Influences of W Content on the Phase Transformation Properties and

Research Article Cite This: ACS Comb. Sci. 2018, 20, 229−236

pubs.acs.org/acscombsci

Influences of W Content on the Phase Transformation Properties and the Associated Stress Change in Thin Film Substrate Combinations Studied by Fabrication and Characterization of Thin Film V1−xWxO2 Materials Libraries Xiao Wang,† Detlef Rogalla,‡ and Alfred Ludwig*,† †

Institute for Materials, Ruhr-Universität Bochum, D-44801 Bochum, Germany RUBION, Ruhr-Universität Bochum, D-44801 Bochum, Germany

‡

S Supporting Information *

ABSTRACT: The mechanical stress change of VO2 film substrate combinations during their reversible phase transformation makes them promising for applications in micro/nanoactuators. V1−xWxO2 thin film libraries were fabricated by reactive combinatorial cosputtering to investigate the effects of the addition of W on mechanical and other transformation properties. High-throughput characterization methods were used to systematically determine the composition spread, crystalline structure, surface topography, as well as the temperature-dependent phase transformation properties, that is, the hysteresis curves of the resistance and stress change. The study indicates that as x in V1−xWxO2 increases from 0.007 to 0.044 the crystalline structure gradually shifts from the VO2 (M) phase to the VO2 (R) phase. The transformation temperature decreases by 15 K/at. % and the resistance change is reduced to 1 order of magnitude, accompanied by a wider transition range and a narrower hysteresis with a minimal value of 1.8 K. A V1−xWxO2 library deposited on a Si3N4/SiO2-coated Si cantilever array wafer was used to study simultaneously the temperature-dependent stress change σ(T) of films with different W content through the phase transformation. Compared with σ(T) of ∼700 MPa of a VO2 film, σ(T) in V1−xWxO2 films decreases to ∼250 MPa. Meanwhile, σ(T) becomes less abrupt and occurs over a wider temperature range with decreased transformation temperatures. KEYWORDS: V1−xWxO2 materials libraries, phase transformation properties, fhigh-throughput characterization, mechanical stress, micro/nanoactuators

■

INTRODUCTION

is surrounded by an octahedron of O atoms. Along the rutile c axis, VO6 octahedra share edges, and V atoms are equally positioned with a distance of 0.288 nm. The monoclinic VO2 (M1) phase belongs to the crystallographic group P21/c, with a = 0.575 nm, b = 0.452 nm, c = 0.538 nm, β = 122.6°. V atoms pair along the rutile c axis in zigzag type, exhibiting alternative distances of 0.312 and 0.265 nm.6 Along the transformation from M phase to R phase, the unit cell of VO2 expands by ∼0.32% (one unit cell of M phase consists of two unit cells of R phase) and the area for (011) M and (110) R planes is

VO2 is of interest because of its promising potential in various applications, such as energy-efficient windows, switching devices, smart radiating devices, and nanoactuators.1−4 It undergoes an insulator-to-metal transformation (IMT) at a critical temperature Tc of 68 °C, accompanied by significant changes in optical and electrical properties, which makes it a multifunctional material. At T < Tc, VO2 is in the semiconductor state with monoclinic structure VO2 (M), while at T > Tc, it transforms to the metallic state with rutile structure VO2 (R). The phase transformation is reversible and fast, occurring within 1 ps.5 Rutile VO2 (R) belongs to the crystallographic group P42/ mnm, with a = b = 0.455 nm, c = 0.285 nm, in which a V atom © 2018 American Chemical Society

Received: December 24, 2017 Revised: February 27, 2018 Published: March 5, 2018 229

DOI: 10.1021/acscombsci.7b00192 ACS Comb. Sci. 2018, 20, 229−236

Research Article

ACS Combinatorial Science

stress measurements, a V1−xWxO2 library was deposited on a Si cantilever array wafer (CAW).14 The 100 mm diameter CAW consists of 30 equidistant parallel cantilevers (15 mm long, 1 mm wide and 150 μm thick, coated with stress-compensated SiO2/Si3N4 (SiO2 100 nm, Si3N4 300 nm). To achieve a composition spread over the CAW and simultaneously uniform thickness on each cantilever, the CAW was positioned in a way that the longitudinal direction of the cantilevers is perpendicular to the V−W target direction. RF power of the W target for deposition on the Si strip and the CAW are 50 and 46 W. The substrates were held on a carrier wafer which was put on a calibrated heater during deposition. The temperature of the carrier wafer was calibrated to be 520 °C. After deposition, the films were cooled down to room temperature in vacuum. Characterization of V1−xWxO2 Thin Film Libraries. The thin film composition spread of the materials library was determined using Rutherford backscattering spectrometry (RBS) at the 4 MV accelerator facility of RUBION/RuhrUniversity Bochum. RBS measurements were carried out using a 2 MeV He+ beam. The beam was collimated to a diameter of 1 mm and its intensity was in the range from 30 to 80 nA on the sample, which was tilted by an angle of typically 7 degrees. The collected beam charge was in the order of 20 μC. For the detection of the backscattered particles, a Si solid state detector was used (160 degrees, Cornell geometry, solid angle 1.9 msrad). All recorded data were analyzed using SIMNRA software.15 Measurements on the V1−xWxO2 library were carried out every 6 mm, along the W gradient. The crystallinity of the thin films in the library was measured using an X-ray diffractometer (XRD, Bruker D8, USA) in θ/2θ geometry, equipped with a microfocus source and a Vantec-500 area detector. The V1−xWxO2 library was measured on the same measurement areas as in the RBS tests. The surface microstructure of the library was measured by atomic force microscopy (AFM, Bruker Dimension Fastscan, USA) using Scanasyst mode. The temperature-dependent resistance R(T) of the library was measured on a custom-made high-throughput test stand (HTTS), specially designed for the rapid characterization of thin-film materials libraries.16 A custom-made four-point probe head with 20 pins (5 × 4 groups) was used. As a result, five measurement areas are measured simultaneously to achieve high throughput. The distance between each pin is 0.5 mm and the distance between each group is 4.5 mm. The library is placed on a temperature-controlled stage, and was measured from −40 to 100 °C. To avoid frost on the sample at low temperatures, the measurement table is enclosed in a box which is flooded with N2 during the measurement. The temperature-dependent stress change σ(T) of the cantilever array was measured on a custom-designed laser test stand by monitoring the deflections of each cantilever of the CAW simultaneously.17 The deflections of the cantilevers are tracked by the movement of the reflected laser spot on the diffusing panel, which is recorded by a CCD camera. Based on Stoney’s law,18 the stress is calculated by

decreased by 1.7%, resulting in a strain of up to 1% along the rutile c axis. Taking the high Young’s modulus E of VO2 (∼140 GPa) into account, this leads to high stresses associated with the phase transformation, which could be used for micro/nanoactuators.7 For instance, it was observed that a VO2-coated Si microcantilever can show a curvature change of over 2000 m−1 and σ(T) of ∼1 GPa across the phase transformation. Wu et al. reported a set of VO2-based microactuators, which exhibit high displacement-to-length ratios up to 1 in the sub-100 μm length scale during bending, and work densities over 0.63 J/cm.7,8 Tc can be influenced by the addition of third elements to VO2: for example, Tc is increased by addition of Cr,9 Ti,10 whereas it is decreased by W11 and Mo.12 The observed changes of Tc are associated with changes in the electrical and optical properties of VO2 during the IMT. It is likely that the mechanical response of VO2 across the IMT is also influenced, considering structural distortions induced by the addition of a third element. So far, only one study reports a curvature change through the IMT of 2900 m−1 for a V0.976Cr0.024O2-coated cantilever, being higher than 1850 m−1 for a VO2-coated cantilever.9 The influence of Tc-decreasing elements on the mechanical performance of VO2 has not been studied. Here, the effect of W additions on the transformation properties of VO2 are efficiently investigated by combinatorial deposition and high-throughput characterization of V1−xWxO2 thin film libraries.

■

EXPERIMENTAL DETAILS Fabrication of V1−xWxO2 Thin Film Libraries. V1−xWxO2 thin film libraries were fabricated by reactive cosputtering from 100 mm diameter elemental targets V (purity 99.95%) and W (purity 99.95%) using a magnetron sputter system (AJA International, USA) as shown in Figure 1. The sputter chamber

Figure 1. Schematic diagram (not to scale) for reactive codeposition of a V1−xWxO2 library from the elemental targets V and W on (a) a strip of Si (side view) and (b) a cantilever array wafer (top view). (c) Side view of a coated cantilever. The unit of all labeled dimensions is mm.

was evacuated to 2 × 10−5 Pa. Then 80 sccm of Ar was introduced into the chamber through a MFC (Mass Flow Controller) and the chamber pressure was equilibrated to 1.3 Pa. After 5 min of presputtering, the pressure was adjusted to 0.4 Pa for deposition. A plasma emission monitor (PEM) system (Flotron, Nova Fabrica, Lithuania) was used to dynamically control the oxygen flow rate.13 V and W targets were positioned apposing each other. Pulsed DC power of 300 W was applied to V target under pulsing frequency of 10 kHz and reverse time of 5 μs, and RF power was used for the W target. To obtain a library, a 90 mm long strip of (100) Si substrate was positioned along the V−W target direction. For

σf =

Es d2 1 × × s × Δx 12 × L × lc (1 − υs) df

in which L is the vertical distance between the cantilever and diffusing panel, lc is the length of the cantilever between the measured position and fixed end, Es and υs are Young’s modulus and Poisson ratio of the substrate, ds is thickness of 230


Research Article

ACS Combinatorial Science the cantilever, df is thickness of the film, and Δx is the shift of the reflected laser spot.

due to a systematic uncertainty of determining the O content, as O and Si substrate signals are overlaying in the RBS spectra. A thickness variation is present along the composition spread because of the geometrically opposite layout of the V and W targets, which is also implied in Figure 2a. Along the arrows’ direction pointing in the direction of increasing W-content, both the signal bands of V and W become narrower gradually, which indicates that film becomes thinner. Film thicknesses could be calculated based on measured areal density and density of VO2. Since there is no referable density data of V1−xWxO2 films, the density of VO2 was used for an estimate. This is acceptable due to the low W contents of x < 0.05. The density of VO2 used for the calculation is 4.67 g/cm3, which is reported to be same for both VO2 M and R phase.19 On the basis of the calculation, the thickness of film varies from 199 nm at the higher V-content end to 85 nm at the higher Wcontent end. The phase constitution of the thin films in the library was analyzed by XRD at room temperature. All diffraction patterns, except of the film with x = 0.044, correspond to polycrystalline VO2 with a dominant peak at 2θ = ∼28° and small peaks at 2θ = ∼40.0°, 55.7°, and 57.9°. In the diffraction pattern of the film with x = 0.044, except for peaks of VO2, there is also an almost negligible peak at 2θ = ∼25.4° belonging to the phase V6O13, which indicates that mixed phases start to grow at the higher W-content end. (For details, see Supporting Information.) Even though the O/(V+W) ratio determined by RBS exhibits a deviation from the stoichiometry of VO2, most of the films in the library show the pure VO2 phase. Therefore, to focus on the effect of W, all films in the library are still referred to as V1−xWxO2 in the discussion. To investigate the change of the phase constitution with addition of W, a comparison of the diffraction patterns focusing on the main peak in the 2θ-region of 27−29° is shown in Figure 3a. The stacked diffraction patterns from bottom to top correspond to V1−xWxO2 films with W content from x = 0.007 to x = 0.044. Probably due to thermal stress, the peak position of VO2 M (011)/VO2 R (110) deviated slightly (∼0.14°) from the reported value. To compare the peak shift caused by addition of W, a pure VO2 film was measured, both at room temperature and 100 °C. The determined peak positions, i.e. VO2 M (011) at 28.03° and VO2 R (110) at 27.81°, were added in Figure 3a as references. The peak position of the diffraction of the (011)M plane shifts from 28.01° to 27.81° with increasing W content, suggesting that the films change their phase from M to R phase. Figure 3b shows the peak position shift of the diffraction angle of the different V1−xWxO2 films compared to the pure VO2 (M) in dependence of W-content. Interestingly, for x > 0.014, a linear relationship between the diffraction peak shift and the W content is observed. Prior literature on the V1−xWxO2 system suggest that addition of W could induce a structural distortion of the VO2 M phase toward the symmetric rutile phase VO2 (R).20,21 Tan et al. found by using X-ray absorption fine structure (XAFS) spectroscopy that with increasing W content, the local structure around V atoms in the M phase gradually evolves into that in R phase.20 Because of the continuous composition gradient of the V1−xWxO2 library, it is possible in our study to observe the gradual evolution of the crystalline structure from M phase to R phase as a consequence of W-addition induced crystalline distortion. Figure 4 shows AFM images of different V1−xWxO2 films from the library. With increasing W content, the roughness of

■

RESULTS AND DISCUSSION Composition and Crystalline Structure. Figure 2 shows the V1−xWxO2 composition spread as determined from RBS.

Figure 2. (a) RBS spectra acquired from the V1−xWxO2 library at different measurement areas. The arrows point to the higher Wcontent end of the library. (b) W content and (c) O/(V+W) ratio along the length of the library calculated from RBS spectra.

The spectra show well-separated bands for V and W. The signal of W is significant despite its small count number, since RBS is highly sensitive to elements with high atomic numbers. All spectra are well simulated in the SIMNRA software and the areal densities for V, W, and O were determined. These results allowed to calculate the W content: along the composition spread it increases from x = 0.007 to x = 0.044, as shown in Figure 2b. Furthermore, the O/(V+W) ratios were calculated (Figure 2c). The weighted average of all ratios of the composition spread is 2.10 ± 0.17. The error of 8% is mainly 231


Research Article


W content indicates a change of −15 ± 0.72 K/at.%. R/Rlt decreases from more than 2 orders of magnitude to 1 order of magnitude, and the transformation range becomes wider (fwhm increases about 5 times to 21.2 K); however, the transformation hysteresis shows about 5 times lower ΔT values, down to 1.8 K. The phase transition of VO2 is considered as being first order, associated with a heterogeneous nucleation process and hysteresis originating from nucleation barriers during the transition.22,23 Moreover, Tan et al.20 have reported that the local structure around W atoms was distorted to form rutile-like VO2 nuclei, and the propagations of these nuclei through the VO2 M1 matrix lower the thermal energy barrier for the phase transition. Thermal kinetic analysis reported by Zhang et al.22 confirms that W addition promotes the transition by reducing the energy barrier. By W addition the hysteresis loop width gets narrower due to an increase of effective nucleation sites. The resistance in the semiconductor state decreases with W addition because of the increase of charge carriers, therefore the magnitude of the resistance change of VO2 across the transition is decreased.24 The reduction in Tc by −15 K/at.% observed in the present work is lower than the reported value of around −22 K/at.%,11,25 which might be caused by the deviation of the O/(V+W) ratio of the films from stoichiometric VO2 as is indicated by RBS measurements. Similarly, the observed resistance changes of the V1−xWxO2 films across the transition were about 1 order of magnitude lower than the value reported by Romanyuk et al.,11 for samples with similar W content. Since the O/(V+W) ratio in the V1−xWxO2 library deviates from stoichiometric VO2, there must be some nonstoichiometric VOx oxides in the film. Those oxides do not experience phase transition in the studied temperature region, and therefore will apparently dominate the total resistance after completion of the phase transition of VO2. As is shown in Figure 5a, at temperatures above Tc, the resistance values of the films exhibit a continuous slight decrease with temperature, confirming a semiconductor behavior instead of a metallic behavior. As a result, the magnitude of total resistance change is deteriorated by the existent nonstoichiometric VOx phase.26 The mechanism of the IMT of the correlated complex oxide VO2 has been debated for decades, leading to two possible explanations: Peierls and Mott transition.27,28 The Peierls transition suggests that the IMT is driven by a structural distortion, while the Mott transition addresses the role of strong electron correlation in the IMT. Research on ternary (V−M)O2 has also been carried out to unveil the mechanism of transition. Recently, some studies by time-resolved spectroscopy and X-ray absorption spectroscopy support the Peierls transition, which demonstrate that third element additions into the VO2 structure induce structural distortions of M to R phase, which promotes the IMT transition, and therefore reduces the transition temperature.5,20 In the V1−xWxO2 library, the structural evolution from VO2 (M) to VO2 (R) by adding W was confirmed. The decrease of Tc with W content supports the assumption of Peierls transition. Temperature-Dependent Stress Change σ(T) measured by coated Si cantilevers. Figure 6 shows the σ(T) curves of a VO2 film. The significant stress change at around 75 °C is related to the phase transformation of VO2. When VO2 transforms from M to R phase, the shrinkage of the unit cell along the rutile c axis induces tensile stress in the film− substrate combination. In addition to the stress change through the phase transformation in the temperature range from 65 to 85 °C, there is compressive stress showing up at T < 65 °C and

Figure 3. (a) XRD diffraction patterns of the V1−xWxO2 library (b) Peak position (rounding error is within 0.005°) shift in dependence of W content.

the film changes from 5.6 nm for x = 0.007 (Figure 4a) to 15.3 nm for x = 0.034 (Figure 4c) and 11.6 nm for x = 0.044 (Figure 4d). The grain size increases from 50 to 90 nm for x = 0.007 to 80−120 nm for x = 0.044. In the films with x > 0.02 (only x = 0.02 and x = 0.044 are shown here), triangular-shaped grains are observed, whereas in a film with x = 0.044, triangularshaped grains fully cover the film surface. Temperature-Dependent Resistance R(T) Measurements. Figure 5a shows results of R(T) measurements of the library. As the W content increases, the transformation properties change gradually, as previously reported.11 Table 1 summarizes detailed information on the transformations. Gauss fitting of the first derivatives of heating and cooling curves for each sample was performed. Tc is determined as the peak point of the Gauss fitting of the first derivative of a heating curve; the hysteresis width value, ΔT, is determined by the difference between the peak points of Gauss fitting for heating and cooling curves; fwhm is the full width at half-maximum of the Gauss fitting for heating curve, indicating the transition width, R/Rlt is the resistance R normalized by the resistance in the low temperature (lt) semiconductor state Rlt. The R(T) curves of the library demonstrate that Tc decreases with increasing W content. Linear fitting of Tc as a function of 232


Research Article


Figure 4. Surface microstructures of V1−xWxO2 films (AFM) of different W content x: (a) x = 0.007, (b) x = 0.02, (c) x = 0.034, (d) x = 0.044.

Table 1. Detailed Parameters Characterizing Transformation Performance of V1−xWxO2 Library V1−xWxO2 x

TC (°C)

ΔT (K)

fwhm (K)

0 0.007 0.011 0.016 0.023 0.034 0.044

73.5 56.6 51 43.8 35.4 19.1 5.2

8.8 8.5 7.7 7.5 6.7 2.6 0.3

4.4 8.6 12.6 17.4 29.6 37.1 43.2

R/Rlt 2.5 9.5 1.8 2.5 3.2 4.9 8.4

× × × × × × ×

10−3 10−3 10−2 10−2 10−2 10−2 10−2

Figure 6. Temperature-dependent stress change σ(T) of a VO2 film/ substrate combination on the CAW.

Figure 5. (a) Results of R(T) measurements on the V1−xWxO2 library: solid lines correspond to heating curves and dotted lines correspond to cooling curves. (b) Transition temperature Tc as a function of W content with a linear fit indicated by the red line.

the σ(T) curves, a stress change of ∼700 MPa was determined for the phase transformation of the VO2 film. A V1−xWxO2 library was deposited on a CAW and measured by the high-throughput cantilever test stand. Figure 7a shows the σ(T) curve of a V1−xWxO2 film with x = 0.013. Figure 7b shows σ(T) curves of V1−xWxO2 films with different W content. The W content determined by RBS ranges from x = 0.013 to x = 0.03 along the cantilever array. In Figure 7a, the heating curve exhibits tensile stress change of ∼130 MPa from 30 to 60 °C through the phase transformation. From −10 to 30 °C, there is

T > 85 °C, which is the result of thermal stress caused by the difference in the thermal expansion coefficients of the VO2 film and the cantilever substrate. Compared to the stress induced by the phase transformation, the thermal stress is less significant and can be ignored when discussing the stress change. From 233


Research Article


between the heating and cooling σ(T) curves exists and results in the unusual hysteresis appearance. To study the stress change in the V1−xWxO2 library, the relative stress change (subtracted by the stress at 100 °C) is adopted, and only heating curves are considered. The σ(T) measurements of Figure 7b show that with increasing W content, the stress change becomes less abrupt and occurs over a wider temperature range, whereas critical temperature for stress change decreases gradually. To compare the stress change through IMT for V1−xWxO2 films with different W content and neglect the effect of thermal stress, the same temperature range (−50−100 °C) was studied: as x increases from 0.013 to 0.019, the stress change increases from ∼120 MPa to ∼200 MPa and then decreases to ∼110 MPa for x from 0.023 to 0.03. To compare the phase-transformation induced σ(T) of a VO2 film and that of V1−xWxO2 films, the same temperature range should be used. For the VO2 film, the stress change of ∼700 MPa is determined between −65 and 85 °C, whereas for V1−xWxO2 films, because of the less abrupt σ(T) curve, a larger temperature span is considered and as a result, more tensile stress could be offset by the compressive thermal stress. However, it still can be concluded that the tensile stress change values induced by phase transformation in V1−xWxO2 films decrease significantly, compared with that in a VO2 film. Taking the film with x = 0.019 as an example, the stress change between −20 and 50 °C is ∼250 MPa. The Tc determined by the first derivative of σ(T) was plotted in Figure 7c, with the Tc from R(T) added as a comparison. As the amount of W increases to x = 0.03, Tc decreases to ∼7.5 °C. Compared with Tc determined from R(T), the Tc for σ(T) is lower for the same W amount. Furthermore, the lowering of Tc by −22.8 ± 0.72 K/at.% for the cantilever samples was determined by linear fit, which is higher than that for the V1−xWxO2 library deposited on Si substrate. The average of the O/(V + W) ratios over the cantilever library is calculated to be 1.98 ± 0.10 based on RBS spectra, which is approximate to the stoichiometric ratio of VO2 and might result in the higher reduction in Tc. With addition of W, the structure distortion promotes the instability of VO2(M) phase and consequently decreases the transition temperature. Accordingly, the distortion of VO2 (M) weakens the stress change of the film through the transformation, as indicated by the σ(T) measurements. However, unlike the monotonous trend of decreased Tc and broadened transition region with increasing W amount x, the amplitude of stress change first experiences a slight increase until x = 0.019, then decreases for x ≥ 0.023. The microstructures of the films from the cantilever library were further investigated by AFM, see Supporting Information. Table S1 and Figure S2 show that the microstructure of VO2 on cantilever is significantly different from that of the V1−xWxO2 library, showing a smoother surface (roughness 6.6 nm) and smaller grains in the range from 70 to 130 nm, which might contribute to the significant difference in stress change between the binary VO2 film and the V1−xWxO2 library. For the V1−xWxO2 films, the roughness of the film decreases monotonously from 28.2 nm for x = 0.013 to 14.3 nm for x = 14.6. The grain size of the films in the library first increases from 130−330 to 130−450 nm until x = 0.019, then decreases to 120−330 nm as x increases to 0.03. Notably, the trend of grain size change is similar to that of the stress change along the V1−xWxO2 library, which suggest that the σ(T) behavior of V1−xWxO2 library is also influenced by the microstructure of the film besides the W content.

Figure 7. Relative temperature-dependent stress change σ(T) of V1−xWxO2 thin films: (a) heating and cooling curves for a film with x = 0.013 and (b) heating curves for V1−xWxO2 films with different W content. The measured stress is subtracted by the stress at 100 °C. (c) Transition temperatures Tc determined by R(T) and σ(T) on two different libraries as a function of W content with linear fits indicated by the colored solid lines. The Tc for σ(T) is determined by the first derivative of σ(T) except the point at x = 0.03, where it was was determined by the temperature of half stress change on the plot.

slight tensile stress (∼35 MPa), which indicates the phase transformation of V0.987W0.013O2 film over a wider temperature range. Because of the offset of thermal compressive stress, the onset of tensile stress change is less significant. Unlike the hysteresis between heating and cooling σ(T) curves of the VO2 film, the hysteresis in σ(T) curves of V0.987W0.013O2 film is ambiguous. At first glance, the cooling σ(T) is ahead of the heating curve in terms of temperature, indicating a reversed hysteresis. However, the first derivative of σ(T) curves indicates that the Tc for heating and cooling curves is ∼49 and 47.5 °C, respectively. At T > Tc, the compressive thermal stress in the heating curve during the heating process cannot be totally relieved as temperature decreases in the cooling curve. A gap 234


Research Article


■

(7) Rúa, A.; Fernández, F. E.; Sepúlveda, N. Bending in VO2-Coated Microcantilevers Suitable for Thermally Activated Actuators. J. Appl. Phys. 2010, 107, 074506−4. (8) Liu, K.; Cheng, C.; Cheng, Z.; Wang, K.; Ramesh, R.; Wu, J. Giant-Amplitude, High-Work Density Microactuators with Phase Transition Activated Nanolayer Bimorphs. Nano Lett. 2012, 12, 6302−6308. (9) Rúa, A.; Cabrera, R.; Coy, H.; Merced, E.; Sepúlveda, N.; Fernández, F. E. Phase Transition Behavior in Microcantilevers Coated with M1-Phase VO2 and M2-Phase VO2: Cr Thin Films. J. Appl. Phys. 2012, 111, 104502−10. (10) Du, J.; Gao, Y.; Luo, H.; Kang, L.; Zhang, Z.; Chen, Z.; Cao, C. Significant Changes in Phase-Transition Hysteresis for Ti-Doped VO2 Films Prepared by Polymer-Assisted Deposition. Sol. Energy Mater. Sol. Cells 2011, 95, 469−475. (11) Romanyuk, A.; Steiner, R.; Marot, L.; Oelhafen, P. Temperature-Induced Metal−Semiconductor Transition in W-Doped VO2 Films Studied by Photoelectron Spectroscopy. Sol. Energy Mater. Sol. Cells 2007, 91, 1831−1835. (12) Jin, P.; Tanemura, S. V1−XMoxO2 Thermochromic Films Deposited by Reactive Magnetron Sputtering. Thin Solid Films 1996, 281−282, 239−242. (13) Wang, X.; Yan, L.; Li, Y.; Cao, Y. Z. Sputtering Deposition of VO2 Film Using Plasma Emission Monitor as Reactive Gas Flow Rate Feedback Control. Wuji Cailiao Xuebao 2015, 30, 1228−1232. (14) Ludwig, A.; Cao, J.; Savan, A.; Ehmann, M. High-Throughput Characterization of Hydrogen Storage Materials Using Thin Films on Micromachined Si Substrates. J. Alloys Compd. 2007, 446−447, 516− 521. (15) Mayer, M. SIMNRA, A Simulation Program for the Analysis of NRA, RBS and ERDA. AIP Conf. Proc. 1998, 475, 541−544. (16) Thienhaus, S.; Hamann, S.; Ludwig, A. Modular HighThroughput Test Stand for Versatile Screening of Thin-Film Materials Libraries. Sci. Technol. Adv. Mater. 2011, 12, 054206. (17) Zarnetta, R.; Ehmann, M.; Savan, A.; Ludwig, A. Identification of Optimized Ti−Ni−Cu Shape Memory Alloy Compositions for High-Frequency Thin Film Microactuator Applications. Smart Mater. Struct. 2010, 19, 065032. (18) Janssen, G. C. A. M.; Abdalla, M. M.; van Keulen, F.; Pujada, B. R.; van Venrooy, B. Celebrating the 100th Anniversary of the Stoney Equation for Film Stress: Developments from Polycrystalline Steel Strips to Single Crystal Silicon Wafers. Thin Solid Films 2009, 517, 1858−1867. (19) Leroux, Ch.; Nihoul, G.; Van Tendeloo, G. From VO2(B) to VO2(R): Theoretical Structures of VO2 Polymorphs and In-Situ Electron Microscopy Theoretical Structures of VO2 Polymorphs. Phys. Rev. B: Condens. Matter Mater. Phys. 1998, 57, 5111−5121. (20) Tan, X.; Yao, T.; Long, R.; Sun, Z.; Feng, Y.; Cheng, H.; Yuan, X.; Zhang, W.; Liu, Q.; Wu, C.; Xie, Y.; Wei, S. Unraveling MetalInsulator Transition Mechanism of VO2 Triggered by Tungsten Doping. Sci. Rep. 2012, 2, 466. (21) Whittaker, L.; Wu, T.-L.; Patridge, C. J.; Sambandamurthy, G.; Banerjee, S. Distinctive Finite Size Effects on the Phase Diagram and Metal−Insulator Transitions of Tungsten-Doped Vanadium(Iv) Oxide. J. Mater. Chem. 2011, 21, 5580−5592. (22) Zhang, H.; Yu, H.; Chen, Z.; Luo, H.; Gao, Y. Thermal Kinetic Analysis of Metal−Insulator Transition Mechanism in W-Doped VO2. J. Therm. Anal. Calorim. 2016, 126, 949−957. (23) Lopez, R.; Haynes, T. E.; Boatner, L. A.; Feldman, L. C.; Haglund, R. F. Size Effects in the Structural Phase Transition of VO2 Nanoparticles. Phys. Rev. B: Condens. Matter Mater. Phys. 2002, 65, 224113. (24) Chae, B. G.; Kim, H. T. Effects of W Doping on the Metal− Insulator Transition in Vanadium Dioxide Film. Phys. B 2010, 405, 663−667. (25) Shibuya, K.; Kawasaki, M.; Tokura, Y. Metal-Insulator Transition in Epitaxial V1−xWxO2(0 ≤ x ≤ 0.33) Thin Films. Appl. Phys. Lett. 2010, 96, 022102.

CONCLUSIONS The combinatorial fabrication of V1−xWxO2 libraries and their high-throughput characterization revealed systematically the influence of W-additions on the transformation properties. Adding W into VO2 induces distortions of its crystal structure which is indicated by gradual phase shift from M to R phase as W content increases from x = 0.007 to x = 0.044. With addition of W, Tc of the library decreases by 15 K/at.%. The resistance change is reduced to 1 order of magnitude, and transition range become wider accompanied by narrower hysteresis of down to 1.8 K. Moreover, distortion of VO2(M) structure caused by W addition reduces the stress change values in V1−xWxO2 films. Compared with the stress change of ∼700 MPa through phase transformation of a VO2 film, σ(T) values of V1−xWxO2 films decrease to ∼250 MPa and becomes less abrupt, occurring at gradually decreased temperatures.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscombsci.7b00192. Complete XRD patterns of the V1−xWxO2 library on Si and AFM images of the V1−xWxO2 library on cantilever array (PDF)

■

AUTHOR INFORMATION

ORCID

Xiao Wang: 0000-0001-5753-5389 Alfred Ludwig: 0000-0003-2802-6774 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to thank Mr. Steffen Salomon and Dr. Sigurd Thienhaus for their assistance during XRD, temperaturedependent resistance, and stress change measurements. ZGH (Zentrum für Grenzflächendominierte Höchstleistungswerkstoffe) is hereby acknowledged for the supply of AFM measurement. This work was supported by the German Research Foundation (DFG) (Project number LU 1175/18-1).

■

REFERENCES

(1) Jerominek, H.; Picard, F.; Vincent, D. Vanadium Oxide Films for Optical Switching and Detection. Opt. Eng. 1993, 32, 2092−2099. (2) Soltani, M.; Chaker, M.; Haddad, E.; Kruzelesky, R. V. Thermochromic Vanadium Dioxide Smart Coatings Grown on Kapton Substrates by Reactive Pulsed Laser Deposition. J. Vac. Sci. Technol., A 2006, 24, 612−617. (3) Wang, X.; Cao, Y.; Zhang, Y.; Yan, L.; Li, Y. Fabrication of VO2Based Multilayer Structure with Variable Emittance. Appl. Surf. Sci. 2015, 344, 230−235. (4) Guo, H.; Wang, K.; Deng, Y.; Oh, Y.; Syed Asif, S. A.; Warren, O. L.; Shan, Z. W.; Wu, J.; Minor, A. M. Nanomechanical Actuation from Phase Transitions in Individual VO2 Micro-Beams. Appl. Phys. Lett. 2013, 102, 231909−5. (5) Cavalleri, A.; Dekorsy, T.; Chong, H. H. W.; Kieffer, J. C.; Schoenlein, R. W. Evidence for a Structurally-Driven Insulator-toMetal Transition in VO2: a View From the Ultrafast Timescale. Phys. Rev. B: Condens. Matter Mater. Phys. 2004, 70, 161102. (6) Eyert, V. The Metal-Insulator Transitions of VO2: a Band Theoretical Approach. Ann. Phys. 2002, 11, 650−702. 235


Research Article

ACS Combinatorial Science (26) Ruzmetov, D.; Zawilski, K. T.; Narayanamurti, V.; Ramanathan, S. Structure-functional Property Relationships in rf-sputtered Vanadium Dioxide Thin Films. J. Appl. Phys. 2007, 102, 113715−7. (27) Yao, T.; Zhang, X.; Sun, Z.; Liu, S.; Huang, Y.; Xie, Y.; Wu, C.; Yuan, X.; Zhang, W.; Wu, Z.; Pan, G.; Hu, F.; Wu, L.; Liu, Q.; Wei, S. Understanding the Nature of the Kinetic Process in a VO2 MetalInsulator Transition. Phys. Rev. Lett. 2010, 105, 226405. (28) Qazilbash, M. M.; Brehm, M.; Chae, B.-G.; Ho, P.-C.; Andreev, G. O.; Kim, B.-J.; Yun, S. J.; Balatsky, A. V.; Maple, M. B.; Keilmann, F.; et al. Mott Transition in VO2 Revealed by Infrared Spectroscopy and Nano-Imaging. Science 2007, 318, 1750−1753.

236


Influences of W Content on the Phase Transformation Properties and

Recommend Documents