Dataset and publication "Strongly Thermochromic W-Doped Vo2 Films with a Large Temperature Coefficient of Electrical Resistance Near Room Temperature"

Full Length Article Strongly thermochromic W-doped VO 2 films with a large temperature coefficient of electrical resistance near room temperature Sadoon Farrukh , Jaroslav Vlˇ cek * , Jiˇ rí Rezek , Jiˇ rí Houˇ ska , Radomír ˇ Cerstvý , Tom´ aˇ s Koz´ ak Department of Physics and NTIS – European Centre of Excellence, University of West Bohemia, Univerzitní 8, 30100 Plzeˇ n, Czech Republic ARTICLE INFO Keywords: W-doped VO 2 Large temperature coefficient of resistance Infrared detectors Decreased transition temperature Low deposition temperature ABSTRACT We report the crystal structure, surface morphology, electronic band structure, optical and electrical properties, and semiconductor-metal transition characteristics of strongly thermochromic W-doped VO 2 films with a large (up to -16 % K −1 ) temperature coefficient of electrical resistance at a small hysteresis width of electrical resistivity (down to 3 ◦C) near room temperature, and with a wide temperature operation range at a high detection sensitivity (≥8 % K −1 ) and low values of the electrical resistivity. They were deposited at a reduced substrate temperature of 350 ◦C onto soda-lime glass (SLG) with two versions of yttria-stabilized zirconia (YSZ) interlayers possessing different cubic crystal orientations, and onto bare SLG and monocrystalline YSZ and Al 2 O 3 substrates for comparison. The W-doped VO 2 depositions were performed using reactive deep oscillation magnetron sputtering with a feedback pulsed O 2 flow control allowing us to increase deposition rate of films up to 20–30 nm min −1 for a target-substrate distance of 100 mm. The results are important for a further improvement of thermochromic performance of VO 2 -based coatings for energy-saving smart windows and for a new design of highperformance infrared detectors and temperature sensors prepared by a fast low-temperature scalable synthesis. 1. Introduction Vanadium dioxide (VO 2 ) is a strongly correlated material that exhibits a reversible phase transition from a low-temperature monoclinic VO 2 (M1) semiconducting phase to a high-temperature tetragonal VO 2 (R) metallic phase at a transition temperature, T tr , of approximately 68 ◦C for the bulk material [1]. This transformation is accompanied by dramatic changes in the electrical, optical, thermal and magnetic properties and can be initiated not only by temperature change but also by several other external stimuli such as light, terahertz pulses, electric and magnetic fields, and mechanical stress and strain. These characteristics make VO 2 a promising candidate for a wide variety of potential technological applications (see reviews [2–11] and the works cited therein). One of the very important potential applications based on a thermally stimulated semiconductor-metal transition of VO 2 is thermochromic VO 2 -based coatings for energy-saving smart windows with automatically varied solar energy transmittance [7–9]. To meet the requirements for large-scale implementation on building glass (glass panes or flexible glass and polymer foils laminated to glass panes), VO 2 -based coatings should satisfy the following strict criteria simultaneously: a maximum substrate temperature, T s , during the preparation (deposition and possible postannealing) close to 300 ◦C or lower, T tr close to 20 ◦C, an integral luminous transmittance T lum >60 %, a modulation of the solar energy transmittance ΔT sol >10 %, long-term environmental stability, and a more appealing color than the usual yellowish or brownish colors in transmission [8]. Here, a major challenge is to achieve the high T lum and ΔT sol at a relatively low T tr and T s . Another area of very important potential applications based on a thermally stimulated semiconductor-metal transition of VO 2 are highperformance infrared detectors (for example, for novel uncooled microbolometers [10–12]), and high-performance temperature and fluid-flow rate sensors (for example, for novel human health and environmental monitoring systems [10]). Here, the main requirements for VO 2 -based films are: a large temperature coefficient of electrical resistance, TCR, for high detection sensitivity, a low electrical resistivity, ρ , to minimize thermal noise and Joule heating, a mitigated or even eliminated thermal hysteresis of ρ for measurement reproducibility, and a reduced T tr [10,12]. The application potential of VO 2 -based films depends on the ability * Corresponding author. E-mail address: [email protected] (J. Vlˇ cek). Contents lists available at ScienceDirect Applied Surface Science Advances journal homepage: www.sciencedirect.com/journal/applied-surface-science-advances https://doi.org/10.1016/j.apsadv.2025.100870 Received 14 July 2025; Received in revised form 2 September 2025; Accepted 2 October 2025 Applied Surface Science Advances 30 (2025) 100870 Available online 8 October 2025 2666-5239/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). to achieve not only the VO 2 stoichiometry but also the crystallization of the VO 2 (M1/R) phase under as industry-friendly process conditions as possible, i.e., at T s close to 300 ◦C or lower (usually used temperatures are higher than 450 ◦C), which is needed for many applications with temperature-sensitive substrates, and without any substrate bias voltage in case of usually used magnetron sputter techniques. Moreover, T tr needs to be decreased to room temperature for many applications. The state-of-the-art way how to achieve that (see [8] for further details and alternative options) is doping by W: the extra valence electron and the structural disorder resulting from the large size of W destabilize the low-temperature semiconducting phase VO 2 (M1) and lower the temperature at which the metallic phase VO 2 (R) forms. Besides the optical transmittance and the electrical resistivity below T tr , the characteristics of the semiconductor-metal transition, such as the corresponding phase-transition amplitude, hysteresis width and phase-transition sharpness, are of key importance. In our recent paper [13], we presented a sputter deposition technique for the preparation of strongly thermochromic YSZ/W-doped VO 2 /YSZ coatings, where YSZ denotes the yttria-stabilized zirconia, with a reduced transition temperature T tr =33–35 ◦C on conventional soda-lime glass (SLG) at T s =350 ◦C and without any substrate bias voltage. The thermochromic W-doped VO 2 layers were deposited using a controlled deep oscillation magnetron sputtering (DOMS) of a single V-W target. The DOMS is a modified version of high-power impulse magnetron sputtering with packages (macropulses) of short high-power micropulses making it possible to increase discharge stability and deposition rate of films. Here, it should be mentioned that reactive high-power impulse magnetron sputtering is a promising scalable deposition technique for a low-temperature (300–350 ◦C) preparation of thermochromic VO 2 -based films [14]. Moreover, a significantly higher environmental stability of the VO 2 films prepared using high-power impulse magnetron sputtering than using a conventional radio frequency magnetron sputtering was demonstrated recently [15]. Our motivation to apply the YSZ (tetragonal Y 0.06 Zr 0.94 O 1.97 phase in [13]) antireflection layers instead of the ZrO 2 antireflection layers [16,17] was to further improve the crystallinity of the thermochromic VO 2 phase in the W-doped VO 2 layers, as the bottom antireflection layer provides also a structure template for their growth. In addition to the VO 2 stoichiometry, the microstructure (particularly the degree of crystallinity, the size and orientation of the W-doped VO 2 crystal grains, and the size of amorphous boundary regions) of the W-doped VO 2 layer is very important for the thermochromic performance (including ΔT sol ) of coatings. Indeed, the high-resolution TEM analyses, presented in [13] and [17], proved a more compact structure of VO 2 crystal grains with much narrower amorphous nanodomains formed along the VO 2 grain boundaries in the W-doped VO 2 layers prepared on the YSZ layers [13]. In this paper, we report on the crystal structure, surface morphology, electronic band structure, optical and electrical properties, and semiconductor-metal transition characteristics of strongly thermochromic W-doped VO 2 films with a large (up to −16 % K −1 ) temperature coefficient of electrical resistance at a small hysteresis width of ρ (down to 3 ◦C) near room temperature, and with a wide temperature operation range at a high detection sensitivity and low values of ρ . They were deposited at T s =350 ◦C onto SLG with two versions of cubic YSZ interlayers possessing different crystal orientations, and onto bare SLG and monocrystalline YSZ and Al 2 O 3 substrates for comparison. The Wdoped VO 2 films were synthesized using reactive DOMS with a feedback pulsed O 2 flow control allowing us to increase deposition rate of films up to 20–30 nm min −1 . No top antireflection overlayers, applied in our earlier works focused exclusively on the optical performance [13,16, 17], were used in this study. The main aim of this work is to present and explain the effect of the YSZ interlayers with preferred (111) or (200) cubic crystal orientation, which were prepared on SLG under different discharge conditions, on the crystal structure, optical and electrical properties, and detection sensitivity of the produced W-doped VO 2 films. This is important not only for a further improvement of thermochromic performance of VO 2 - based coatings, but also for a new design of high-performance infrared detectors and temperature sensors prepared by a fast lowtemperature scalable synthesis. 2. Experimental details 2.1. Film preparation The W-doped VO 2 films were deposited under the same conditions onto 5 different, highly transparent substrates, characterized in Table 1, and onto a Si (100) substrate using a strongly unbalanced magnetron source with a directly water-cooled single V-W (3.0 wt. % corresponding to 0.85 at. %) target (99.95 % V and W purity, diameter of 100 mm and thickness of 6 mm) in a standard stainless-steel vacuum chamber (diameter of 507 mm and length of 520 mm), which was evacuated by a diffusion pump (2 m 3 s −1 ) backed up with a rotary pump (30 m 3 h −1 ). The base pressure before deposition was 10 −3 Pa. The substrates at the distance of 100 mm from the target were at a floating potential. The substrate surface temperature, maintained during the deposition by a builtin heating system, was 350 ◦C. The magnetron was driven by a DOMS power supply (HIPIMS Cyprium plasma generator, Zpulser Inc.). In this work, the macropulse (composed of 10 micropulses) duration was 500 µs at a repetition frequency of 640 Hz (see Fig. 1). The deposition-averaged target power density (spatially averaged over the total target area) was 34 Wcm −2 . Oxygen was admitted into the vacuum chamber via mass flow controller and two corundum conduits. Two O 2 inlets with a diameter of 1 mm were placed symmetrically above the V-W target racetrack at the same distance of 20 mm from the V-W target surface and oriented to the substrate [18]. The to-substrate O 2 injection into the dense plasma in front of the sputtered target is very suitable for reactive HiPIMS depositions of oxide films. It leads to a substantially (2–3 times in [18]) increased local oxygen partial pressure in front of the O 2 inlets, compared with the pO2 in the vacuum chamber, at a very high degree of dissociation of O 2 molecules in the high-density plasma in front of the target. As a result, the pO2 needed for preparation of stoichiometric VO 2 films can be very low, resulting in the required low compound (oxide) fraction in the target surface layer. This is important for process stability, for increased sputtering of V and W atoms, leading to a high deposition rate of W-doped VO 2 films, and for low production of high-energy O − ions. Table 1 Basic characteristics of five substrates used for synthesis of the W-doped VO 2 films. In case of the YSZ layers on SLG, the refractive index at a wavelength of 550 nm, n 550 , and the total deposition-averaged power delivered into the system during preparation of these layers with the extinction coefficient k 550 <10 −4 are also given. Substrate Basic characteristics SLG 1 mm thick conventional soda-lime glass YSZ 3000 W on SLG 170 nm thick YSZ layer with n 550 =2.19, discharge power of 3000 W YSZ 100 W on SLG 215 nm thick YSZ layer with n 550 =1.81, discharge power of 100 W Mono YSZ 0.5 mm thick YSZ cubic (100) monocrystal Mono Al 2 O 3 0.5 mm thick Al 2 O 3 hexagonal (0001) monocrystal S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 2 The argon flow rate was 25 sccm corresponding to p Ar =0.5 Pa, while the total oxygen flow rate, ΦO2, in both conduits was not fixed but alternating between 0 and 14 sccm (see Fig. 1). The moments of switching of the ΦO2 pulses were determined during the deposition by a programmable logic controller using a pre-selected critical value of the oxygen partial pressure (pO2)cr =85 mPa : when pO2(t) <(pO2)cr,ΦO2= 14 sccm and when pO2(t) ≥(pO2)cr,ΦO2 =0 sccm. The values of p Ar and of p Ar +pO2were measured at the chamber wall using a high stability capacitance manometer (Baratron, Type 127, MKS) with the accuracy much better than 1 %. As can be seen in Fig. 1, the waveforms of the magnetron voltage U d (t) and the target current density J t (t), averaged over the total target area, oscillate during a deposition in a dependence on the value of pO2, being in the range from 39 mPa to 102 mPa. The corresponding maximum target power densities in a micropulse were 1.38 kWcm −2 and 1.59 kWcm −2 , respectively. A detailed characterization of the chemical and physical processes on the sputter V-W target, in the discharge plasma and on the surface of growing W-doped VO x films produced under similar conditions using this fast low-temperature controlled synthesis is given in [19]. Two versions of YSZ interlayers with different preferred cubic (111) or (200) crystal orientations, given in terms of planes parallel to the sample surface, were deposited onto SLG by reactive mid-frequency ac magnetron sputtering without any substrate bias voltage at a substrate surface temperature T s <300◦C in an argon-oxygen gas mixture (oxide mode) at p Ar =1 Pa. The depositions were performed using two strongly unbalanced magnetrons with Zr-Y (15.0 wt.% corresponding to 15.3 at.%) targets (99.5% Zr and Y purity, diameter of 100 mm and thickness of 6 mm) driven by a mid-frequency ac power supply (TruPlasma MF 3010, TRUMPF Huettinger). The oscillation frequency was close to 85 kHz at the total deposition-averaged power delivered into the system of 3000 W for pO2=0.6 Pa [preferred YSZ (111) orientation], and 100 W for pO2=0.1 Pa [preferred YSZ (200) orientation], see below. The deposition-averaged target power density for each magnetron was 19.1 Wcm −2 in the first case, while it was 0.64 Wcm −2 in the second one. Note that the operations of both magnetrons in the dual system used are the same but shifted by a half of the power supply period. When one magnetron target is sputtered by positive ions in the negative-voltage half of the period, the other one is discharged by electrons and can be bombarded also by negative ions in the positive-voltage half of the period [20]. Let us recall that the positive potential (20 V for the discharge power of 3000 W and close to 5 V for the discharge power of 100 W, not shown here) switching between the magnetron targets results in an increase of the energies of the ions bombarding the YSZ layers growing on a floating SLG substrate [20,21]. Much higher fluxes, and the increased energy and momentum of the ions bombarding the YSZ layers at the discharge power of 3000 W resulted not only in the aforementioned change of their preferred orientation, but also in a decrease [20] of their root-mean-square surface roughness, R rms , from 5.8 nm to 1.9 nm, and in an increase of their refractive index, n 550 , from 1.81 to 2.19 (see Table 1). Note that the key characteristic of YSZ for optical performance is the optical path length equal to the product of the thickness and n 550 , not these two quantities separately. We therefore control the optical path length to be 170×2.19 nm (3000 W) or 215×1.81 nm (100 W). Both these lengths are within 5% from that of the optimum three-quarter wavelength layer, allowing simultaneous optimization of integral luminous transmittance and modulation of solar energy transmittance (defined next) and reported previously as 180×2.15 nm [22]. 2.2. Film characterization The elemental composition of the V 33.1 W 0.6 O 66.3 films, i.e., the Wdoped VO 2 films with 1.8 at. % of W in the metal sublattice of V 0.982 W 0.018 O 2 , was measured on dedicated 300 nm thick layer on Si (100) substrate in a scanning electron microscope (SU-70, Hitachi) using wave-dispersive spectroscopy (Magnaray, Thermo Scientific) at a low primary electron energy of 7.5 keV. Standard reference samples of pure V, W and Fe 2 O 3 (Astimex Scientific Ltd.) were utilized. Based on the standards’ purity and our experience, the measurement errors are estimated to be 1.0 at. % for V, 0.04 at. % for W and 2.0 at. % for O. X-ray diffraction (XRD) measurements were carried out at room temperature (25 ◦C) on a PANalytical X’Pert PRO MPD diffractometer working in the Bragg-Brentano geometry using a CuK α (40 kV, 40 mA) radiation, 0.25◦divergence slit, 0.5◦anti-scatter slit, 0.04 rad Soller slits, Ni filter for the CuKβ elimination and an ultrafast semiconductor detector X’Celerator. To avoid a strong reflection from monocrystalline substrates (Table 1), a slightly asymmetrical diffraction geometry with an ω -offset of 1.5◦was also used. Samples were scanned over the 2θrange from 10◦to 65◦. The surface morphology of the films was determined by atomic force microscopy (AFM) using a SmartSPM Microscope (AIST-NT) with a Si tip (nominal radius below 10 nm) in semicontact mode. The R rms values were computed from a randomly selected square area of 2 ×2 µm 2 . The thickness of all individual layers and the optical constants (refractive index, n 550 , and extinction coefficient, k 550 , at a wavelength of 550 nm) of both YSZ layers were measured by spectroscopic ellipsometry using the J.A. Woollam Co. Inc. VASE instrument [13]. No change was observed in the characteristics of the YSZ layers determined before and after deposition of the W-doped VO 2 films. The coating transmittance, T, and reflectance, R, were measured by spectrophotometry using the Agilent CARY 7000 instrument with an inFig. 1. Waveforms of the magnetron voltage, U d , and the target current density, J t , for a 500 μ s macropulse composed of 10 micropulses during depositions of all thermochromic W-doped VO 2 films at a pre-selected critical value of the oxygen partial pressure (pO2)cr =85 mPa. Time evolution of the oxygen partial pressure, pO2, during the deposition is shown in the inset. The pre-selected critical value (pO2)cr =85 mPa determining the switching-on and switchingoff of the oxygen flow rate ΦO2=14 sccm is marked by dots. The dashed (blue) and full (red) lines represent the waveforms (both U d and J t ) measured at the minimum (indicated) and maximum pO2, during the deposition, respectively. S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 3 house made heat/cool cell. The measurements were performed in the wavelength range of 300–2500 nm at the angles of incidence of 0◦(T) and 7◦(R) for T ms = − 20 ◦C and T mm =70 ◦C. Hysteresis curves were measured for T at λ =2500 nm, T 2500 , in the temperature range T m = −20 ◦C to 70 ◦C. The thermochromic performance is quantified by means of integral luminous transmittance, T lum (T m ), and modulation of the solar energy transmittance, ΔT sol , which are defined as Tlum(Tm) = ∫780 380 φlum(λ)φsol(λ)T(Tm,λ)dλ ∫780 380 φlum(λ)φsol(λ)dλ and ΔTsol =Tsol(Tms) − Tsol(Tmm), where Tsol(Tm) = ∫2500 300 φsol(λ)T(Tm,λ)dλ ∫2500 300 φsol(λ)dλ . Here, φ lum is the luminous sensitivity of the human eye and φ sol is the solar irradiance spectrum at an air mass of 1.5 [23]. The average integral luminous transmittance is defined as T lum =[T lum (T ms ) +T lum (T mm )]/2. Note that ΔT sol is a much better measure of the thermochromic performance than e.g. ΔT 2500 (for multilayered coatings the relationship between these two quantities is not even monotonic [8]). The optical band gaps E g1 and E g2 were determined from Tauc plots by utilizing the relation ( α E) 1/2 ∼E - E g , valid for indirect-allowed optical transitions [24–26] where E is the photon energy and α is the absorption coefficient (neglecting the absorption in substrates) calculated as α =-{ln[T/(1-R)]}/d, where d is the thickness of the thermochromic film [24,26]. The temperature-dependent electrical resistivity, ρ (T m ), was measured in the temperature range T m = − 20 ◦C to 80 ◦C with the step of 10 ◦C except near T tr where the step was 5 ◦C in the Van der Pauw configuration using a Variable Temperature Hall Measurement System (MMR Technologies) equipped with a Joule-Thomson refrigerator and a heating stage. Gold contacts (thickness of 50 nm) were sputter-deposited in the corners of the 9 ×9 mm 2 samples to improve the contact between the sample and the gold spring-loaded probe tips. The temperature coefficient of electrical resistance, defined by TCR(T m ) =(1/ ρ )(d ρ /dT m ), was calculated from an equivalent formula TCR(T m ) =d(ln ρ )/dT m in accordance with [12,27,28]. 3. Results and discussion 3.1. Crystal structure and surface morphology The XRD patterns of the W-doped VO 2 films on various substrates are shown in Fig. 2. The crystalline phases identified include VO 2 (M1) (PDF [29] #04–003–2035), VO 2 (R) (#01–073–2362), V 3 O 7 (#01–071–1591), YSZ (cubic Y 0.14 Z 0.86 O 1.93 ; #04–023–7233) and Al 2 O 3 [hexagonal (0001) monocrystal; #00–010–0173]. Regarding the thermochromic W-doped VO 2 films on various substrates, there are not detected any contributions from undesired non-thermochromic phases except for a slight contribution from the V 3 O 7 phase in the W-doped VO 2 film on the monocrystalline Al 2 O 3 substrate. This cannot be considered to be behind the presented differences in thermochromic performance: while specific templates facilitate the nucleation of various phases and (as discussed below) lead to their growth at various strains, the stoichiometry and crystallinity of minor inclusions of any non-thermochromic phase is of secondary importance as long as it is truly minor and sufficiently transparent and insulating. All other peaks are identified as diffraction peaks of either the low-temperature monoclinic VO 2 (M1) phase or the high-temperature tetragonal VO 2 (R) phase. These two thermochromic phases are difficult to distinguish and they are actually expected to be present simultaneously because the measurement temperature T m =25 ◦C is close to the transition temperatures of the investigated films (as shown later). Fig. 2 confirms two different orientations of the YSZ interlayers deposited onto SLG, see the dominant (111) peak for the YSZ 3000 W and the dominant (200) peak for the YSZ 100 W at the same position as for the YSZ monocrystal. A detailed analysis showed that a position with 2θ =29.23◦for the strongest (111) peak of the YSZ interlayer prepared at a high discharge power of 3000 W is considerably shifted from a theoretical position with 2θ =30.10◦, in contrast to the same peak of the YSZ interlayer prepared at the discharge power of 100 W with 2θ =30.07◦ Regarding the effect of the template on the crystal orientation of thermochromic films, first, the XRD patterns do not reveal any convincing preferential orientation of W-doped VO 2 grown on SLG and on YSZ (200) (both the monocrystal and the 100 W interlayer). Second, there is a considerable preferential orientation VO 2 (M1) (011) ⇔ VO 2 (R) (110) of W-doped VO 2 grown on YSZ (111) (the 3000 W interlayer). However, probably owing to stress (the lattice misfit is ≈10 %), this strongest peak on YSZ 3000 W with 2θ =27.70◦is shifted from the corresponding theoretical positions of the VO 2 (M1) (011) planes (2θ = 27.80◦) and the VO 2 (R) (110) planes (2θ =27.91◦). The use of the Bragg-Brentano geometry with an ω -offset of 1.5◦did not significantly change this XRD pattern, indicating that while the distribution of angles between the surface and the aforementioned crystallographic planes is Fig. 2. X-ray diffraction patterns taken at T m =25 ◦C from five W-doped VO 2 films deposited onto various substrates (Table 1). The main diffraction peaks of VO 2 (M1), VO 2 (R), V 3 O 7 , YSZ (cubic Y 0.14 Z 0.86 O 1.93 ) and Al 2 O 3 [hexagonal (0001)] are marked. The red and blue lines represent the patterns obtained using a standard Bragg-Brentano geometry and the Bragg-Brentano geometry with an ω -offset of 1.5◦, respectively. Very low additional peaks marked by asterisks in the diffractogram of the W-doped VO 2 film on the monocrystalline YSZ are the (200) peaks of the YSZ monocrystal from the small W contamination on the Cu anode originating from the heated W cathode in the X-ray tube. S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 4 centered around zero, it has a non-negligible width (similarly to the underlying YSZ). Third, there is a strong preferential orientation VO 2 (M1) (020) and (002) ⇔ VO 2 (R) (200) of W-doped VO 2 grown on the Al 2 O 3 monocrystal (lattice misfit of only ≈4 %). This time the use of the Bragg-Brentano geometry with an ω -offset of 1.5◦led to significantly weaker XRD peaks, indicating truly epitaxial growth [30,31]. The lower background level of the XRD patterns measured in case of the used ω -offset, particularly for both monocrystalline samples, is due to a partial defocusing of the diffracted X-ray beam. In parallel, the growth template affected the deposition rate (thicknesses - including half of the surface roughness layer in the corresponding optical model - between 62 and 86 nm in Table 2, fully acceptable for the present purpose) and especially the surface roughness (R rms between 5.3 and 19.9 nm in Fig. 3) of W-doped VO 2 . The interpretation is not straightforward due to the superposition of numerous phenomena such as roughness of the glass substrate, roughness of interlayers, effect of (local) epitaxy on strain and in turn density, effect of (local) epitaxy on crystal size or even slightly varied O content. Nevertheless, several observations can be made. First, the presented deposition technique of YSZ is able to yield a very low roughness in a wide range of sputtering powers, even suppressing that of the underlying glass. Second, the only truly epitaxial growth on the Al 2 O 3 monocrystal, arguably leading to the largest horizontal crystal size (direction which is most important for the electrical resistivity presented below), also led to - by far - the largest roughness. Third, the only other template allowing at least local epitaxy, YSZ 3000 W, affected the crystal orientation but arguably did not support a large crystal formation (let us recall the aforementioned lattice misfit and stress), preserving a low roughness. Fourth, there is an imperfect but considerable correlation between the thickness and the roughness, indicating that the mass of W-doped VO 2 deposited may be less template-dependent than its volume. 3.2. Electronic band structure Fig. 4(a) shows a schematic energy band diagram for the lowtemperature VO 2 (M1) semiconducting phase with two band gaps. E g2 is a band gap between the filled lower part of the split d || band and the empty π * band while E g1 is a band gap between the filled π band and the empty π * band. The width of the E g2 gap is in the infrared spectral range with reported values of about 0.6 eV [34–36]. At the transition to the high-temperature VO 2 (R) metallic phase, E g2 closes (Fig. 4(b)). This results in an increased density of free charge carriers with a direct effect mainly on the electrical conductivity of the high-temperature state and on the contribution of infrared wavelengths to ΔT sol . The width of the E g1 gap is in the visible range with reported values of 1.5 – 1.7 eV [24,25, 37,38] As can be seen in Fig. 4(c), the incorporation of W into the metal sublattice of VO 2 resulted in decreased values of E g2 <0.25 eV, which are associated predominantly with a decreased T tr due to the W doping [5,10], at E g1 =1.63 – 1.76 eV for the low-temperature state (T ms = − 20 ◦C), being almost the same as E g1 for the high-temperature state (T mm = 70 ◦C), of all W-doped VO 2 films. The α values are in the infrared range considerably higher for the high-temperature state (T mm =70 ◦C) of each film than for the low-temperature state (T ms = − 20 ◦C), leading to a substantially reduced transmittance. This is very important for the thermochromic performance (including ΔT sol ) of films. 3.3. Optical and electrical properties The optical and electrical properties of all five coatings are presented in Fig. 5 and Table 2. Let us start by recalling that all requirements of a specific application have to be fulfilled in parallel, which in the first place leads to the special status of T tr : each presented functional property should be evaluated and compared with the state of the art in parallel to at which T tr it has been achieved. In the present work this Table 2 Optical and electrical properties of five W-doped VO 2 films with a thickness d deposited onto various substrates. Here, T lum is the average integral luminous transmittance, ΔT sol is the modulation of the solar energy transmittance, (T tr ) T and (T tr ) ρ are the transition temperatures corresponding to the middle of hysteresis curves [Figs. 5(b) and 5(c), respectively], both horizontally and vertically, (ΔH) T and (ΔH) ρ are the hysteresis widths, ρ (T ms ) and ρ (T mm ) are the electrical resistivities at T ms = −20 ◦C and T mm =70 ◦C, respectively, and ρ (T tr ) is the electrical resistivity at a transition temperature. Substrate d (nm) T lum ( %) ΔT sol ( %) (T tr ) T ( ◦C) (ΔH) T ( ◦C) ρ (T ms )/ ρ (T mm ) ρ (T tr ) (Ωcm) (T tr ) ρ ( ◦C) (ΔH) ρ ( ◦C) SLG 70 32.1 6.8 33 16 105 0.11 34 4.4 YSZ 3000 W on SLG 62 46.5 7.2 28 15 67 0.39 30 3.9 YSZ 100 W on SLG 73 39.9 8.4 32 11 101 0.07 35 4.3 Mono YSZ 86 26.9 9.2 31 7 91 0.13 36 3.7 Mono Al 2 O 3 86 37.1 5.9 33 7 275 0.02 39 2.5 Fig. 3. Surface morphology and the root-mean-square roughness of the surface for five W-doped VO 2 films deposited onto various substrates. S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 5 means transmittance-based (T tr ) T =28 ◦C −33 ◦C (relevant for energysaving smart windows) and resistivity-based (T tr ) ρ =30 ◦C-39 ◦C (relevant for infrared detectors and temperature sensors; as usual [8], similar but not equal to (T tr ) T ). The spectral transmittance of all five coatings both below and above T tr is shown in Fig. 5(a). From the qualitative point of view, note the features resulting from the presence of the YSZ interlayers in two of the coatings. First, the transmittance maximum in the visible and near infrared, which is to some extent exhibited by all coatings due to the minimum of the extinction coefficient [8], is stronger and much better centered in the visible due to the second-order interference maximum resulting from the antireflection (AR) role of the three-quarter wavelength YSZ (Section 2.1). Second, the second-order maximum is accompanied by first-order maximum at ≈3×longer wavelength, particularly pronounced below T tr (where it is not suppressed by enhanced absorption) and therefore contributing to enhanced ΔT sol . From the quantitative point of view, it is worth pointing out the exceptionally high T(λ) of the film deposited on the Al 2 O 3 monocrystal (compared to the other films deposited without AR interlayers). This is consistent with the slightly O-rich composition (traces on V 3 O 7 in Fig. 2) and exceptionally low absorption coefficient (Fig. 4(c)) of this specific sample, let alone slight contribution of the AR effect of the relatively Fig. 4. Schematic energy band diagram, based on [32], with two gaps E g1 and E g2 according to Goodenough [24,33] for pure VO 2 (M1) semiconducting phase (a) and for VO 2 (R) metallic phase (b), E F denotes the Fermi energy. (c) ( α E) 1/2 as a function of the photon energy E, where α is the absorption coefficient, measured at T ms = −20 ◦C and T mm =70 ◦C for five W-doped VO 2 films deposited onto various substrates. At −20 ◦C, linear fittings are performed to determine E g1 and E g2 (Tauc plots). The shaded area represents the visible range of the electromagnetic spectrum (380–780 nm). Fig. 5. (a) Spectral transmittance measured at T ms = − 20 ◦C and T mm =70 ◦C for five W-doped VO 2 films deposited onto various substrates. The contours of the shaded areas represent the luminous sensitivity of the human eye (φ lum ) and the solar irradiance spectrum (φ sol ), normalized to maxima of 100 %. Temperature dependences of the transmittance at 2500 nm (b) and the electrical resistivity (c) for the same films as in (a). S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 6 thick surface roughness layer (Fig. 3(e)). The film deposited on the YSZ monocrystal exhibits the opposite: exceptionally low T(λ) (at the same d), consistent with the exceptionally high absorption coefficient (Fig. 4 (c)), possibly indicating slightly metal-rich composition (although this time not leading to any new XRD peaks). The spectral transmittance has been used to calculate the integral quantities T lum and ΔT sol of all five coatings (Table 2). First, the table shows that the performance of the three films prepared without YSZ interlayers is competitive (for homogeneous densified VO 2 ). Note the role of the modulation in the visible and near infrared, small but multiplied by high φ sol . The film deposited on the Al 2 O 3 monocrystal exhibits the undesired sign of this modulation and its high T lum resulting from the aforementioned high T(λ) is therefore accompanied by relatively low ΔT sol , while the film deposited on the YSZ monocrystal once again exhibits the opposite and in turn very high ΔT sol . Second, the table quantifies the improvement of the thermochromic performance achieved by the AR YSZ interlayers. While the tradeoff between T lum and ΔT sol depends on the thickness of W-doped VO 2 (see, e.g., the particularly high T lum of the particularly thin film deposited on YSZ 3000 W), there is a significant enhancement of T lum at a given ΔT sol or vice versa. Third, it is necessary to emphasize that the present paper is focused also on the electrical properties and temperature sensing ability. Thus, while two of the coatings include AR interlayers at the bottom, the coatings do not include similar state-of-the-art second-order AR overlayers at the top [21] which have been used in our papers focused exclusively on the thermochromic performance [13,16,17]. Quantitatively speaking, the present values of T lum =46.5 % and 39.9 % achieved for the W-doped VO 2 films on the YSZ 3000 W and YSZ 100 W interlayers, respectively, would be enlarged by up to 10 % [16], while the corresponding values of ΔT sol =7.2 % and 8.4 % would be enlarged by up to 5 % with an AR overlayer [13]. Fourth, the achievement of the presented thermochromic performance in parallel to the low transition temperatures (T tr ) T = 28 ◦C and 32 ◦C demonstrates the ability of the presented fast scalable preparation technique (contrary to some other techniques cited in [8]) to dope VO 2 with W without suppressing the thermochromic effect. Furthermore, the different growth mechanisms on various templates affected not only the transmittance at temperatures below and above T tr but also the width of transmittance hysteresis, (ΔH) T (Table 2). See the low (ΔH) T of 7 ◦C for the films grown on monocrystalline templates (both epitaxial Al 2 O 3 and non-epitaxial YSZ), compared to (ΔH) T of 11 ◦C-16 ◦C achieved on amorphous and polycrystalline templates. The electrical resistivity and its modulation are shown in Fig. 5(c) and the width of its hysteresis, (ΔH) ρ , is once again quantified in Table 2. The resistivity is not affected by the film thickness, it is not affected by the AR layers, and its differences are explainable by the reported [39] anisotropy of individual crystals (which are however not prone to be equally oriented in the surface plane) only to a small extent in the case of VO 2 (M1) (anisotropic due to V-V dimerization) and even less in the case of VO 2 (R). Thus, the resistivity captures the effect of the growth template, in the first place the effect on the large crystal formation and the importance of this formation, in a more straightforward way than the transmittance. First, the larger values of (ΔH) T (7 ◦C-16 ◦C) compared to (ΔH) ρ (2.5 ◦C-4.4 ◦C) may be explained by larger hysteresis width of smaller crystals [40,41]. On the one hand, the transmittance is given by characteristics of all crystals including the relatively small ones, leading to a considerable hysteresis width. On the other hand, a case can be made that most of the electrical conduction takes place through channels which consist of relatively few relatively large crystals, decreasing the hysteresis width into the range required by applications discussed in Section 1. It is also worth mentioning the beneficial effect of W doping not only in terms of lowering T tr but also in terms of lowering ΔH, see review [5] and the works (dealing with various mechanisms of tailoring the hysteresis by doping) cited therein. Second, the three templates which do not possess any potential for epitaxial growth (glass, monocrystalline YSZ and YSZ 100 W with the same preferred orientation) led to about the same ρ and its modulation. Third, the aforementioned (Fig. 3) ability of the epitaxial template Al 2 O 3 to support the formation of large high-quality crystals of W-doped VO 2 has been captured also in terms of the lowest ρ and the largest ρ modulation. Indeed, the low ρ has to be due to the crystal size and quality, not due to the elemental composition (which is on the contrary slightly O-rich, Fig. 2). The exceptionally low (ΔH) ρ of 2.5 ◦C (well below 3.7 ◦C-4.4 ◦C exhibited by the other coatings) and correlated with low (ΔH) T ) is also consistent with that. The ρ modulation (275×) is not only the largest out of all coatings presented here, but also fully competitive with the literature, taking into account that it has been achieved (i) for VO 2 doped by W (the doping can reduce this amplitude by 1–2 orders of magnitude, see review [5] and the works cited therein) and (ii) at the low substrate temperature T s =350 ◦C (which has also been reported to reduce the amplitude [30,42]). Fourth, the YSZ 3000 W template, second best in terms of at least local epitaxy (Fig. 2), led to the highest ρ and lowest ρ modulation of W-doped VO 2 . While this also cannot be explained by the elemental or phase composition (see the high T and its switching in Fig. 5), it supports the case (Fig. 3) that YSZ (111) is able to affect the preferred orientation of VO 2 but only at a cost of lattice misfit, stress and relatively small crystal size. 3.4. Infrared detection and temperature sensitivity The performance of W-doped VO 2 films in terms of infrared detection and temperature sensitivity is quantified using a temperature dependence of |TCR(T m )| in Fig. 6 and Table 3, and temperature operation ranges at a high detection sensitivity, together with the corresponding electrical resistivities ρ , in Fig. 7 and Table 3. Fig. 6 and Table 3 show very high values of |TCR| for all films, particularly for that deposited onto monocrystalline Al 2 O 3 (14–16 % K −1 ). Taking into account that the highest |TCR| values were achieved, except for the film on the YSZ 3000 W interlayer, in the T m range of 35–40 ◦C, which is close to human body temperature, and that the developed sputter technique can be used for fast deposition of W-doped VO 2 films on large-scale lightweight flexible temperature-sensitive Fig. 6. Temperature dependence of the temperature coefficient of electrical resistance during heating (solid lines) and cooling (dashed lines) for five Wdoped VO 2 films deposited onto various substrates. S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 7 substrates (0.1 mm thick flexible glass in [14]), the obtained results open up new opportunities for smart wearable technology in healthcare, particularly respiratory monitoring (see review [10] and the works cited therein). The relation |TCR| max,h <|TCR| max,c valid for all films (see Fig. 6 and Table 3) is a consequence of a lower ρ during cooling at almost the same value of d ρ /dT m for T m near T tr , see Fig. 5(c). Note that a higher |TCR| max,c is one of the major reasons for the use of cooling curves in VO 2 -based microfluidic gas flow sensors [43]. As can be seen in Fig. 7 and Table 3, a high |TCR| ≥8 % K −1 was achieved during heating of the W-doped VO 2 film on monocrystalline Al 2 O 3 in the T m range from 25 ◦C to 49 ◦C at the corresponding decrease in ρ from 0.08 to 0.004 Ωcm. These very low values of ρ are a result of the epitaxial growth (see Fig. 2) of the film on monocrystalline Al 2 O 3 . In terms of high-performance VO 2 -based sensing layers on low-cost, and even amorphous, substrates, it may be important that |TCR| ≥8 % K −1 was achieved during heating of the film on the YSZ 100 W interlayer in the T m range from 29 ◦C to 44 ◦C at the corresponding decrease in ρ from 0.14 to 0.03 Ωcm, and even during heating of the film on bare SLG in the T m range from 27 ◦C to 43 ◦C at the corresponding decrease in ρ from 0.23 to 0.05 Ωcm. For application to uncooled microbolometers, TCR(T m ) should remain constant over the full operation range of T m in order to avoid additional corrections of the TCR(T m ) dependence. Recently, V 1-x W x O 2 - based multilayer structures with x increasing with height, which exhibited almost constant |TCR| near 10 % K −1 [27] and 7 % K −1 [12] in a wide temperature operation range, were proposed and fabricated. Here, it should be mentioned that only a small modification of the presented sputter technique (single V and W targets as in [14]) would make it possible to prepare W-doped VO 2 multilayer coatings with an upward gradient of the W content on large-scale substrates at a substantially decreased T s =330 ◦C-350 ◦C. 4. Conclusions Thermochromic films of V 33.1 W 0.6 O 66.3 have been deposited using a controlled deep oscillation magnetron sputtering of a single V-W target at a low temperature of 350 ◦C onto five different templates: soda-lime glass, monocrystalline Al 2 O 3 (0001), monocrystalline YSZ (100) and two YSZ interlayers sputtered on glass with preferred cubic orientations (200) and (111). The W doping successfully lowered the transition temperature, not only toward room temperature according to the transmittance hysteresis (28 ◦C-33 ◦C; applications on energy-saving smart windows) but at the same time also toward the human body temperature according to the resistivity hysteresis (30 ◦C-39 ◦C; applications in temperature and fluid-flow rate sensors, including novel human health monitoring systems). The application potential in the field of infrared detectors (including novel uncooled microbolometers), and temperature and fluid-flow rate sensors is further emphasized by the low electrical resistivity and the very low width of the resistivity hysteresis of 2.5 ◦C-4.4 ◦C. The use of Al 2 O 3 led to an epitaxial growth of large crystals of VO 2 (R) (200) resulting from the small ≈4 % lattice misfit, the use of YSZ (111) supported the preferred orientation of VO 2 (R) (110) but at a cost of stress and small crystal size resulting from the large ≈10 % lattice misfit, and the other three templates did not lead to any preferred orientation. Consequently, the W-doped VO 2 film on Al 2 O 3 exhibits the lowest resistivity, the largest resistivity modulation (275×), the largest temperature coefficient of electrical resistance, TCR, (up to −16 % K −1 ) and the broadest (25 ◦C to 49 ◦C during heating) operation range where |TCR| ≥ 8 % K −1 . The W-doped VO 2 film on more industrially relevant YSZ with preferred orientation (200) exhibits the second lowest resistivity (closely followed by the other films without any preferred orientation), only slightly smaller TCR (up to −13 % K −1 ) and slightly narrower (29 ◦C to 44 ◦C during heating) operation range. The W-doped VO 2 on YSZ with preferred orientation (111) exhibits the least attractive resistivityrelated properties, emphasizing the importance of template orientation. The results are important for a new design of high-performance infrared detectors, and high-performance temperature and fluid-flow rate sensors prepared by a fast low-temperature scalable synthesis. Promising optical properties were achieved for the films on both YSZ interlayers supported by their antireflection role due to the optimum thickness. The key integral quantities T lum =46.5 % at ΔT sol =7.2 % and T lum =39.9 % at ΔT sol =8.4 % obtained without any antireflection overlayer should be understood as a basis for further improvement of T lum by up to 10 % and ΔT sol by up to 5 % due to such an overlayer. Taking into account that our fast low-temperature scalable sputter deposition technique can be modified easily to make it possible to prepare more complex W and Sr codoped VO 2 films and W-doped VO 2 -SiO 2 films on these cubic YSZ interlayers with expected much higher T lum and ΔT sol compared to those achieved for the presented W-doped VO 2 films, the results constitute a pathway toward further improvement of thermochromic performance of VO 2 -based coatings with a low transition temperature for energy-saving smart windows. CRediT authorship contribution statement Sadoon Farrukh: Visualization, Methodology, Investigation, Data Table 3 Maximum |TCR| values and operation ranges of the temperatures T m , together with the corresponding electrical resistivities ρ , during heating (h) and cooling (c) for five W-doped VO 2 films deposited onto various substrates. The operation ranges are defined by the T m values for which |TCR(T m )| ≥8 % K −1 . Substrate Maximum |TCR| Operation range, heating Operation range, cooling |TCR| max,h ( % K −1 ) |TCR| max,c ( % K −1 ) T m ( ◦C) ρ (Ωcm) T m ( ◦C) ρ (Ωcm) SLG 11.5 13.1 27 – 43 0.23 – 0.05 46 – 25 0.03 – 0.23 YSZ 3000 W on SLG 8.9 9.4 24 – 33 0.72 – 0.37 37 – 26 0.20 – 0.48 YSZ 100 W on SLG 11.1 13.2 29 – 44 0.14 – 0.03 48 – 27 0.01 – 0.13 Mono YSZ 12.0 12.6 35 – 46 0.15 – 0.05 53 – 33 0.02 – 0.13 Mono Al 2 O 3 14.1 15.9 25 – 49 0.08 – 0.004 57 – 28 0.002 – 0.05 Fig. 7. Temperature coefficient of electrical resistance during heating and the corresponding electrical resistivity for five W-doped VO 2 films deposited onto various substrates. The temperatures defining the operation ranges with |TCR (T m )| ≥8 % K −1 are also given for all films. S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 8 curation. Jaroslav Vlˇ cek: Writing – original draft, Investigation. Jiˇ rí Rezek: Investigation. Jiˇ rí Houˇ ska: Writing – original draft, Investigation. Radomír ˇ Cerstvý: Investigation. Tom´ aˇ s Koz´ ak: Investigation. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This work was supported by the project Quantum materials for applications in sustainable technologies (QM4ST), funded as Project No CZ.02.01.01/00/22_008/0004572 by Programme Johannes Amos Comenius, call Excellent Research. The authors acknowledge Dr. Stanislav Haviar for his elemental composition analysis of the films and Mr. Tom´ aˇ s T˝ olg for the technical support. Data availability Data will be made available on request. References [1] F.J. Morin, Oxides which show a metal-to-insulator transition at the Neel temperature, Phys. Rev. Lett. 3 (1959) 34–36. [2] K. Liu, S. Lee, S. Yang, O. Delaire, J. Wu, Recent progresses on physics and applications of vanadium dioxide, Mater. Today 21 (2018) 875. [3] Y. Ke, S. Wang, G. Liu, M. Li, T.J. White, Y. Long, Vanadium dioxide: the multistimuli responsive material and its applications, Small. 14 (2018) 1802025. [4] R. Shi, N. Shen, J. Wang, W. Wang, A. Amini, N. Wang, C. Cheng, Recent advances in fabrication strategies, phase transition modulation, and advanced applications of vanadium dioxide, Appl. Phys. Rev. 6 (2019) 011312. [5] Y. Xue, S. Yin, Element doping: a marvelous strategy for pioneering the smart applications of VO 2 , Nanoscale 14 (2022) 11054–11097. [6] E. Haddad, R.V. Kruzelecky, P. Murzionak, W. Jamroz, K. Tagziria, M. Chaker, B. Ledrogoff, Review of the VO 2 smart material applications with emphasis on its use for spacecraft thermal control, Front. Mater. 9 (2022) 1013848. [7] X. Cao, T. Chang, Z. Shao, F. Xu, H. Luo, P. Jin, Challenges and opportunities toward real application of VO 2 -based smart glazing, Matter 2 (2020) 862–881. [8] J. Houska, Design and reactive magnetron sputtering of thermochromic coatings, J. Appl. Phys. 131 (2022) 110901. [9] Z. Zhang, L. Zhang, Y. Zhou, Y. Cui, Z. Chen, Y. Liu, J. Li, Y. Long, Y. Gao, Thermochromic energy efficient windows: fundamentals, recent advances, and perspectives, Chem. Rev. 123 (2023) 7025–7080. [10] Y. Zhou, S. Su, Z. Zhu, D. Hou, H. Zhang, Y. Cao, Thermal-stimulated phase transition of vanadium dioxide enabling versatile transduction and smart applications, J. Appl. Phys. 136 (2024) 130901. [11] P.V.K. Yadav, I. Yadav, B. Ajitha, A. Rajasekar, S. Gupta, Y.A.K. Reddy, Advancements of uncooled infrared microbolometer materials: a review, Sens. Actuators A: Phys. 342 (2022) 113611. [12] C. Wheeler, Y. Zhu, K. Sun, B. Ding, R. Huang, O.L. Muskens, C.H. (Kees) de Groot, Multilayer W-doped vanadium dioxide thermal sensors with extended operation region, iScience 28 (2025) 112528. [13] M. Kaufman, J. Vlˇ cek, J. Houˇ ska, R. ˇ Cerstvý, S. Farrukh, M. Chargaoui, S. Haviar, J. C. Jiang, E.I. Meletis, ˇ S. Kos, High-performance thermochromic YSZ/ V 0.986 W 0.014 O 2 /YSZ coatings for energy-savings smart windows, Sol. Energy Mater. Sol. Cells 263 (2023) 112570. [14] J. Rezek, J. Szelwicka, J. Vlˇ cek, R. ˇ Cerstvý, J. Houˇ ska M. Fahland, J. Fahlteich, Transfer of the sputter technique for deposition of strongly thermochromic VO 2 - based coatings on ultrathin flexible glass to large-scale roll-to-roll device, Surf. Coat. Technol. 442 (2022) 128273. [15] S. Loquai, B. Baloukas, J.E. Klemberg-Sapieha, L. Martinu, HiPIMS-deposited thermochromic VO 2 films with high environmental stability, Sol. Energy Mater. Sol. Cells 160 (2017) 217–224. [16] D. Kolenatý, J. Vlˇ cek, T. B´ arta, J. Rezek, J. Houˇ ska, S. Haviar, High-performance thermochromic VO 2 -based coatings with a low transition temperature deposited on glass by a scalable technique, Sci. Rep. 10 (2020) 11107. [17] J.C. Jiang, T. B´ arta, J. Vlˇ cek, J. Houˇ ska, E.I. Meletis, Microstructure of highperformance thermochromic ZrO 2 /V 0.984 W 0.016 O 2 /ZrO 2 coating with a low transition temperature (22 ◦C) prepared on flexible glass, Surf. Coat. Technol. 424 (2021) 127654. [18] T. Koz´ ak, J. Vlˇ cek, A parametric model for reactive high-power impulse magnetron sputtering of films, J. Phys. D 49 (2016) 055202. [19] J. Vlˇ cek, M. Kaufman, A.D. Pajdarov´ a, S. Haviar, R. ˇ Cerstvý, J. Houˇ ska, M. Farahani, Detailed pathway for a fast low-temperature synthesis of strongly thermochromic W-doped VO 2 films with a low transition temperature, J. Phys. D 56 (2023) 505301. [20] J. Rezek, J. Vlˇ cek, J. Houˇ ska, J. ˇ Capek, P. Baroch, Enhancement of the deposition rate in reactive mid-frequency ac magnetron sputtering of hard and optically transparent ZrO 2 films, Surf. Coat. Technol. 336 (2018) 54–60. [21] S. J¨ ager, B. Szyszka, J. Szczyrbowski, G. Br¨ auer, Comparison of transparent conductive oxide thin films prepared by a.c. and d.c. reactive magnetron sputtering, Surf. Coat. Technol. 98 (1998) 1304–1314. [22] J. Houska, D. Kolenaty, J. Vlcek, T. Barta, J. Rezek, R. Cerstvy, Significant improvement of the performance of ZrO 2 /V 1-x W x O 2 /ZrO 2 thermochromic coatings by utilizing a second-order interference, Sol. Energy Mater. Sol. Cells 191 (2019) 365–371. [23] Available online: rredc.nrel.gov/solar/spectra/am1.5/; http://hyperphysics.phyastr.gsu.edu/hbase/vision/efficacy.html; (accessed on 14 June 2016). [24] S. Hu, S.Y. Li, R. Ahuja, C.G. Granqvist, K. Hermansson, G.A. Niklasson, R. H. Scheicher, Optical properties of Mg-doped VO 2 : absorption measurements and hybrid functional calculations, Appl. Phys. Lett. 101 (2012) 201902. [25] M.K. Dietrich, B.G. Kramm, M. Becker, B.K. Meyer, A. Polity, P.J. Klar, Influence of doping with alkaline earth metals on the optical properties of thermochromic VO 2 , J. Appl. Phys. 117 (2015) 185301. [26] B. Li, S. Tian, Z. Wang, B. Liu, X. Gong, X. Zhao, Thermochromic Ta doped VO 2 films: enhanced luminous transmittance, significantly depressed phase transition temperature and hysteresis width, Appl. Surf. Sci. 568 (2021) 150959. [27] N. Emond, A. Hendaoui, M. Chaker, Low resistivity W x V 1-x O 2 -based multilayer structure with high temperature coefficient of resistance for microbolometer applications, Appl. Phys. Lett. 107 (2015) 143507. [28] S. Lee, C. Cheng, H. Guo, K. Hippalgaonkar, K. Wang, J. Suh, K. Liu, J. Wu, Axially engineered metal-insulator phase transition by graded doping VO 2 nanowires, J. Am. Chem. Soc. 135 (2013) 4850–4855. [29] The International Centre for Diffraction Data, PDF-4+Database, The International Centre for Diffraction Data, Newtown Square, PA, USA, 2022. [30] J. Narayan, V.M. Bhosle, Phase transition and critical issues in structure-property correlations of vanadium oxide, J. Appl. Phys. 100 (2006) 103524. [31] K. Okimura, J. Sakai, Changes in lattice parameters of VO 2 films grown on c-plane Al 2 O 3 substrates across metal-insulator transition, Jpn. J. Appl. Phys. 48 (2009) 045504. [32] N.B. Aetukuri, A.X. Gray, M. Drouard, M. Cossale, L. Gao, A.H. Reid, R. Kukreja, H. Ohldag, C.A. Jenkins, E. Arenholz, K.P. Roche, H.A. Durr, M.G. Samant, S.S. P. Parkin, Control of the metal-insulator transition in vanadium dioxide by modifying orbital occupancy, Nat. Phys. 9 (2013) 661–666. [33] J.B. Goodenough, The two components of the crystallographic transition in VO 2 , J. Solid. State Chem. 3 (1971) 490. [34] T.C. Koethe, Z. Hu, M.W. Haverkort, C. Schüssler-Langeheine, F. Venturini, N. B. Brookes, O. Tjernberg, W. Reichelt, H.H. Hsieh, H.-J. Lin, C.T. Chen, L.H. Tjeng, Transfer of spectral weight and symmetry across the metal-insulator transition in VO 2 , Phys. Rev. Lett. 97 (2006) 116402. [35] M.M. Qazilbash, K.S. Burch, D. Whisler, D. Shrekenhamer, B.G. Chae, H.T. Kim, D. N. Basov, Correlated metallic state of vanadium dioxide, Phys. Rev. B 74 (2006) 205118. [36] M.M. Qazilbash, A.A. Schafgans, K.S. Burch, S.J. Yun, B.G. Chae, B.J. Kim, H. T. Kim, D.N. Basov, Electrodynamics of the vanadium oxides VO 2 and V 2 O 3 , Phys. Rev. B 77 (2008) 115121. [37] M.K. Dietrich, F. Kuhl, A. Polity, P.J. Klar, Optimizing thermochromic VO 2 by codoping with W and Sr for smart window applications, Appl. Phys. Lett. 110 (2017) 141907. [38] Z. Gao, Z. Liu, Y. Ping, Z. Ma, X. Li, C. Wei, C. He, Y. Liu, Low metal-insulator transition temperature of Ni-doped vanadium oxide films, Ceram. Int. 47 (2021) 28790–28796. [39] J. Lu, K.G. West, S.A. Wolf, Very large anisotropy in the dc conductivity of epitaxial VO 2 thin films grown on (011) rutile TiO 2 substrates, Appl. Phys. Lett. 93 (2008) 262107. [40] F. Dumas-Bouchiat, M. Gaudin, I.A. Alonzo Zapata, C. Champeaux, VO 2 thin films: various microstructures for hysteresis manipulations, Vacuum. 227 (2024) 113408. [41] H. Zhang, Z. Wu, X. Wu, W. Yang, Y. Jiang, Transversal grain size effect on the phase-transition hysteresis width of vanadium dioxide films comprising spheroidal nanoparticles, Vacuum. 104 (2014) 47. [42] D. Ruzmetov, K.T. Zawilski, V. Narayanamurti, S. Ramanathan, Structurefunctional property relationships in rf-sputtered vanadium dioxide thin films, J. Appl. Phys. 102 (2007) 113715. [43] Y. Cao, T. Wang, N. Sepulveda, Microfluidic thermal flow sensor based on phasechange material with ultra-high thermal sensitivity, J. Mater. Chem. C 11 (2023) 1278–1284. S. Farrukh et al. Applied Surface Science Advances 30 (2025) 100870 9