Procedimiento de Creación

5.3.1 Precursors

["Bu4N]3[TaW5 0i9] is a white powder soluble in aprotic solvents such as acetone and

acetonitrile. The heteropolyoxometalate ["B^NBITaW.sOiy] has a similar structure to the isopolyoxometalate ["B ^N htW ^O ^] except that one tungsten atom is substituted by a tantalum atom. fB ^ N H T a W ^ O ^ ] was synthesized according to literature procedure as described below. 81

Preparation o f BU4N]2[Ta W50 / 9]

The preparation of ["ButNhPTaWsOu] required the use of ["Bu4N]3[W0 4], the synthesis of

which is described in Section 3.2.

A solution comprising of tantalum ethoxide, [TaCCX^H.^] (2.78 g, 6.85 mmol) and trichloroacetic acid, CI3C2O2H, (2.24 g, 13.7 mmol) in acetonitrile (5 cm3) was added drop

wise to a solution of ["BU4NM W O4] (10.0 g, 13.7 mmol) in acetonitrile (10 cm3 ) whilst

stirring vigorously. The solution was stirred for a further 30 minutes, resulting in the formation of a white precipitate. The precipitation was completed by addition of diethyl ether (150 cm ). The precipitate was collected under suction filtration, washed with diethyl ether (25 cm3) and dried in air.

IR (KBr disk, cm '1): 430.6 (vs), 573.3 (m), 790.3 (m), 885.3 (m, sh), 916.1 (m), 957.6 (m), 1026.5 (w), 1152.4 (w), 1381.4 (s), 1488.9 (m).

The TGA of ["B^NhlTaW sOiy] showed a total mass loss of 23 % at 400°C, corresponding to decomposition to WO3 and Ta2C>5, which is theoretically associated with a mass loss of

24 %. However a small mass loss was also observed prior to the onset of decomposition, which is probably due to removal of solvent from the sample. AACVD reactions using

["B^NhfTaW.sO^] above 400°C are therefore expected to yield either mixed phased WO3

5.3.2 Depositions using["Bu4N]3[TaWsO 1 9]

AACVD reactions using ["Bi^NfilTaW.sO^] were carried out using a precursor solution of the polyoxometalate (0.25 g) in acetonitrile (50 cm3) at substrate temperatures between 500 - 600°C and a flow rate of 0.5 L m in'1.

As with depositions using isopolyoxometalates and ["B^NHNbW sOufi, film deposition occurred exclusively on the top plate for AACVD reactions using ["Bu-iNhfTaWsOufi, indicating that film growth occurs via gas phase nucleation (Figure 1.3, Process 4), rather than aerosol droplet settling (Process 1 and 2) or molecular diffusion to the substrate as was the case for tungsten hexacarbonyl (Process 3).

The films deposited from ["Bu-iNHTaW^O^] were composed of multiple regions. At 500°C the film was comprised of three distinct regions: a powdery brown deposit near the precursor inlet (region A), an intermediate grey/blue partially adherent film (region B) and an adherent blue coating with interference fringes in the regions furthest from the precursor inlet (region C). Increasing the substrate temperature to 550°C afforded a film comprised only of the adherent blue region (region A) with interference fringes and the grey/blue partially adherent coating (region B). Deposition at 600°C yielded a predominately powdery brown coating (region A), with a small adherent blue region furthest from the precursor inlet (region B). Complete coverage of the substrate occurred at deposition temperatures between 500 - 550°C whereas no deposition occurred in the third of the substrate furthest from the precursor inlet at 600°C.

Wavelength dispersive analysis by X-rays (WDX) was carried out on the films deposited from ["Bu-jN^ITaWsOicd in order to determine their W:Ta and W :0 ratios. WDX showed that the 5:1 tungsten: tantalum ratio of the precursor had been approximately retained in all regions of the films (Table 5.2). It was not possible however to determine the tungsten: oxygen ratios of the films as they were not sufficiently thick to prevent appreciable breakthrough to the underlying substrate.

Deposition Temperature / °C Tungsten:Tantalum ratio 500-region A 4.41:1 500-region B 4.23:1 550-region A 5.07:1 550-region B 4.54:1 600 4.68:1

Table 5.2 The W:Ta ratios of the films deposited from ["BiMNfotTaWsOu*], determined by WDX. The ratios represent an average of a minimum of three spots across the film. For films composed of multiple regions; region A is the closest to the precursor inlet.

Films deposited from ["Bi^NhlTaW ^O^] were analysed by XRD. The films deposited at substrate temperatures of 500 and 550°C were composed of multiple regions. In both cases the diffraction pattern of the regions of the film closest to the precursor inlet were dominated by broad peaks indicative of a poorly crystalline material, and those furthest

from the precursor inlet by the 0 2 0 reflection of Y-WO3 (March Dollase r factor = 0.59 and

0.67 respectively) . 121 The film deposited at 600°C was also comprised of W O3 crystallites,

which were preferentially orientated along the <0 1 0> direction (r = 0.74). AACVD reactions of ["Bu-tNfitNhW^O^] also afforded films composed of multiple regions which showed an increase in the randomisation of the crystallites with increasing deposition temperature. The XRD patterns of the films following annealing in air at 550°C were

characteristic of randomly orientated WO3 (r = 1), which indexed in the P2\/n space group

with typical cell parameters of a = 7.35, b = 7.55, c = 7.73

A,

p = 90.44°. Figure 5.5 shows the XRD patterns of the films deposited from rB ^ N h tT a W sO ^ ] at 500 - 600°C.

The absence of any peaks corresponding to tantalum oxide or tantalum tungsten oxides phases in the diffraction patterns of both the as-deposited and annealed films is indicative

of the formation of either tantalum doped W O3 (Tao.17Wo.83O3) films or composite films in

which the tantalum oxide is a poorly crystalline phase. The incorporation of dopants into a film is expected to shift the lattice parameters and change the relative intensities of the

peaks. However the size of the tantalum 5+ and tungsten 6+ cations is very similar (0.64 and

0.60

A

respectively) 9 therefore only a 0.008

A

increase in the lattice parameters, compared

tantalum. Such a shift is within error of the experimental data and would be difficult to observe due to limitations in the instrument resolution. Furthermore the preferred orientation of the films and the poor resolution of the data made it difficult to assess any changes in the relative intensity of the peaks.

(0 4 4 c 3 _{Post-annealing} l r = 1 600°C~RegiorTB ^ r = 0.74 >. w CB 4 4 !o_w < >> m c © 4 4 550°C Region B r = 0.67 c 500°C Region B ( r = 0.59) 10 20 30 40 50 60 28 / d egrees

Figure 5.5 XRD patterns of films deposited from ["Bu^NfetTaWsOi®]. The diffraction pattern of the region furthest from the precursor inlet is shown (region B). The diffraction pattern of the region closest to the precursor inlet (region A) contained broad peaks characteristic of a poorly crystalline material.

XRD has shown that as with depositions using ["B^NFfNbW SO^] films deposited from ["ButNhlTaW.sO^] are either poorly crystalline or composed of < 0 1 0> preferentially

orientated WO3 crystallites. The tantalum is thought to be either a dopant in the WO3 films

or have formed a separate phase which is poorly crystalline.

The Raman patterns of the films deposited from ["B^NFfTaW .sO^l were characteristic of

amorphous WO3-X, exhibiting bands at 220, 770 and 950 cm' 1 corresponding to the

vibrations of W4+-0 , W6+- 0 and W ^ O bands respectively. 44,45 Similar Raman patterns

were recorded for the films deposited from ["B utN ^N bW sO ^], ['!Bu4N]3[W0 4] and

[W(CO)6]. The presence of the reduced tungsten cations is consistent with the blue regions of the films. The broad nature of the bands and their low intensity made it difficult to assess

any shift in their position relative to WO3 films deposited from undoped precursors. The Raman patterns of the annealed films were characteristic of crystalline monoclinic W O 3,

displaying bands at 804 and 709 cm' 1 arising from W6+- 0 stretching vibrations and a band

at 269 cm' 1 which was assigned to the W 6+-0-W 6+ stretching mode. 127 The bands

corresponding to the vibrational modes of WO3 were shifted to lower wavenumbers compared to films deposited from isopolyoxotungstates, indicating that the tantalum is a dopant in the WO3 films (Figure 5.6) . 134 However the shift in the position of the bands is

smaller than that observed for depositions using [''B^N^tNbWsO^]. Such a small shift may be less than the error associated with the measurement. Furthermore the broad nature of the peaks made it difficult to assess the peak maxima. The absence of any bands arising from Ta-O-Ta vibrational modes confirms the formation of doped films rather than multiphase films. Intense Raman bands corresponding to vibrational modes of TSL2O5 are

expected at wavenumbers of 610, 500 and 240 cm' 1, 137 which do not coincide with any

peaks corresponding to y-WCh, discounting the possibility of masking.

« 'c 3 >• k . < >> 0) c © c 810 830 850 670 690 710 730 750 770 790 Wavenumber / cm'1

Figure 5.6 The shift in the position of the W6+- 0 vibrational bands of the annealed film deposited from ["B^NMTaWsO^] at 550°C compared to the film deposited from [nBu4N]2[W6 0 19] at the same temperature.

XRD and Raman spectroscopy showed that AACVD reactions of ["B^N^fTaW.sOio] afforded films with a partially reduced WO3-X stoichiometry in which the crystallites were

preferentially orientated along the <0 1 0> direction. The annealed films were composed of randomly orientated crystallites of monoclinic WO3 (r = 1). These results are consistent

with the correlation between the stoichiometry of the films and the randomisation of the crystallites, which was observed for films deposited from ["B^NhfNbW sOig],

isopolyoxometalates and [W(CO>6]. The observation of peaks corresponding to only WO3

in the XRD pattern indicated that the tantalum had either been incorporated into the WO3

films as a dopant, or had formed a secondary phase which was amorphous. The absence of any bands arising from Ta-O-Ta vibrational modes and the shift in the position of the bands for the annealed film relative to films deposited from undoped precursors in the Raman data is also consistent with the formation of tantalum doped tungsten oxide films.

The film deposited from [”Bu4N]3[TaW5 0i9] at 550°C was analyzed by XPS. The surface of

the film exhibited a doublet corresponding to W6+ 4f7 /2 and W6* 4fs/2 photoelectrons at

binding energies of 36.5 and 38.7 eV respectively and an O Is ionization at 530.4 eV. These tungsten and oxygen chemical shifts are in good agreement with both previous

p o 1 9 9

studies of WO3 ~ ‘ and with WO3 films deposited from isopolyoxometalates and tungsten

hexacarbonyl. The absence of any splitting or broadening of the W 4f doublet indicates that the tungsten was present in a single environment, which is contradictory with the Raman data in which bands arising from vibrational modes of reduced tungsten cations were observed and the blue colour of the film which is indicative of a partially reduced WO3-X

stoichiometry. This discrepancy was also observed for films deposited from ["B^NMNbWsOiy] and may be attributed to surface oxidation of the films after deposition and prior to the XPS analysis. The ratio of the W 4f doublet and the O Is peak at 530.4 eV,

taking into account their sensitivity factors is consistent with a WO3 stoichiometry,

confirming that the surface of the film is fully oxidised.

Tantalum was also present in the sample, with Ta Aini and Ta 4f5/2 ionizations occurring at

27.1 and 29.1 eV respectively and an additional O Is environment was observed at 532.4 eV. There are no reported XPS studies of tantalum doped tungsten oxides; however these Ta 4f and O Is chemical shifts are in good agreement with a range of tantalum compounds in oxidation state 5+ including Ta2C>5 138 which would also be expected for Tao.17Wo.8 3O3.

The ratio of the Ta 4f doublet and O Is peak at 530.4 eV, taking into account empirically derived sensitivity factors confirmed that the oxidation state of the tantalum was 5+. XPS

showed that the 5:1 tungsten to tantalum ratio of the precursor was preserved in the film. This is in good agreement with the WDX analysis.

As with the film deposited from ["B^NfifNbW.sO^] nitrogen was present in the sample with the N Is photoelectron at 400.8 eV. The N Is chemical shift is in good agreement with a range of ammonia compounds including (CH3)4NBr130 and [NH4]io[Wi20 4i]131, indicating

that the nitrogen in the polyoxometalates does not get eliminated during the deposition. In addition XPS revealed that the level of carbon contamination on the surface of the film was 4 %. Etching showed that the carbon contamination was entirely surface limited.

XPS and WDX have therefore confirmed the presence of tantalum in the films obtained via AACVD of [”Bu4N]3[TaW5 0i9] and have shown that the W:Ta ratio of the precursor has

been retained. Furthermore Ta 4f and O Is environments which would be expected for Ta 5+doped WO3 films were observed by XPS. This is consistent with the shift in the position

of the bands corresponding to WO3 stretching modes in the Raman spectrum and the lack

SEM showed that the films deposited from ["B^NM TaW sO^] had a similar microstructure to those deposited from ["B u^M N bW sO ^], comprising of an agglomeration of spherical particles, which is indicative of film growth through gas phase nucleation, with needle-type particles growing out of the spheres (Figure 5.7a). Depositions using ["Bu4N]3[TaW5Oi9] afforded films with a smaller mean particle size (200 nm),

narrower size distribution and lower concentration of needle agglomerates (Figure 5.7b) compared to those using ["Bu-tNMNhW.sO^] at the same temperature.

^ O O v ^ v O* O f t ’ O ’ v ^

Particl* Siza /

b

Figure 5.7 Scanning electron micrograph of the film deposited from [nBu4N]3[TaW5Oi9] at 550°C and its size distribution, (a) and a higher magnification

5.4 AACVD o f Tungsten Oxide Films from Mixtures o f [nBu4N]2[W60 19]

In document PROSPECTO DEFINITIVO. (página 82-85)