IGR Report on Grants GR

1 IGR Report on Grant GR/R03020/01 Prof. I. P. Parkin, Department of Chemistry, University College London Intelligent Thermochromic Coatings; APCVD Prepared Metal Doped Vanadium Oxides 1.1 Background/ Context This project was part of a three university investigation – UCL, Salford University and Oxford Brookes University. The UCL part of the project employed a project studentship and was timed to start some six months (Salford) to a year (Oxford Brookes) before the other universities. UCL was the pathfinder to come up with new synthetic routes to vanadium dioxide and doped vanadium dioxide thin films on glass substrates by APCVD. It should be noted that UCL involvement was only in original objectives 1-4 as listed in the case for support – these remained unchanged during the project which has proven to be spectacularly successful. A final grant report will be forthcoming separately from Salford University who started their leg of the project some five months behind UCL. Oxford Brookes University did not participate in the work. Their section on optical modelling has been taken up in part by Pilkington Glass. This final report reflects the work undertaken at UCL. 1.2 Aim The primary aim of the project was to prepare vanadium dioxide and metal-doped vanadium dioxide thermochromic films by CVD. The secondary objective was to measure the thermochromic transition temperature of the new films. 1.3 Introduction The ability to vary the quantity of solar radiation passing through a window in response to the needs of the internal environment, can avoid overheating during peak periods of solar availability, hence reducing cooling demand, and allow increased glazing area to enhance solar gain during the heating season. Control of how much energy is transmitted through a window could be achieved by a variable transmittance coating. In addition to perceived energy and thermal comfort benefits, controlling the solar gain could enhance visual comfort by reducing glare and, undoubtedly, a window whose appearance can be varied in situ, presents new and exciting opportunities in architectural design. Materials whose optical properties vary in response to an external stimulus are termed “chromogenic”. A thin film of a thermochromic material on an exterior window could modify its reflectance properties, dependant on outside ambient temperature.1 The solar radiation that is not able to pass through the window when it is in its darkened state must be either reflected or absorbed. Ideally, in a cooling dominated application, the window would pass all or part of the visible radiation incident on the window and reflect the majority the Sun’s near infrared radiation. Incident solar radiation, that is not transmitted, is absorbed. Absorption will cause significant heating of the window if left to stagnate in a coloured state under conditions of high irradiance. Temperature rises in the window will give rise to a radiant heat source adjacent to the room, potentially leading to thermal discomfort, and will impose additional demands on the temperature stability of the materials used in the smart window. Effective thermochromic window coatings would respond to this heating by increasing their reflectance and “compensating” for the increased heating by reflecting more heat away. Such intelligent coatings could be used in a myriad of applications including windscreens of automobiles, sunscreens and greenhouses. The development of such coatings will lead to large savings in energy costs (e.g. power to air conditioning units), improved building environments, and environmental benefits (eg. reduced CO2 emmissions). Thermochromic properties and thermochromic materials for window glazing applications have been investigated2, 3 . By the late 80's thermochromic properties had been demonstrated by a number of groups1. Coating techniques employed were PVD or sol gel based4, 5. The technology advances were not transferable into commercially valuable exploitation. The primary reasons for this can be summarized: 1. Optical: transition temperatures too far above room typical ambient temperatures. Visible colour unattractive or too dark. 2. Scientific understanding: lack of in-depth understanding of: how thermochromics will work as energy efficient /comfort enhancing glazings, material properties, film structure, and stoichiometry control. 3. Process: complex and /or difficult to control processes for film manufacture. Films required separate annealing steps. 4. Significant film durability limitations (e.g. "soft" and moisture sensitive). This proposal set out to find a novel CVD approach to investigate the science and technology limitations of earlier work and was based on the formation of vanadium dioxide films by APCVD. Vanadium dioxide is a thermochromic material known to undergo a reversible semiconductor-metal phase transition at 70°C associated with a structural phase change from monoclinic to tetragonal. Below the transition temperature monoclinic VO2 is a narrow band gap semiconductor (0.7 eV). Above the transition temperature the tetragonal (rutile) material exhibits metallic properties (105 increase in electrical conductivity, increased IR reflectance). For intelligent coatings to function effectively the switching temperature has to be reduced to a desirable room temperature (18-28°C). Alterations to the switching temperature of VO2 have been shown to be effective on small-scale sol-gel derived films by the incorporation of dopant levels of W, Mo, Nb and F2. The best record to date is a 2% tungsten loaded sol-gel derived film which is thermochromic at 25°C3. Thermochromic measurements on VO2 single crystals show that the material does not withstand repeated thermal cycling because of mechanical stresses associated with the phase change. Thin films of sol-gel derived VO2, however, are significantly more robust and reports of many thousands of thermal cycles have been recorded before loss of switching characteristics6. For intelligent thermochromic films to be inexpensive and widely available a manufacturing method other than sol-gel is required. For the glass manufacturing 2 industry one of the most important coating methods is APCVD. This is because the appropriate spraying nozzels can be added directly to float glass plants, so that the glass is coated 'in situ' as it is made. Such processes are in wide spread usage for solar control purposes. The challenge at the outset of this project was to prepare and fully characterise thermochromic metaldoped coatings on glass using APCVD. A number of scientific objectives will be achieved in the CVD field including the first demonstration of APCVD of metal-doped vanadium oxide on glass. Previous chemical vapour deposition studies on vanadium oxide films has focused on the formation of V2O5 by low pressure work2. It has proved possible to post-treat the V2O5 films with hydrogen to reduce them to VO2. This procedure requires additional steps and can lead to multiphasic films. Ideally a precursor system should be devised in which the VO2 could be formed directly from the CVD process. Precursors that have been investigated for vanadium oxide deposition have been primarily restricted to the use of oxoalkoxides such as VO(OR)3 and VO(OC3H7)3,6. We have shown that vanadium chloride can be used with various oxygen sources to form vanadium oxide coatings with some control of crystallinity7. 1.4 Overview of Project All of the UCL objectives listed in the initial programme (objectives 1-4) were investigated in detail. All of the planned experiments were studied and to our surprise after initial tinkering all of the CVD experiments worked. This enabled us to study the materials in greater detail than proposed in the original proposal - this included much fuller characterisation both of the VO2 thin films and of the functional switching temperatures. We were awarded eight days of beamtime at the SRS at Daresbury. We also developed a variable temperature cell for study of the VO2 films by reflectance/transmission and Raman measurements. As will be detailed in this report this work has led to five publications in international journals8 with a further three papers in press9. It has also been the subject of two press releases and two patient submissions. The thermochromic work has been highlight as a technological breakthrough in a Nature scientific highlight10. 2.1 Preparation of Vanadium Dioxide Films by APCVD 1:0 1:0 Increasing VOCl 3 5:1 5:1 2:1 VOCl3 : H2O ratio 1:1 V2O5 V2O5 /V6O13 VO2 VCl4 : H2O Ratio Increasing VCl 4 V2O5 3:1 VOx 1:1 NO FILM VO x 1:2 Increasing H 2O 1:3 VO 2/V 6 O13 1:4 1:6 1:7 1:8 V 2O 5 VO2 1 : 20 1 : 57 0:1 350 375 400 425 450 475 500 525 550 575 600 625 650 0:1 250 300 350 400 450 500 550 600 650 700 Substrate temperature / ¼C Vanadium dioxide films on glass were obtained from a wide range of vanadium precursors (VCl4, VOCl3, V(OEt)4) and oxygen sources (H2O, CH3OH, C2H5OH, ethyl acetate). Water proved to be the best source of oxygen. Although methanol and ethanol both gave excellent films some slight carbon contamination was observed. A detailed reaction landscape was investigated of flow conditions, concentrations of precursor and temperature of substrate required to form the VO2 phase. It was found that only particular conditions gave VO2 – other reaction conditions giving rise to V2O5, V6O13 and VOx phases8. Vanadium dioxide films could only be grown with substrate Substrate temperature / ¼C temperatures in excess of 550°C using VOCl3 as a precursor and at 600°C and above using VCl4 (Figure 1). In both systems an excess gas phase concentration of water compared to the vanadium precursor was required to form VO2. The VO2 thin films were characterised by a variety of techniques including Raman, XPS, SEM/EDX, RBS, reflectance-transmission and XRD. In all cases the films showed single phase VO2. Growth rates could be controlled by the precursor gas-phase concentrations and were typically 200 nm min-1 at 600°C. The VO2 films were adhesive (passed the Scotch tape test), yellow and resistant to rubbing. The vanadium dioxide films could be scratched with a steel scalpel. The vanadium dioxide films were shown to be completely stable in air for over two years. All of the VO2 films showed a fully reversible thermochromic transition at 68-70°C. This transition was studied by XRD, reflectance/transmission and Raman measurements (Section 3.1). The transition matched that seen previously in films prepared by PVD and was associated with the monoclinic to tetragonal phase transition. It was also found that for very thin films of VO2 – that is films of less than 250 nm in thickness - a reduction in the switching temperature was observed down to ca 55°C. This observation has been hinted at before for very thin films prepared by PVD and has been attributed to a strain effect that lowers the thermochromic transition11. Figure 1 CVD deposition landscape of VOCl3 (left) or VCl4 (right) to water concentrations against substrate temperature. 2.2 Preparation of Doped Vanadium Dioxide films by APCVD Metal-doped vanadium dioxide films were prepared by APCVD. The process was relatively simple. It employed the same conditions found in Section 2.1 for the formation of the VO2 phase but a secondary precursor was also added via a separate bubbler – so that both the vanadium and dopant metal precursor were in the gas phase at the same time. A number of Increasing H 2O 3 secondary precursors were investigated including WCl6, WCl4, W(OEt)6, NbCl5, TaCl5, TiCl4, SnCl4, MoCl6 and CrO2Cl2. For the W, Mo, Nb and Ta precursors a solid solution was formed V1-xMXO2 (x = 0.001 – 0.03), for titanium and tin composite films VO2 / TiO2 and VO2 / SnO2 were formed whilst for chromium a new phase Cr3VO8 was formed12. The reactions studied, phases observed, range of dopant element and the corresponding thermochromic transition temperatures are summarised in Table 1. It was found that solid solutions were readily formed for W, Mo, Nb and Ta substitutions. In these films the secondary element was found to be homogeneously dispersed throughout the film by XPS, spot EDX and WDX analysis. Raman showed the expected pattern for VO2 and XRD showed again a single phase of material. A slight shift in the X-ray lines was seen with dopant concentration – a good indication of a solid solution. Further band gap measurements using a Tauc plot showed a single transition that was perturbed by dopant concentration. Attempts to dope titanium or tin into the VO2 lattice resulted in a composite material. This was readily picked up by Raman – which showed bands for each individual phase; XRD which showed the patterns for both TiO2 and SnO2 as well as VO2; band gap measurements which showed two discrete band gaps and SEM/EDAX which showed two different sorts of particles. Table 1 Reactions, Conditions and Thermochromic Transitions for Doped Vanadium Dioxide Films Reagents Substrate temperatures/°C 600-650 550-650 550-650 600-650 500-650 550-650 650 650 650 550-650 550-650 Phase observed at room temperature VO2(M) VO2(M) VO2(M) VO2(M) or VO2(R) Cr3VO8 VO2(M) or VO2(R) VO2(M) VO2(M) VO2(M) VO2(M) / TiO2 VO2(M) / SnO2 Percentage range of metal dopant – atom% n.a. n.a. 0.1-1.3 [W] 0.1-3.0 [W] n.a. 0.1 - 3.1 [W] 0.02 - 1.0 [Mo] 0.2-0.5 [Nb] 0.01-0.05 [Ta] composite composite Thermochromic transition temperatures/ °C 58 - 70 58 - 70 42 - 70 5 - 70 No Tc 5 - 70 48 - 70 50-70 50-70 58-70 50-70 Remarks VCl4 + H2O VOCl3 + H2O VOCl3 + [W(OEt)6] + H2O VCl4 + WCl6 + H2O VOCl3 + CrO2Cl2 + H2O VOCl3 + WCl6 + H2O VOCl3 + MoCl6 + H2O VOCl3 + NbCl5 + H2O VOCl3 + TaCl5 + H2O VOCl3 + TiCl4 + H2O VOCl3 + SnCl4 + H2O Low Tc - strain Low Tc - strain Solid solution Solid solution New phase Solid solution Solid solution Solid solution Solid solution Photocatalytic activity Composite film Table 2 APCVD Reaction Conditions and Measured Thermochromic Transitions for VxW1-xO2 Films Formed from the APCVD of VCl4, WCl6 and H2O Reactor Gas phase Gas phase Gas phase Phase from W Transition temp. / ºC amount of WCl6 amount VOCl3 amount H2O / Raman and atom% in temperature ± / mol min-1 / mol min-1 mol min-1 XRD film hysteresis width / ºC 650 0.002 0.014 0.026 VO2(M) 0.6 43 ± 6 650 0.004 0.026 0.026 VO2(M) 0.3 55 ± 7 650 0.008 0.024 0.026 VO2(M) 1.2 35 ± 3 650 0.010 0.019 0.026 VO2(R) 2.6 10(a) 650 0.001 0.021 0.026 VO2(M) 0.4 55 ± 6 650 0.010 0.015 0.026 VO2(R) 3.1 5(a) 650 0.006 0.017 0.026 VO2(M) 0.9 41 ± 3 650 0.008 0.021 0.026 VO2(M) 0.7 45 ± 6 650 0.009 0.020 0.026 VO2(M) 1.9 29(a) 3.1 Functional Analysis – Overview The thermochromic switching temperatures of the doped vanadium dioxide thin films was measured by four techniques. • • X-ray measurements – A variable temperature X-ray cell was built and the thermochromic transition was monitored by X-ray powder diffraction using station 2.6 at the SRS Daresbury. This showed the change in structure from monoclionic to tetragonal that occurs at the transition temperature. (Figure 3) Optical measurements – The variable temperature UV-vis spectra showed significant changes above and below the transition temperature. In particular there was little change in the visible region of the spectra – however very significant increases in reflectance were noted for the tetragonal phase in the far infra-red region and consequential decrease in transmission. This is exactly the behaviour required for a functioning intelligent window coating – that is its optical properties change with ambient temperature in a predictable and controllable way. (Figures 2 and 5). This process was also followed by measuring the absorbance at a set frequency - 2.5 µm. This enabled the hysteresis associated with the transitions to be monitored. (Figure 4) • • 4 Raman measurements – A variable temperature Raman cell enabled the thermochromic transition to be studied. The monoclinic phase gave a good Raman scattering pattern, whilst the high temperature tetragonal phase gave a featureless Raman signature (which can be related to the fact that the phase becomes metallic). (Figure 7) 0.6 (110) tetragonal 0.5 Transmittance Tc Tc= 58 ¼C 0.3 Reflectance T c Intensity 0.2 Temperature /¼C 0.1 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 23.23 23.26 23.29 23.32 23.35 23.38 23.2 23.41 23.44 23.47 23.5 23.53 23.56 23.59 23.62 23.65 23.68 23.71 23.74 23.77 2θ / deg Figure 2 Reflectance transmission plots for V0.994W0.006O2 above and below the thermochromic transition temperature Tc Figure 3 X-ray powder diffraction patterns for V0.99Mo0.01O2 against temperature. (Change from VO2(M) - VO2(R)) Two factors were found to affect the transition temperature – the presence of a secondary dopant element and the thickness/ strain in the films. For films of thickness less than 250 nm some slight reduction in Tc was noted – this was attributed to a strain effect11. Introduction of Mo, Nb, Ta and W into the VO2 host lattice in the form of a solid solution gave a reduction in 0.2 Direction of heating 0.18 0.16 Reflectance-Transmittance 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 350 550 750 950 1150 1350 1550 1750 1950 23.8 Wavelength / nm Transmittance Tc Transmittance >Tc Direction of cooling Reflectance