007819) and by the Ministry of Trade, Industry and Energy (10076608). In addition, this study was partially supported by the Brain Korea 21 PLUS project (Center for Creative Industrial Materials).


  1. Y. Wang, E. L. Runnerstrom and D. J. Milliron, Annu. Rev. Chem. Biomol. Eng., 2016, 7, 283–304 CrossRef CAS PubMed.
  2. C. G. Granqvist, Thin Solid Films, 2014, 564, 1–38 CrossRef CAS.
  3. I. P. Parkin and T. D. Manning, J. Chem. Educ., 2006, 83, 393 CrossRef CAS.
  4. G. B. Smith and C.-G. S. Granqvist, Green nanotechnology: solutions for sustainability and energy in the built environment, CRC Press, 2010 Search PubMed.
  5. F. Morin, Phys. Rev. Lett., 1959, 3, 34 CrossRef CAS.
  6. D. Adler, Rev. Mod. Phys., 1968, 40, 714 CrossRef CAS.
  7. M.-H. Lee, Sol. Energy Mater. Sol. Cells, 2002, 71, 537–540 CrossRef CAS.
  8. S. Wang, M. Liu, L. Kong, Y. Long, X. Jiang and A. Yu, Prog. Mater. Sci., 2016, 81, 1–54 CrossRef CAS.
  9. M. E. A. Warwick and R. Binions, J. Mater. Chem. A, 2014, 2, 3275–3292 CAS.
  10. J. Narayan and V. Bhosle, J. Appl. Phys., 2006, 100, 103524 CrossRef.
  11. R. Binions, G. Hyett, C. Piccirillo and I. P. Parkin, J. Mater. Chem., 2007, 17, 4652–4660 RSC.
  12. S. Chen, H. Ma, J. Dai and X. Yi, Appl. Phys. Lett., 2007, 90, 101117 CrossRef.
  13. Y. Gao, C. Cao, L. Dai, H. Luo, M. Kanehira, Y. Ding and Z. L. Wang, Energy Environ. Sci., 2012, 5, 8708–8715 CAS.
  14. Y. F. Gao, S. B. Wang, H. J. Luo, L. Dai, C. X. Cao, Y. L. Liu, Z. Chen and M. Kanehira, Energy Environ. Sci., 2012, 5, 9947 Search PubMed.
  15. Y. F. Gao, S. B. Wang, L. T. Kang, Z. Chen, J. Du, X. L. Liu, H. J. Luo and M. Kanehira, Energy Environ. Sci., 2012, 5, 8234–8237 CAS.
  16. Z. Chen, Y. Gao, L. Kang, C. Cao, S. Chen and H. Luo, J. Mater. Chem. A, 2014, 2, 2718–2727 CAS.
  17. S.-Y. Li, G. A. Niklasson and C.-G. Granqvist, J. Appl. Phys., 2010, 108, 063525 CrossRef.
  18. A. B. Huang, Y. J. Zhou, Y. M. Li, S. D. Ji, H. J. Luo and P. Jin, J. Mater. Chem. A, 2013, 1, 12545–12552 CAS.
  19. Y. F. Gao, C. X. Cao, L. Dai, H. J. Luo, M. Kanehira, Y. Ding and Z. L. Wang, Energy Environ. Sci., 2012, 5, 8708–8715 CAS.
  20. Y. Zhang, X. Zhang, Y. Huang, C. Huang, F. Niu, C. Meng and X. Tan, Solid State Commun., 2014, 180, 24–27 CrossRef CAS.
  21. N. Shen, S. Chen, Z. Chen, X. Liu, C. Cao, B. Dong, H. Luo, J. Liu and Y. Gao, J. Mater. Chem. A, 2014, 2, 15087–15093 CAS.
  22. N. Shen, B. Dong, C. Cao, Z. Chen, H. Luo and Y. Gao, RSC Adv., 2015, 5, 108015–108022 RSC.
  23. Z. Cao, X. Xiao, X. Lu, Y. Zhan, H. Cheng and G. Xu, Sci. Rep., 2016, 6, 39154 CrossRef CAS PubMed.
  24. J. Zou, Y. Peng and H. Lin, J. Mater. Chem. A, 2013, 1, 4250–4254 CAS.
  25. 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, 4520–4523 CAS.
  26. N. Kimizuka, M. Ishii, I. Kawada, M. Saeki and M. Nakahira, J. Solid State Chem., 1974, 9, 69–77 CrossRef CAS.
  27. J. MacChesney and H. Guggenheim, J. Phys. Chem. Solids, 1969, 30, 225–234 CrossRef CAS.
  28. C. Griffiths and H. Eastwood, J. Appl. Phys., 1974, 45, 2201–2206 CrossRef CAS.
  29. L. Austin, K. Shoji and P. T. Luckie, Powder Technol., 1976, 14, 71–79 CrossRef.
  30. A. Makarevich, I. Sadykov, D. Sharovarov, V. Amelichev, A. Adamenkov, D. Tsymbarenko, A. Plokhih, M. Esaulkov, P. Solyankin and A. Kaul, J. Mater. Chem. C, 2015, 3, 9197–9205 RSC.
  31. H. Kim, Y. Kim, K. S. Kim, H. Y. Jeong, A.-R. Jang, S. H. Han, D. H. Yoon, K. S. Suh, H. S. Shin and T. Kim, ACS Nano, 2013, 7, 5769–5776 CrossRef CAS PubMed.
  32. G. Silversmit, D. Depla, H. Poelman, G. B. Marin and R. De Gryse, J. Electron Spectrosc. Relat. Phenom., 2004, 135, 167–175 CrossRef CAS.
  33. G. K. Williamson and W. H. Hall, Acta Metall., 1953, 1, 22–31 CrossRef CAS.
  34. M. Powell, P. Marchand, C. Denis, J. Bear, J. Darr and I. Parkin, Nanoscale, 2015, 7, 18686–18693 RSC.
  35. R. Lopez, T. Haynes, L. Boatner, L. Feldman and R. Haglund Jr, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 65, 224113 CrossRef.
  36. A. Tselev, E. Strelcov, I. A. Luk’yanchuk, J. D. Budai, J. Z. Tischler, I. N. Ivanov, K. Jones, R. Proksch, S. V. Kalinin and A. Kolmakov, Nano Lett., 2010, 10, 2003–2011 CrossRef CAS PubMed.
  37. W. Fan, J. Cao, J. Seidel, Y. Gu, J. Yim, C. Barrett, K. Yu, J. Ji, R. Ramesh and L. Chen, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 83, 235102 CrossRef.
  38. Y. Zhou, A. Huang, Y. Li, S. Ji, Y. Gao and P. Jin, Nanoscale, 2013, 5, 9208–9213 RSC.
  39. I. Avramov, T. Vassilev and I. Penkov, J. Non-Cryst. Solids, 2005, 351, 472–476 CrossRef CAS.
  40. X. Cao, M. N. Thet, Y. Zhang, S. C. J. Loo, S. Magdassi, Q. Yan and Y. Long, RSC Adv., 2015, 5, 25669–25675 RSC.
  41. J. Nag and R. Haglund Jr, J. Phys.: Condens. Matter, 2008, 20, 264016 CrossRef.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c7tc05876d

This journal is © The Royal Society of Chemistry 2018