- Email: [email protected]
Templated growth of smart coatings: Hybrid chemical vapour deposition of vanadyl acetylacetonate with tetraoctyl ammonium bromide
Applied Surface Science 255 (2009) 7291–7295
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Templated growth of smart coatings: Hybrid chemical vapour deposition of vanadyl acetylacetonate with tetraoctyl ammonium bromide Manfredi Saeli a, Russell Binions b,*, Clara Piccirillo b, Ivan P. Parkin b a b
Universita` degli Studi di Palermo, Dipartimento di Progetto e Costruzione Edilizia (DPCE), Viale delle Scienze 90128 Palermo, Italy Christopher Ingold Laboratories, Department of Chemistry, University College London, 20 Gordon Street, London, WC1H 0AJ, United Kingdom
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 January 2009 Received in revised form 24 March 2009 Accepted 24 March 2009 Available online 1 April 2009
Hybrid aerosol assisted and atmospheric pressure chemical vapour deposition methodology has been utilised to produce thin films of vanadium dioxide from vanadyl acetylacetonate. Tetraoctyl ammonium bromide (TOAB) was used in the aerosol precursor solution. The films were analysed by X-ray diffraction, scanning electron microscopy (SEM) and X-ray photoelectron spectroscopy. Their optical and thermochromic behaviour was also determined. It was found that the use of TOAB had a templating effect that led to a halving in the particle size and that this consequently led to a significant decrease in the thermochromic transition temperature of the films to 34 8C. Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved.
Keywords: CVD Vanadium dioxide Thermochromism Aerosol
1. Introduction Vanadium dioxide is a material that has shown potential for use as an intelligent glazing material [1,2]. The material has a metal to semiconductor transition (MST) where there is a structural change from a higher temperature rutile structure to a lower temperature monoclinic structure [3]. This structural transition results in a change in electrical conductivity and optical properties. The rutile material is metallic and reflects a wide range of solar radiation; whereas the monoclinic phase is a semiconductor and generally transparent to solar radiation. The temperature this phase change occurs as is 68 8C in single crystal vanadium dioxide [3], it is considered that in order for vanadium dioxide to be a useful material in intelligent glazing this transition temperature needs to be reduced to around 25 8C [4]. Work has been conducted in recent years to try and reduce the transition temperature of the vanadium dioxide MST by adding dopants [5,6], of which tungsten has been the most successful dopant; reducing the MST by 20–25 8C per at.% of tungsten dopant [7]. There are however a variety of problems with adding dopants to the films. Results on the effect of tungsten and fluorine dopants show that fluorine doping increased the transmission of the metallic state thus diminishing the overall level of switch and
* Corresponding author. Tel.: +44 20 7679 1460; fax: +44 20 7679 7463. E-mail address: [email protected] (R. Binions).
subsequent efficiency of the film [8]. The same effect is seen in the case of tungsten doping where the transmission of the semiconducting state is reduced also lowering the overall switch [9]. In order to have as large a switch as possible the transition temperature must be reduced in an alternative manner. It has been demonstrated that by adding strain into a VO2 film the transition temperature may also be reduced. This may be done by substrate matching [10], using sub 50 nm thin films (in physical vapour deposition processes) [11] or by introducing preferential orientation through the careful choice of chemical vapour deposition (CVD) growth conditions [12]. Surfactant templates have been used in CVD previously to control the size of deposited gold nanoparticles in both aerosol assisted CVD (AACVD) and the hybrid system [13–15]. In each case, the surfactant is added to the precursor solution and mixed before use. In other examples a template is printed onto the substrate prior to deposition [16] or the surface is saturated by gaseous exposure before the main deposition process begins [17]. A hybrid chemical vapour deposition methodology has previously been used to deposit thin films of gold nanoparticle doped vanadium dioxide from the chemical vapour deposition reaction of vanadyl acetylacetonate and auric acid in methanol [15]. This hybrid technique shows great potential as the film characteristics are similar to those produced by atmospheric pressure chemical vapour deposition (APCVD – good adhesion, uniformity and coverage) but with the precursor versatility afforded by AACVD. Here we report on depositions carried out using this hybrid methodology, vanadyl acetylacetonate and the
0169-4332/$ – see front matter . Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2009.03.083
7292
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295
surfactant molecule, tetraoctyl ammonium bromide (TOAB), to directly influence film growth. It was found that the use of TOAB in the hybrid reaction lead to template affects where the VO2 crystallite size was reduced; subsequently adding strain into the films and significantly reducing the metal semi-conductor transition temperature. 2. Experimental A 98% nitrogen, 2% oxygen mixture was obtained from the British Oxygen Company (BOC) and used as supplied in the plain line. Nitrogen (99.99%) was obtained from BOC and used as supplied in the bubbler lines. Coatings were obtained on SiO2 coated float glass. Combined AA/AP CVD experiments were conducted on 150 mm 45 mm 3 mm pieces of glass using a flat-bed cold-walled CVD reactor. The glass was cleaned prior to use by washing with petroleum ether (60–80 8C) and isopropanol and then dried in air. A graphite block containing a Whatman cartridge heater was used to heat the glass substrate. The temperature of the substrate was monitored by a Pt–Rh thermocouple. Independent thermocouple measurements indicated that temperature gradients of up to 50 8C cm 1 were observable at 600 8C across the surface of the glass. The rig was designed so that four independent gas lines could be used. All gas handling lines, regulators and flow valves were made of stainless steel and were 6.5 mm internal diameter except for the inlet to the mixing chamber and the exhaust line from the apparatus that were 13 mm in diameter. In these experiments three gas lines were used. Gases came directly from a cylinder and were preheated by passing along 2 m lengths of stainless steel tubing, which were curled and inserted inside a tube furnace. The temperatures of all the gas inlet lines were monitored by Pt–Rh thermocouples and Eurotherm heat controllers. Vanadyl acetylacetonate (99.99%) – [VO(acac)2] – was obtained from Aldrich and used without further purification and was placed into a stainless steel bubbler. The bubbler was heated to 175 8C by a heating jacket and vanadyl acetylacetonate introduced into the gas streams by passing hot nitrogen gas through the bubbler. Tetraoctyl ammonium bromide was obtained from Aldrich and used without further purification. In deposition experiments using TOAB, between 1 g and 0.1 g was added to the reaction flask and stirred for 10 min prior to the deposition experiment. An aerosol was generated at room temperature by use of a PIFCO air humidifier. Nitrogen was passed through the aerosol mist, thus forcing the aerosol particles encapsulated with precursor into the heated reaction chamber. The two components of the system were transported using stainless steel pipes 13 mm internal diameter respectively. The pipes were attached directly to the reaction chamber of the coater. Gas flows were adjusted using suitable regulators and flow controllers. The exhaust from the reactor was vented directly into the extraction system of a fume cupboard. This deposition set up has been described in more detail previously [15]. All of the apparatus was baked out with nitrogen at 150 8C for 30 min before use. Deposition experiments were conducted by heating the horizontal-bed reactor and the bubblers to the desired temperatures before diverting the nitrogen line through the bubbler and hence to the reactor. Deposition experiments were timed by use of a stopwatch and were conducted for typically 15 min. The maximum possible deposition temperature with this equipment was 600 8C. At the end of the deposition only nitrogen was allowed to flow over the glass substrate until the substrate had sufficiently cooled to handle (60 8C). Samples were handled and stored in air. Substrate temperature for deposition was kept constant at 525 8C as this has been found to be the optimal temperature to grow
monoclinic vanadium dioxide thin films from vanadyl acetylacetonate. Due to the nature of the graphite heating block there are subtle changes in thickness that correlate with the temperature gradient across the substrate however, a highly uniform area 2 cm 5 cm in the middle of the substrate is always observed, it is this area that is referred to when discussing the uniformity of the films. Electron microprobe analysis was obtained on a JEOL EMA and referenced against vanadium and oxygen standards. Energy dispersive analysis of X-rays (EDAX) and wavelength dispersive analysis of X-rays (WDAX) were conducted using a Phillips XL30 ESEM instrument. Scanning electron microscopy (SEM) images were acquired on a Jeol 6301F field emission instrument. X-ray diffraction (XRD) patterns were measured on a Bruker Gadds D8 diffractometer using monochromated (CuKa1+2) radiation in the reflection mode using a glancing incident angle of 58. X-ray photoelectron spectra were recorded with a VG ESCALAB 220I XL instrument using a focused (300 mm spot) monochromatic AlKa radiation at a pass energy of 20 eV. Scans were acquired with steps of 50 meV. A flood gun was used to control charging and the binding energies were referenced to an adventitious C 1s peak at 284.6 eV. Depth profiling measurements were obtained by using argon beam sputtering. Reflectance and transmission spectra were recorded between 300 nm and 2500 nm on a PerkinElmer Lambda 950 UV–vis spectrometer. UV–vis spectra were obtained using a Helios double beam instrument. Raman spectra were acquired on a Renishaw Raman system 1000 using a helium-neon laser of wavelength 632.8 nm. The Raman system was calibrated against the emission lines of neon. Transmittance-temperature studies (transition temperature and hysteresis width measurements) were performed on a PerkinElmer 457 grating spectrometer set to 4000 cm 1. An aluminium temperature cell controlled by RS resistive heaters, Eurotherm temperature controllers and k-type thermocouples was used to manipulate sample temperature. Sample temperature was measured using a k-type thermocouple taped directly onto the film surface. Film thickness was measured directly by scanning electron microscopy then correlated with EDAX and optical transmission data. 3. Results VO2 films were grown by the use of hybrid aerosol assisted atmospheric pressure CVD from TOAB in methanol and [VO(acac)2]. The films showed good surface coverage, uniformity and reproducibility. Film thickness could be easily varied by increasing or decreasing the time of deposition. In all cases at least the first 75% of the substrate is covered, similar to that observed previously with VO2 films produced from the APCVD
Table 1 Summary of experimental conditions and chemical analysis. Flow conditions and deposition time were constant for all experiments; plain flow = 2 L min 1, VO(acac)2 bubbler flow = 4 L min 1, aerosol flow = 1 L min 1 and deposition time 15 min; content of AACVD flask: 25 ml of methanol with variable amount of TOAB as shown in the table. Sample
TOAB amount (g)
EDAX/WDAX/XRD phase
lmax in visible transmission (film colour) (nm)
1 2 3 4 5 6
Not used 1.00 0.50 0.25 0.12 0.06
VO2 VO2 VO2 VO2 VO2 VO2
569 568 566 570 566 567
(monoclinic) (monoclinic) (monoclinic) (monoclinic) (monoclinic) (monoclinic)
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295
7293
Fig. 1. Secondary electron scanning electron microscopy pictures of typical samples of a VO2 film produced using the hybrid AA/AP CVD methodology grown (A) with TOAB , (B) without TOAB.
Fig. 2. X-Ray Diffraction patterns of typical samples of a VO2 film produced using the hybrid AA/AP CVD methodology grown (A) with TOAB , (B) without TOAB.
reaction of vanadyl acetylacetonate [12]. Similarly there are changes in thickness that correlate with the temperature gradient across the substrate surface as seen previously [12] and a highly uniform area 2 cm 5 cm in the middle of the substrate. For the results reported here deposition time was 15 min, which corresponded to a film thickness of 110 10 nm for all samples. The films appeared yellow/brown in colour as has been observed previously. Adjusting the amount of TOAB in the reaction mixture did not alter the colour of the films significantly (lmax in visible transmission as shown in Table 1). The films were fairly adherent to the glass substrate; they could not be rubbed off by
abrasion with a tissue and were not marked by a brass stylus. The films also passed the Scotch tape test. The films could however be marked with a steel scalpel. These mechanical properties are identical to those VO2 films grown here without an AACVD component. Scanning electron microscopy (Fig. 1.) of the samples indicates an island growth morphology. Typically for samples produced with TOAB (Samples 2–6) a smaller particle size of 50 nm (Fig. 1A) is seen compared with 100 nm (Fig. 1B) for those produced without. Analysis with energy dispersive analysis of X-rays indicated that the films contained oxygen and vanadium only, with no contaminant, at least to the limit of detection of the methodology (around 0.5 at.% depending on the element). Data for carbon was not collected as the samples were coated in preparation for EDAX. X-ray diffraction analysis is shown in Fig. 2; because the films were quite thin, a large broad hump centred at 248 2u is seen which results from the amorphous nature of the glass substrate. The films grown with TOAB were less crystalline than those grown without; however a peak corresponding to VO2(monoclinic) could be detected unambiguosly. Annealing in nitrogen at 500 8C for 2 h did not improve the crystallinity of the VO2 phase in the films, however peaks relating to V2O5 appeared in the X-ray diffraction pattern. Previous work has shown that small amounts of V2O5 are present at the surface, but that this does not affect the thermochromic behaviour of the bulk film [4,7,12]. X-ray photoelectron analysis (XPS, Fig. 3A) of the sample surface suggests a variety of vanadium environments, we believe
Fig. 3. X-Ray photoelectron spectroscopy profiles of (A) the vanadium and oxygen region, (B) the carbon region of a typical of a typical VO2 film produced using the hybrid AA/ AP CVD methodology grown with TOAB both before (surface) and after sputtering. Each 30 s sputter corresponds to removal of approximately 10 nm of material.
7294
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295 Table 2 Summary of thermochromic behaviour. Flow conditions and deposition time were constant for all experiments; plain flow = 2 L min 1, VO(acac)2 bubbler flow = 4 L min 1, aerosol flow = 1 L min 1 and deposition time 15 min; content of AACVD flask: 25 ml of methanol with variable amount of TOAB as shown in the table. Sample
TOAB amount (g)
Transition temperature (8C)
Hysteresis width (8C)
1 2 3 4 5 6
Not used 1.00 0.50 0.25 0.12 0.06
52 38 35 43 34 36
15 13 14 10 8 10
Fig. 4. Raman spectroscopy pattern of a VO2 film produced using the hybrid AA/AP CVD methodology grown with TOAB.
V2O5 to be present only at the surface; indeed, on etching only a single vanadium environment is observed consistent with vanadium dioxide. XPS also indicated that carbon was present (1 at.%) in the bulk of the film (Fig. 3B) and not just at the surface as has been observed previously [4,7,12]. Raman spectroscopy (Fig. 4) indicated the presence of VO2 (m) with large peaks at 192 cm 1, 222 cm 1, 388 cm 1 and 611 cm 1. There were also broad peaks centred at 1371 cm 1and 1591 cm 1, which correspond to graphitic carbon. However, considering that the colour of the sample is similar to the ones grown without TOAB, it can be assumed that the amount of graphitic carbon is relatively low. As has been seen previously with un-doped and tungsten doped samples prepared by CVD a large change in transmission is observable at 2500 nm (Fig. 5.), typically 35–40% comparable with previous literature values.[4,7,12] The films produced with by the hybrid methodology with TOAB (Fig. 5A) had virtually identical transmission spectra to films produced without the use of TOAB from the APCVD reaction of [VO(acac)2] (Fig. 5B). The reflectance spectra are different. The samples produced by the hybrid method had a markedly lower change in infrared reflectance above the MST. All of the films show thermochromic behaviour with reduced transition temperatures. Typically the hysteresis width (Table 2.) of vanadium dioxide thin films with is in the range of 10–15 8C. The films produced with TOAB in (Samples 2–6) had lower transition temperatures, typically 34–43 8C, than the films without TOAB (Sample 1) where the transition temperature was 52 8C.
4. Discussion The hybrid aerosol assisted atmospheric pressure chemical vapour deposition reaction of vanadyl acetylacetonate and TOAB lead to the production of thin films of monoclinic vanadium dioxide thin films. Within the range tested here the amount of TOAB added to the precursor flask did not affect the resulting properties of the deposited film, which were similar for all films produced using TOAB. The films produced from the APCVD reaction of [VO(acac)2] had significantly different thermochromic properties. The microstructure of the films is significantly altered. In depositions without TOAB the particle size of the films is around 100 nm and there is significant agglomeration (Fig. 1B). The use of TOAB reduces particle size to around 50 nm and leads to a more rounded island (Fig. 1A). It is known that TOAB causes a templating effect for gold nanoparticles that has been observed previously [13,14]. It was not expected that the use of TOAB would cause a change in the overall film morphology. It is unlikely that a gas phase reaction is taking place. If a gas phase reaction were taking place we would anticipate finding excess material in the exhaust of our deposition set up and evidence on the surface of film such as pin hole defects or similar [18]. Neither case occurs. Additionally the TOAB is transported to the reactor in an aerosol mist separately from the vanadium precursor and is mixed in the reactor making the likelihood of a gas phase reaction occurring smaller. Therefore we suggest that a surface reaction is occurring where TOAB molecules coordinate to growing vanadium dioxide particles on the substrate surface effectively capping their growth. This
Fig. 5. Transmission and Reflection Data for typical samples of a VO2 film produced using the hybrid AA/AP CVD methodology grown (A) with TOAB, (B) without TOAB.
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295
would explain the smaller particle size of the vanadium dioxide and TOAB films as well as the diminished level of agglomeration. In all cases a large change in transmission is observable at 2500 nm (Fig. 5), typically 30–40% comparable with previous literature values [4,7,12]. The reflectance spectra of films produced with TOAB (Fig. 5A) show a smaller change of 10–15% on undergoing the MST. This is most likely to be as a result of the change in film morphology (Fig. 1) caused by the use of TOAB in the deposition reaction. All of the films show thermochromic behaviour with reduced transition temperatures. Typically the hysteresis width (Table 2) vanadium dioxide thin films produced under these experimental condition is within the range of 10–15 8C [4,7,12] and comparable to monoclinic vanadium oxide thin films produced using a sol–gel methodology [11]. This is less than gold nanoparticle doped vanadium dioxide films produced previously using the hybrid method where hysteresis widths were slightly larger, around 15– 20 8C [15]. The hysteresis width seemed to be independent of the amount of TOAB used in the deposition of the film (Table 2). Ideally the hysteresis width would be as narrow as possible in order to maximise the energy saving effect of the coating. Hysteresis width has been attributed to several factors such as a crystallographic orientation [12] and grain size; [19] in this instance however the hysteresis width is similar in both samples produced with TOAB and without (Table 2) despite their being a significant difference in particle size and morphology (Fig. 1). This suggests that this feature is less important in determining the hysteresis width than it has been previously thought. As shown in Table 2, the thermochromic transition temperature is reduced to 50 8C for samples grown without the use of TOAB and to as low as 34 8C with films grown with TOAB. This is independent of the amount of TOAB used in the deposition reaction. It is likely that this reduction for the films grown without TOAB is caused by strain as a result of preferential orientation as observed previously [12,15], although this is difficult to quantify as the films are thin and poorly crystalline. The additional reduction of the transition temperature to 34 8C seen in films produced by TOAB is harder to explain. The average particle and island size is smaller for the films produced with TOAB (Fig. 1); as such the particles have a higher surface to bulk ratio and this may lead to higher strain in the films. However, this is difficult to evaluate in the absence of quality X-ray diffraction data. It is possible that carbon doping is taking place, as carbon is found throughout the film (XPS, Fig. 3); we consider this unlikely as this effect has never been observed in any previous experiments conducted within our laboratories and carbon doping of vanadium dioxide has never been reported in the literature. The effect of crystallographic and morphologically induced strain, on the other hand, has been observed by several groups [10–12]. TOAB has been used previously to template gold nanoparticles in AACVD [13,14] causing a surfactant effect in solution before aerosol generation. The effect was to direct the size and shape of the nanoparticles. In this instance the TOAB is influencing the growth of vanadium dioxide particles. As stated above, we believe this is most likely to be taking place on the substrate surface rather than in the gas phase. The use of TOAB in hybrid aerosol assisted atmospheric pressure chemical vapour deposition reactions leads to several interesting effects. There is a change in morphology of the overall film that is unexpected and can be potentially exploited to enhance the properties of the film. We anticipate that the use of
7295
a different surfactant will lead to different morphological effects. Despite the films being less crystalline than the one produced without TOAB, the change in the transmittance with the temperature is quite big; this makes them suitable films to be used as intelligent material. The use of TOAB also leads to an additional reduction on the thermochromic transition temperature, this is attributed to the introduction of additional strain in the film caused by this template effect. To our knowledge, this surfactant effect has never reported before. 5. Conclusion The use of hybrid aerosol assisted and atmospheric pressure chemical vapour deposition methodology has afforded thin films of monoclinic vanadium dioxide. The use of TOAB in the aerosol solution has been shown to have an effect on the films properties. In fact the TOAB was found to template the growth of the film; the particle size was halved from 100 nm to 50 nm. This smaller particle size led to additional strain in the film and caused a further reduction in the thermochromic transition temperature from 52 8C to 34 8C compared to films grown under the same conditions but without the use of the surfactant TOAB. These results, never reported before, show how the morphology and the properties of VO2 can be tailored not just using dopants but with the use of a surfactant agent. This opens new possibilities for the development of this material and its application as intelligent window coating. Acknowledgments Pilkington Glass are thanked for the provision of glass substrates. RB thanks the Royal Society for a Dorothy Hodgkin Fellowship. IPP thanks the Wolfson trust for a merit award. Electron microscopy technician Mr. Kevin Reeves is thanked for invaluable assistance with electron microscopy. Manfredi Saeli thanks the University of Palermo ‘‘Esperto di Nanotecnologie per i Beni Culturali’’ for funding. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
C.G. Granqvist, Thin Solid Films 193 (1990) 730–740. C.G. Granqvist, Adv. Mater. 15 (2003) 1789–1803. K.D. Rogers, Powder Diffract 8 (1993) 240–244. I.P. Parkin, R. Binions, C. Piccirillo, C.S. Blackman, T.D. Manning, J. Nano Res. 2 (2008) 1–20. T.E. Phillips, R.A. Murray, T.D. Poelher, Mater. Res. Bull. 22 (1987) 1113–1123. F. Beteille, R. Morineau, J. Livage, M. Nagano, Mater. Res. Bull. 32 (1997) 1109– 1117. T.D. Manning, I.P. Parkin, M.E. Pemble, D.W. Sheel, D. Vernardou, Chem. Mater. 16 (2004) 744–749. W. Burkhardt, T. Christmann, B.K. Meyer, W. Niessner, D. Schalch, A. Scharmann, Thin Solid Films 345 (1999) 229–235. W. Burkhardt, T. Christmann, S. Franke, W. Kriegseis, D. Meister, B.K. Meyer, W. Niessner, D. Schalch, A. Scharmann, Thin Solid Films 402 (2002) 226–231. F.C. Case, J. Vac. Sci. Technol. A 2 (1984) 1509–1512. E. Cavanna, J.P. Segaud, J. Livage, Mater. Res. Bull. 34 (1999) 167–177. R. Binions, C. Piccirillo, G. Hyett, I.P. Parkin, J. Mater. Chem. 17 (2007) 4652–4660. R.G. Palgrave, I.P. Parkin, J. Am. Chem. Soc. 128 (2006) 1587–1597. R.G. Palgrave, I.P. Parkin, Chem. Mater. 19 (2007) 4639–4647. R. Binions, I.P. Parkin, C. Piccirillo, C.J. Carmalt, R.G. Palgrave, Chem. Vap. Depos. 1–2 (2008) 33–39. C. Bae, S. Kim, B. Ahn, J. Kim, M. Sungc, H. Shin, J. Mater. Chem. 18 (2008) 1362– 1367. E.S. Hwang, J. Lee, Chem. Mater. 12 (2000) 2076–2081. R. Binions, C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, Polyhedron 21 (2002) 1943–1947. F. Beleille, L. Mazerolles, J. Livage, Mater. Res. Bull. 34 (1999) 2177–2184.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Templated growth of smart coatings: Hybrid chemical vapour deposition of vanadyl acetylacetonate with tetraoctyl ammonium bromide Manfredi Saeli a, Russell Binions b,*, Clara Piccirillo b, Ivan P. Parkin b a b
Universita` degli Studi di Palermo, Dipartimento di Progetto e Costruzione Edilizia (DPCE), Viale delle Scienze 90128 Palermo, Italy Christopher Ingold Laboratories, Department of Chemistry, University College London, 20 Gordon Street, London, WC1H 0AJ, United Kingdom
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 January 2009 Received in revised form 24 March 2009 Accepted 24 March 2009 Available online 1 April 2009
Hybrid aerosol assisted and atmospheric pressure chemical vapour deposition methodology has been utilised to produce thin films of vanadium dioxide from vanadyl acetylacetonate. Tetraoctyl ammonium bromide (TOAB) was used in the aerosol precursor solution. The films were analysed by X-ray diffraction, scanning electron microscopy (SEM) and X-ray photoelectron spectroscopy. Their optical and thermochromic behaviour was also determined. It was found that the use of TOAB had a templating effect that led to a halving in the particle size and that this consequently led to a significant decrease in the thermochromic transition temperature of the films to 34 8C. Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved.
Keywords: CVD Vanadium dioxide Thermochromism Aerosol
1. Introduction Vanadium dioxide is a material that has shown potential for use as an intelligent glazing material [1,2]. The material has a metal to semiconductor transition (MST) where there is a structural change from a higher temperature rutile structure to a lower temperature monoclinic structure [3]. This structural transition results in a change in electrical conductivity and optical properties. The rutile material is metallic and reflects a wide range of solar radiation; whereas the monoclinic phase is a semiconductor and generally transparent to solar radiation. The temperature this phase change occurs as is 68 8C in single crystal vanadium dioxide [3], it is considered that in order for vanadium dioxide to be a useful material in intelligent glazing this transition temperature needs to be reduced to around 25 8C [4]. Work has been conducted in recent years to try and reduce the transition temperature of the vanadium dioxide MST by adding dopants [5,6], of which tungsten has been the most successful dopant; reducing the MST by 20–25 8C per at.% of tungsten dopant [7]. There are however a variety of problems with adding dopants to the films. Results on the effect of tungsten and fluorine dopants show that fluorine doping increased the transmission of the metallic state thus diminishing the overall level of switch and
* Corresponding author. Tel.: +44 20 7679 1460; fax: +44 20 7679 7463. E-mail address: [email protected] (R. Binions).
subsequent efficiency of the film [8]. The same effect is seen in the case of tungsten doping where the transmission of the semiconducting state is reduced also lowering the overall switch [9]. In order to have as large a switch as possible the transition temperature must be reduced in an alternative manner. It has been demonstrated that by adding strain into a VO2 film the transition temperature may also be reduced. This may be done by substrate matching [10], using sub 50 nm thin films (in physical vapour deposition processes) [11] or by introducing preferential orientation through the careful choice of chemical vapour deposition (CVD) growth conditions [12]. Surfactant templates have been used in CVD previously to control the size of deposited gold nanoparticles in both aerosol assisted CVD (AACVD) and the hybrid system [13–15]. In each case, the surfactant is added to the precursor solution and mixed before use. In other examples a template is printed onto the substrate prior to deposition [16] or the surface is saturated by gaseous exposure before the main deposition process begins [17]. A hybrid chemical vapour deposition methodology has previously been used to deposit thin films of gold nanoparticle doped vanadium dioxide from the chemical vapour deposition reaction of vanadyl acetylacetonate and auric acid in methanol [15]. This hybrid technique shows great potential as the film characteristics are similar to those produced by atmospheric pressure chemical vapour deposition (APCVD – good adhesion, uniformity and coverage) but with the precursor versatility afforded by AACVD. Here we report on depositions carried out using this hybrid methodology, vanadyl acetylacetonate and the
0169-4332/$ – see front matter . Crown Copyright ß 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2009.03.083
7292
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295
surfactant molecule, tetraoctyl ammonium bromide (TOAB), to directly influence film growth. It was found that the use of TOAB in the hybrid reaction lead to template affects where the VO2 crystallite size was reduced; subsequently adding strain into the films and significantly reducing the metal semi-conductor transition temperature. 2. Experimental A 98% nitrogen, 2% oxygen mixture was obtained from the British Oxygen Company (BOC) and used as supplied in the plain line. Nitrogen (99.99%) was obtained from BOC and used as supplied in the bubbler lines. Coatings were obtained on SiO2 coated float glass. Combined AA/AP CVD experiments were conducted on 150 mm 45 mm 3 mm pieces of glass using a flat-bed cold-walled CVD reactor. The glass was cleaned prior to use by washing with petroleum ether (60–80 8C) and isopropanol and then dried in air. A graphite block containing a Whatman cartridge heater was used to heat the glass substrate. The temperature of the substrate was monitored by a Pt–Rh thermocouple. Independent thermocouple measurements indicated that temperature gradients of up to 50 8C cm 1 were observable at 600 8C across the surface of the glass. The rig was designed so that four independent gas lines could be used. All gas handling lines, regulators and flow valves were made of stainless steel and were 6.5 mm internal diameter except for the inlet to the mixing chamber and the exhaust line from the apparatus that were 13 mm in diameter. In these experiments three gas lines were used. Gases came directly from a cylinder and were preheated by passing along 2 m lengths of stainless steel tubing, which were curled and inserted inside a tube furnace. The temperatures of all the gas inlet lines were monitored by Pt–Rh thermocouples and Eurotherm heat controllers. Vanadyl acetylacetonate (99.99%) – [VO(acac)2] – was obtained from Aldrich and used without further purification and was placed into a stainless steel bubbler. The bubbler was heated to 175 8C by a heating jacket and vanadyl acetylacetonate introduced into the gas streams by passing hot nitrogen gas through the bubbler. Tetraoctyl ammonium bromide was obtained from Aldrich and used without further purification. In deposition experiments using TOAB, between 1 g and 0.1 g was added to the reaction flask and stirred for 10 min prior to the deposition experiment. An aerosol was generated at room temperature by use of a PIFCO air humidifier. Nitrogen was passed through the aerosol mist, thus forcing the aerosol particles encapsulated with precursor into the heated reaction chamber. The two components of the system were transported using stainless steel pipes 13 mm internal diameter respectively. The pipes were attached directly to the reaction chamber of the coater. Gas flows were adjusted using suitable regulators and flow controllers. The exhaust from the reactor was vented directly into the extraction system of a fume cupboard. This deposition set up has been described in more detail previously [15]. All of the apparatus was baked out with nitrogen at 150 8C for 30 min before use. Deposition experiments were conducted by heating the horizontal-bed reactor and the bubblers to the desired temperatures before diverting the nitrogen line through the bubbler and hence to the reactor. Deposition experiments were timed by use of a stopwatch and were conducted for typically 15 min. The maximum possible deposition temperature with this equipment was 600 8C. At the end of the deposition only nitrogen was allowed to flow over the glass substrate until the substrate had sufficiently cooled to handle (60 8C). Samples were handled and stored in air. Substrate temperature for deposition was kept constant at 525 8C as this has been found to be the optimal temperature to grow
monoclinic vanadium dioxide thin films from vanadyl acetylacetonate. Due to the nature of the graphite heating block there are subtle changes in thickness that correlate with the temperature gradient across the substrate however, a highly uniform area 2 cm 5 cm in the middle of the substrate is always observed, it is this area that is referred to when discussing the uniformity of the films. Electron microprobe analysis was obtained on a JEOL EMA and referenced against vanadium and oxygen standards. Energy dispersive analysis of X-rays (EDAX) and wavelength dispersive analysis of X-rays (WDAX) were conducted using a Phillips XL30 ESEM instrument. Scanning electron microscopy (SEM) images were acquired on a Jeol 6301F field emission instrument. X-ray diffraction (XRD) patterns were measured on a Bruker Gadds D8 diffractometer using monochromated (CuKa1+2) radiation in the reflection mode using a glancing incident angle of 58. X-ray photoelectron spectra were recorded with a VG ESCALAB 220I XL instrument using a focused (300 mm spot) monochromatic AlKa radiation at a pass energy of 20 eV. Scans were acquired with steps of 50 meV. A flood gun was used to control charging and the binding energies were referenced to an adventitious C 1s peak at 284.6 eV. Depth profiling measurements were obtained by using argon beam sputtering. Reflectance and transmission spectra were recorded between 300 nm and 2500 nm on a PerkinElmer Lambda 950 UV–vis spectrometer. UV–vis spectra were obtained using a Helios double beam instrument. Raman spectra were acquired on a Renishaw Raman system 1000 using a helium-neon laser of wavelength 632.8 nm. The Raman system was calibrated against the emission lines of neon. Transmittance-temperature studies (transition temperature and hysteresis width measurements) were performed on a PerkinElmer 457 grating spectrometer set to 4000 cm 1. An aluminium temperature cell controlled by RS resistive heaters, Eurotherm temperature controllers and k-type thermocouples was used to manipulate sample temperature. Sample temperature was measured using a k-type thermocouple taped directly onto the film surface. Film thickness was measured directly by scanning electron microscopy then correlated with EDAX and optical transmission data. 3. Results VO2 films were grown by the use of hybrid aerosol assisted atmospheric pressure CVD from TOAB in methanol and [VO(acac)2]. The films showed good surface coverage, uniformity and reproducibility. Film thickness could be easily varied by increasing or decreasing the time of deposition. In all cases at least the first 75% of the substrate is covered, similar to that observed previously with VO2 films produced from the APCVD
Table 1 Summary of experimental conditions and chemical analysis. Flow conditions and deposition time were constant for all experiments; plain flow = 2 L min 1, VO(acac)2 bubbler flow = 4 L min 1, aerosol flow = 1 L min 1 and deposition time 15 min; content of AACVD flask: 25 ml of methanol with variable amount of TOAB as shown in the table. Sample
TOAB amount (g)
EDAX/WDAX/XRD phase
lmax in visible transmission (film colour) (nm)
1 2 3 4 5 6
Not used 1.00 0.50 0.25 0.12 0.06
VO2 VO2 VO2 VO2 VO2 VO2
569 568 566 570 566 567
(monoclinic) (monoclinic) (monoclinic) (monoclinic) (monoclinic) (monoclinic)
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295
7293
Fig. 1. Secondary electron scanning electron microscopy pictures of typical samples of a VO2 film produced using the hybrid AA/AP CVD methodology grown (A) with TOAB , (B) without TOAB.
Fig. 2. X-Ray Diffraction patterns of typical samples of a VO2 film produced using the hybrid AA/AP CVD methodology grown (A) with TOAB , (B) without TOAB.
reaction of vanadyl acetylacetonate [12]. Similarly there are changes in thickness that correlate with the temperature gradient across the substrate surface as seen previously [12] and a highly uniform area 2 cm 5 cm in the middle of the substrate. For the results reported here deposition time was 15 min, which corresponded to a film thickness of 110 10 nm for all samples. The films appeared yellow/brown in colour as has been observed previously. Adjusting the amount of TOAB in the reaction mixture did not alter the colour of the films significantly (lmax in visible transmission as shown in Table 1). The films were fairly adherent to the glass substrate; they could not be rubbed off by
abrasion with a tissue and were not marked by a brass stylus. The films also passed the Scotch tape test. The films could however be marked with a steel scalpel. These mechanical properties are identical to those VO2 films grown here without an AACVD component. Scanning electron microscopy (Fig. 1.) of the samples indicates an island growth morphology. Typically for samples produced with TOAB (Samples 2–6) a smaller particle size of 50 nm (Fig. 1A) is seen compared with 100 nm (Fig. 1B) for those produced without. Analysis with energy dispersive analysis of X-rays indicated that the films contained oxygen and vanadium only, with no contaminant, at least to the limit of detection of the methodology (around 0.5 at.% depending on the element). Data for carbon was not collected as the samples were coated in preparation for EDAX. X-ray diffraction analysis is shown in Fig. 2; because the films were quite thin, a large broad hump centred at 248 2u is seen which results from the amorphous nature of the glass substrate. The films grown with TOAB were less crystalline than those grown without; however a peak corresponding to VO2(monoclinic) could be detected unambiguosly. Annealing in nitrogen at 500 8C for 2 h did not improve the crystallinity of the VO2 phase in the films, however peaks relating to V2O5 appeared in the X-ray diffraction pattern. Previous work has shown that small amounts of V2O5 are present at the surface, but that this does not affect the thermochromic behaviour of the bulk film [4,7,12]. X-ray photoelectron analysis (XPS, Fig. 3A) of the sample surface suggests a variety of vanadium environments, we believe
Fig. 3. X-Ray photoelectron spectroscopy profiles of (A) the vanadium and oxygen region, (B) the carbon region of a typical of a typical VO2 film produced using the hybrid AA/ AP CVD methodology grown with TOAB both before (surface) and after sputtering. Each 30 s sputter corresponds to removal of approximately 10 nm of material.
7294
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295 Table 2 Summary of thermochromic behaviour. Flow conditions and deposition time were constant for all experiments; plain flow = 2 L min 1, VO(acac)2 bubbler flow = 4 L min 1, aerosol flow = 1 L min 1 and deposition time 15 min; content of AACVD flask: 25 ml of methanol with variable amount of TOAB as shown in the table. Sample
TOAB amount (g)
Transition temperature (8C)
Hysteresis width (8C)
1 2 3 4 5 6
Not used 1.00 0.50 0.25 0.12 0.06
52 38 35 43 34 36
15 13 14 10 8 10
Fig. 4. Raman spectroscopy pattern of a VO2 film produced using the hybrid AA/AP CVD methodology grown with TOAB.
V2O5 to be present only at the surface; indeed, on etching only a single vanadium environment is observed consistent with vanadium dioxide. XPS also indicated that carbon was present (1 at.%) in the bulk of the film (Fig. 3B) and not just at the surface as has been observed previously [4,7,12]. Raman spectroscopy (Fig. 4) indicated the presence of VO2 (m) with large peaks at 192 cm 1, 222 cm 1, 388 cm 1 and 611 cm 1. There were also broad peaks centred at 1371 cm 1and 1591 cm 1, which correspond to graphitic carbon. However, considering that the colour of the sample is similar to the ones grown without TOAB, it can be assumed that the amount of graphitic carbon is relatively low. As has been seen previously with un-doped and tungsten doped samples prepared by CVD a large change in transmission is observable at 2500 nm (Fig. 5.), typically 35–40% comparable with previous literature values.[4,7,12] The films produced with by the hybrid methodology with TOAB (Fig. 5A) had virtually identical transmission spectra to films produced without the use of TOAB from the APCVD reaction of [VO(acac)2] (Fig. 5B). The reflectance spectra are different. The samples produced by the hybrid method had a markedly lower change in infrared reflectance above the MST. All of the films show thermochromic behaviour with reduced transition temperatures. Typically the hysteresis width (Table 2.) of vanadium dioxide thin films with is in the range of 10–15 8C. The films produced with TOAB in (Samples 2–6) had lower transition temperatures, typically 34–43 8C, than the films without TOAB (Sample 1) where the transition temperature was 52 8C.
4. Discussion The hybrid aerosol assisted atmospheric pressure chemical vapour deposition reaction of vanadyl acetylacetonate and TOAB lead to the production of thin films of monoclinic vanadium dioxide thin films. Within the range tested here the amount of TOAB added to the precursor flask did not affect the resulting properties of the deposited film, which were similar for all films produced using TOAB. The films produced from the APCVD reaction of [VO(acac)2] had significantly different thermochromic properties. The microstructure of the films is significantly altered. In depositions without TOAB the particle size of the films is around 100 nm and there is significant agglomeration (Fig. 1B). The use of TOAB reduces particle size to around 50 nm and leads to a more rounded island (Fig. 1A). It is known that TOAB causes a templating effect for gold nanoparticles that has been observed previously [13,14]. It was not expected that the use of TOAB would cause a change in the overall film morphology. It is unlikely that a gas phase reaction is taking place. If a gas phase reaction were taking place we would anticipate finding excess material in the exhaust of our deposition set up and evidence on the surface of film such as pin hole defects or similar [18]. Neither case occurs. Additionally the TOAB is transported to the reactor in an aerosol mist separately from the vanadium precursor and is mixed in the reactor making the likelihood of a gas phase reaction occurring smaller. Therefore we suggest that a surface reaction is occurring where TOAB molecules coordinate to growing vanadium dioxide particles on the substrate surface effectively capping their growth. This
Fig. 5. Transmission and Reflection Data for typical samples of a VO2 film produced using the hybrid AA/AP CVD methodology grown (A) with TOAB, (B) without TOAB.
M. Saeli et al. / Applied Surface Science 255 (2009) 7291–7295
would explain the smaller particle size of the vanadium dioxide and TOAB films as well as the diminished level of agglomeration. In all cases a large change in transmission is observable at 2500 nm (Fig. 5), typically 30–40% comparable with previous literature values [4,7,12]. The reflectance spectra of films produced with TOAB (Fig. 5A) show a smaller change of 10–15% on undergoing the MST. This is most likely to be as a result of the change in film morphology (Fig. 1) caused by the use of TOAB in the deposition reaction. All of the films show thermochromic behaviour with reduced transition temperatures. Typically the hysteresis width (Table 2) vanadium dioxide thin films produced under these experimental condition is within the range of 10–15 8C [4,7,12] and comparable to monoclinic vanadium oxide thin films produced using a sol–gel methodology [11]. This is less than gold nanoparticle doped vanadium dioxide films produced previously using the hybrid method where hysteresis widths were slightly larger, around 15– 20 8C [15]. The hysteresis width seemed to be independent of the amount of TOAB used in the deposition of the film (Table 2). Ideally the hysteresis width would be as narrow as possible in order to maximise the energy saving effect of the coating. Hysteresis width has been attributed to several factors such as a crystallographic orientation [12] and grain size; [19] in this instance however the hysteresis width is similar in both samples produced with TOAB and without (Table 2) despite their being a significant difference in particle size and morphology (Fig. 1). This suggests that this feature is less important in determining the hysteresis width than it has been previously thought. As shown in Table 2, the thermochromic transition temperature is reduced to 50 8C for samples grown without the use of TOAB and to as low as 34 8C with films grown with TOAB. This is independent of the amount of TOAB used in the deposition reaction. It is likely that this reduction for the films grown without TOAB is caused by strain as a result of preferential orientation as observed previously [12,15], although this is difficult to quantify as the films are thin and poorly crystalline. The additional reduction of the transition temperature to 34 8C seen in films produced by TOAB is harder to explain. The average particle and island size is smaller for the films produced with TOAB (Fig. 1); as such the particles have a higher surface to bulk ratio and this may lead to higher strain in the films. However, this is difficult to evaluate in the absence of quality X-ray diffraction data. It is possible that carbon doping is taking place, as carbon is found throughout the film (XPS, Fig. 3); we consider this unlikely as this effect has never been observed in any previous experiments conducted within our laboratories and carbon doping of vanadium dioxide has never been reported in the literature. The effect of crystallographic and morphologically induced strain, on the other hand, has been observed by several groups [10–12]. TOAB has been used previously to template gold nanoparticles in AACVD [13,14] causing a surfactant effect in solution before aerosol generation. The effect was to direct the size and shape of the nanoparticles. In this instance the TOAB is influencing the growth of vanadium dioxide particles. As stated above, we believe this is most likely to be taking place on the substrate surface rather than in the gas phase. The use of TOAB in hybrid aerosol assisted atmospheric pressure chemical vapour deposition reactions leads to several interesting effects. There is a change in morphology of the overall film that is unexpected and can be potentially exploited to enhance the properties of the film. We anticipate that the use of
7295
a different surfactant will lead to different morphological effects. Despite the films being less crystalline than the one produced without TOAB, the change in the transmittance with the temperature is quite big; this makes them suitable films to be used as intelligent material. The use of TOAB also leads to an additional reduction on the thermochromic transition temperature, this is attributed to the introduction of additional strain in the film caused by this template effect. To our knowledge, this surfactant effect has never reported before. 5. Conclusion The use of hybrid aerosol assisted and atmospheric pressure chemical vapour deposition methodology has afforded thin films of monoclinic vanadium dioxide. The use of TOAB in the aerosol solution has been shown to have an effect on the films properties. In fact the TOAB was found to template the growth of the film; the particle size was halved from 100 nm to 50 nm. This smaller particle size led to additional strain in the film and caused a further reduction in the thermochromic transition temperature from 52 8C to 34 8C compared to films grown under the same conditions but without the use of the surfactant TOAB. These results, never reported before, show how the morphology and the properties of VO2 can be tailored not just using dopants but with the use of a surfactant agent. This opens new possibilities for the development of this material and its application as intelligent window coating. Acknowledgments Pilkington Glass are thanked for the provision of glass substrates. RB thanks the Royal Society for a Dorothy Hodgkin Fellowship. IPP thanks the Wolfson trust for a merit award. Electron microscopy technician Mr. Kevin Reeves is thanked for invaluable assistance with electron microscopy. Manfredi Saeli thanks the University of Palermo ‘‘Esperto di Nanotecnologie per i Beni Culturali’’ for funding. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
C.G. Granqvist, Thin Solid Films 193 (1990) 730–740. C.G. Granqvist, Adv. Mater. 15 (2003) 1789–1803. K.D. Rogers, Powder Diffract 8 (1993) 240–244. I.P. Parkin, R. Binions, C. Piccirillo, C.S. Blackman, T.D. Manning, J. Nano Res. 2 (2008) 1–20. T.E. Phillips, R.A. Murray, T.D. Poelher, Mater. Res. Bull. 22 (1987) 1113–1123. F. Beteille, R. Morineau, J. Livage, M. Nagano, Mater. Res. Bull. 32 (1997) 1109– 1117. T.D. Manning, I.P. Parkin, M.E. Pemble, D.W. Sheel, D. Vernardou, Chem. Mater. 16 (2004) 744–749. W. Burkhardt, T. Christmann, B.K. Meyer, W. Niessner, D. Schalch, A. Scharmann, Thin Solid Films 345 (1999) 229–235. W. Burkhardt, T. Christmann, S. Franke, W. Kriegseis, D. Meister, B.K. Meyer, W. Niessner, D. Schalch, A. Scharmann, Thin Solid Films 402 (2002) 226–231. F.C. Case, J. Vac. Sci. Technol. A 2 (1984) 1509–1512. E. Cavanna, J.P. Segaud, J. Livage, Mater. Res. Bull. 34 (1999) 167–177. R. Binions, C. Piccirillo, G. Hyett, I.P. Parkin, J. Mater. Chem. 17 (2007) 4652–4660. R.G. Palgrave, I.P. Parkin, J. Am. Chem. Soc. 128 (2006) 1587–1597. R.G. Palgrave, I.P. Parkin, Chem. Mater. 19 (2007) 4639–4647. R. Binions, I.P. Parkin, C. Piccirillo, C.J. Carmalt, R.G. Palgrave, Chem. Vap. Depos. 1–2 (2008) 33–39. C. Bae, S. Kim, B. Ahn, J. Kim, M. Sungc, H. Shin, J. Mater. Chem. 18 (2008) 1362– 1367. E.S. Hwang, J. Lee, Chem. Mater. 12 (2000) 2076–2081. R. Binions, C.S. Blackman, C.J. Carmalt, I.P. Parkin, S.A. O’Neill, K.C. Molloy, L. Apostolico, Polyhedron 21 (2002) 1943–1947. F. Beleille, L. Mazerolles, J. Livage, Mater. Res. Bull. 34 (1999) 2177–2184.