Iron doped titania thin films prepared by spin coating

Abstract

Thin films of titania (TiO₂) doped with 1–5 wt-%Fe were spin coated on glass slides and then annealed at 500°C for 2 h. Results revealed that all of the films were ∼500 nm in thickness and consisted of anatase. Increasing the dopant levels revealed the following trends: the major (101) anatase peak shifted to larger lattice spacing [the peaks were normalised using the principal (210) chromium coating peak as an in situ standard]; the microstructures showed increasing shrinkage cracking, a phenomenon that has not been reported previously; the optical transmittance generally decreased; the amplitude of the transmittance curves generally decreased; and the optical indirect band gap decreased from 3·40 to 3·18 eV, possibly due to silicon contamination, residual thermal stress and/or the formation of midgap states from Fe doping.

Keywords

Titania Spin coating Thin film Single beam focused ion beam milling Glancing angle X-ray diffraction Field emission scanning electron microscopy UV–vis spectrophotometry Indirect band gap

Introduction

Presently, titanium dioxide (titania) thin films are important materials for applications in dye sensitised solar cells1, 2 and photovoltaic cells3, 4 owing to their advantages, such as long term stability against photochemical corrosion, energy band edges that are well matched with the redox level of water, electronic properties that can be varied by changing the lattice defect chemistry or the oxygen stoichiometry and minimal or zero environmental impact. It is well known that TiO₂ (anatase) can absorb light below the wavelength of 387·5 nm, producing a band gap of 3·20 eV; thus, it absorbs very little visible light (400–800 nm).5 Consequently, one of the key technological issues with regard to titania is the extension of the absorption region into the visible light range.

One of the most commonly used methods to enhance visible light absorption is to dope the material with cations and anions, including vanadium,6 chromium,7 iron,8 cobalt,9 copper,10 zinc,11 nitrogen12 and boron.13 Of the preceding, Fe is of interest owing to its octahedral trivalent radius of 0·079 nm, which is similar to that of octahedral tetravalent Ti (0·075 nm).14 Therefore, it can be inferred that Fe ions may be incorporated in the Ti sublattice in TiO₂, and this would enhance oxygen vacancy formation, leading to an improvement in the photocatalytic performance of TiO₂.15 However, it should be noted that, while doping with transition metals is an attractive method, it can also lower the crystallinity of the films.16

There are several methods to prepare Fe doped TiO₂ films, including sputtering,17 pulsed laser deposition,18 metal organic chemical vapour deposition,19 anodic oxidation,20 dip coating,21 hydrothermal growth,8 electrohydrodynamic atomisation,22 ultrasonic spray pyrolysis23 and spin coating.24

The first three techniques in the preceding paragraph require a vacuum system to undertake the deposition, and this implies higher infrastructural costs; the latter three techniques are less suitable for large scale implementation. Moreover, the ultrasonic spray pyrolysis technique has a very low yield rate (∼5%).23 However, the advantages of spin coating over the other methods is its effectiveness, versatility and practicality with regard to the preparation of TiO₂ thin films, and furthermore, the process can be carried out in atmospheric conditions, thus eliminating the need for a vacuum chamber and additional systems. The properties of the film depend only on three parameters, which are the spinning speed, solution concentration and annealing temperature. It is commonly known that high spinning speeds (⩾5000 rev min⁻¹) will produce ultrathin films (<200 nm).25 On the other hand, thick films (about 1–2 μm) can be obtained using low spinning speeds (∼1000 rev min⁻¹).25 Furthermore, with increasing solution concentration, the film thickness increases linearly. In the previous study by the authors,26 the annealing temperature was shown to cause both grain growth and phase transformation of anatase→rutile.

In the present work, Fe doped TiO₂ films were prepared on soda lime silica glass slides using spin coating with a view to investigate the effects of the dopant concentration on the structural, morphological and optical properties of the thin films. An important aspect of this work is the use of a chromium coating, which was carried out as part of the sample preparation for analysis using scanning electron microscopy (SEM). This coating was found to be an essential step since it could be used as an in situ normalisation standard for critical data interpretation of the X-ray diffraction results.

Experimental

Solution precursors were prepared using titanium isopropoxide (TIP, reagent grade, 97 wt-%, Sigma-Aldrich) mixed with 10 mL of 12M HCl, followed by dissolution in isopropanol (reagent plus ⩾99 wt-%, Sigma-Aldrich) at a titanium concentration of 0·1M (2·8 g of TIP diluted to 100 mL volume with isopropanol); the solutions were mixed by stirring in glass at 400 rev min⁻¹ for 15 min without heating. The Fe dopant concentration was varied in proportions of 1, 3 and 5 wt-%Fe (metal basis in TIP) by adding iron chloride tetrahydrate (FeCl₂.4H₂O, reagent grade ⩾98 wt-%, Sigma-Aldrich) to the solution. Spin coating (Laurell WS-65052) was carried out by dropping ∼0·2 mL of solution onto a soda lime silica glass microscope slide (25×25×1 mm) being spun at 2000 rev min⁻¹ in air. The film was dried by spinning for 15 s, followed by instantaneous heating at 75°C for 5 min. This process was repeated four more times in order to obtain as deposited thin films of ∼500 nm thicknesses. Subsequent annealing was carried out in a muffle furnace at 500°C, with a soaking time of 4 h, heating rate of 300°C h⁻¹ and natural air cooling.

All four resultant thin films were sputter coated simultaneously with chromium to a thickness of 20 nm after the spectrophotometric analysis was carried out. The thin films were characterised using the following methods:

single beam focused ion beam (FIB) milling (FEI XP200) for the determination of film thickness

glancing angle X-ray diffraction (GAXRD, angle of incidence of 1° 2θ, penetration depth of <300 nm (Ref. 22), speed of 1° min⁻¹ 2θ, step size of 0·02° 2θ, Philips X'Pert materials research diffractometer) for mineralogical analysis and measurement of lattice volume changes (GAXRD data were obtained from sequential analysis in order to minimise the potential of unit misalignment)

iii

field emission SEM (FESEM, 5 kV accelerating voltage, secondary electron emission mode, Hitachi S4500) for microstructural examination

UV–vis (Perkin Elmer Lambda 35) dual beam spectrophotometry for analysis of the transmittance and optical indirect band gap.

Results and discussion

The GAXRD patterns of the films (Fig. 1) show that all the films consisted solely of anatase; the corresponding peak positions are listed in Table 1. The data in Table 1 indicate that, without normalisation of the peak positions with regard to the major Cr peak (210), the peak positions alone do not provide any meaningful data. However, after normalisation using Cr as the in situ standard, the peak positions showed a trend of decreasing 2θ angles with increasing dopant level. The absence of any consistent error in the measurements shows that misalignment of the unit was not responsible for the observed differences in 2θ; these are likely to have resulted from individual sample loading variations. Although internal standards cannot be used with thin films, the normalisation procedure using the conducting coating can be considered to be an essential component in the comparative assessment of X-ray diffraction data. There does not appear to be any instances where any standard has been used to report thin film GAXRD data in the previous literature.

Figure 1

Glancing angle X-ray diffraction data of TiO₂ films as function of dopant concentration

Table 1

Summary of analytical data

Sample	(101) peak position 2θ			Thickness/nm	Transmittance (400–800 nm)/%	Optical indirect band gap/eV
Sample	Original	Normalised	Difference	Thickness/nm	Transmittance (400–800 nm)/%	Optical indirect band gap/eV
Undoped	25·39	25·72	+0·33	∼500	∼80	3·40
1 wt-%Fe	25·27	25·52	+0·25		∼80	3·35
3 wt-%Fe	25·48	25·34	–0·14		∼70	3·28
5 wt-%Fe	25·15	25·10	–0·05		∼65	3·17

From Fig. 1, it can be seen that the peak shifts of the principal (101) peak of anatase indicate that Fe doping caused lattice expansion, which can be attributed to the substitution of Ti⁴⁺ (ionic radius^VI = 0·075 nm) by Fe³⁺ (ionic radius^VI = 0·079 nm).14 The maintenance of charge balance that arises due to the substitute would require the formation of oxygen vacancies, and this can affect the band gap.15

Figure 2 shows low resolution FIB cross-sectional images of the films. The thickness of the films was consistent and similar for all dopant concentrations. Furthermore, it can be seen that the surface roughness increased with increasing Fe dopant concentration. The surface morphologies were confirmed using high resolution FESEM images obtained at high magnification (×20 000), as shown in Fig. 3. The FESEM images show that the surface roughness, as suggested by the FIB images, was the result of shrinkage and subsequent cracking of the doped films. Previous studies have not reported shrinkage cracks for TiO₂ thin films doped at similar Fe levels and fabricated by spin coating, although the precursors used were different.24, 27, 28

Figure 2

Single FIB images of TiO₂ films as function of dopant concentration:

Figure 3

Images (FESEM) of TiO₂ films as function of dopant concentration:

The transmittance data of the films produced in the current work are shown in Fig. 4. It is seen that the thin films reported in previous studies had higher transmittance compared to the ones obtained in the present work.24, 27, 28 Therefore, it can be concluded that these shrinkage cracks resulted in an enhancement of the scattering of light that consequently reduced the transmittance. The notable differences between the previous studies and the present work are that different metallorganic precursors were used and a drying stage at 75°C was absent in the former. Both of these factors are likely to have caused the observed differences in results with regard to shrinkage cracking.

Figure 4

UV–vis spectroscopy data of TiO₂ films as function of dopant concentration

Figure 5

Optical indirect band gap data of TiO₂ films as function of dopant concentration

Figure 4 shows that the transmittance decreased with increasing Fe doping concentration, which is in agreement with previous works.24, 27, 28 In the present case, however, decreasing transmittance can be attributed to both increasing absorption by Fe and increasing scattering from the cracks. The optical indirect band gap of TiO₂ films can be calculated using UV–vis spectroscopy data, the details of which have been described elsewhere.29 Figure 5 and the data in Table 1 suggest that the optical indirect band gap of the films decreased from 3·40 to 3·17 eV with increasing dopant concentration, similar to previous studies.24, 27, 28

It should also be noted that two other important factors could have contributed to the alteration of the optical indirect band gap, namely silicon contamination and thermal contraction stresses. Si⁴⁺ (ionic radius^VI = 0·054 nm), which is substantially smaller than Ti⁴⁺ (ionic radius^VI = 0·075 nm), could cause lattice contraction (with substitutional solid solution), which would negate the expansive effect of Fe³⁺ (ionic radius^VI = 0·079 nm).14 Alternatively, the entry of Si into the TiO₂ lattice interstitially could cause lattice expansion. The effect of Si contamination of TiO₂ films by substrates of soda lime silica glass26 and fused SiO₂ (Refs. 23 and 30) has been reported previously. Since Si⁴⁺ and Ti⁴⁺ are isovalent, the lattice expansion and decreased band gap are consistent with interstitial solid solution formation (Si in TiO₂). The presence of interstitial Si⁴⁺ (source of electrons) could raise the conduction band and the band gap, resulting in the measured value of 3·40 eV being higher than the commonly reported value of 3·20 eV for bulk anatase.5 However, such comparisons are obscured by the wide range of reported optical indirect band gaps for undoped anatase films, which range from 3·20 to 3·60 eV.23 With regard to thermal expansion cracking, the typical thermal expansions at 500°C are ∼8×10⁻⁶°C⁻¹ for TiO₂ (Ref. 31) and ∼5×10⁻⁶°C⁻¹ for soda lime silica glass.32

The expansion of the film from the greater cooling contraction could cause lattice expansion of TiO₂, resulting in a residual tensile stress on the film. Conversely, undoped TiO₂ films on fused SiO₂ (thermal expansion at 500°C of only ∼0·3×10⁻⁶°C⁻¹)30 have been reported to show an increase in the optical indirect band gap to 3·40 eV (from 3·20 eV for bulk TiO₂) owing to possibly residual compressive stresses.30 The authors’ explanation appears to derive from that of Thomas and Dube33 via Du et al.,34 who also used published band gap information (for SrTiO₃) as a comparison. They33 speculated that intrinsic stress resulted in a raised energy configuration; however no physical rationalisation was provided for this possible state. These results raise the possibility of a direct relationship between residual stress and band gap (tensile stress→band gap decrease; compressive stress→band gap increase).

It is well known that the establishment of midgap states between the conduction and valence bands can decrease the band gap and increase the electrical conductivity.35 If midgap states exist, they probably are in the form of Fe donor bands and oxygen vacancy acceptor bands, which would explain the observed decrease in band gap in Fe doped films in the present work.

Yu et al.36, 37 also observed a decrease in the optical indirect band gap in Fe doped TiO₂ films. They stated that the absorption in the visible region may be induced by a sub-band gap transition corresponding to the excitation of 3d electrons of Fe³⁺ to the TiO₂ conduction band (charge transfer transition) at 415 nm and the d–d transition ²T_2g→²A_2g or ²T_1g of Fe³⁺ or the charge transfer transition between iron ions (Fe³⁺+Fe³⁺→Fe⁴⁺+Fe²⁺) at 500 nm.

Summary

The present work reports the synthesis of Fe doped TiO₂ thin films of ∼500 nm thickness by spin coating. Annealing at 500°C generated films comprised solely of anatase. The GAXRD data, normalised using the Cr (210) peak, indicate lattice expansion with increasing dopant concentration. However, drying at 75°C resulted in surface shrinkage cracking in the doped films only. This suggests that doping and drying were responsible for the cracking. Light absorption from Fe and scattering from the cracking are considered to be responsible for the decrease in transmission. The optical indirect band gap of the films decreased from 3·40 to 3·17 eV with increasing Fe dopant concentration, probably owing to the formation of midgap states, although residual tensile stress and Si contamination could be contributory factors. Since the mechanisms of crack propagation are closely related to the forming techniques used, further work is planned to investigate this aspect.

Footnotes

Acknowledgements

The authors are grateful for the financial support of the Australian Research Council (ARC), which have allowed this and other developmental work to be undertaken. The authors also would like to acknowledge access to the UNSW node of the Australian Microscopy and Microanalysis Research Facility (AMMRF).

References

Zhang

, Zhu

, Ke

, Liu

, Ramakrishna

: ‘Anatase mesoporous TiO₂ nanofibers with high surface area for solid-state dye-sensitized solar cells’, Small, 2010, 6, 2176–2185.

Wen

, Tao

, Ishikawa

, Itoh

, Feng

: ‘Dye-sensitized solar cells based on anatase TiO₂ nanocrystals exposing a specific lattice plane on the surface’, Appl. Phys. Lett., 2010, 97, 131906.

Foong

TRB

, Shen

, Hu

, Sellinge

: ‘Template-directed liquid ALD growth of TiO₂ nanotube arrays: properties and potential in photovoltaic devices’, Adv. Funct. Mater., 2010, 20, 1390–1396.

, Chang

P.-C

, Chien

C.-J

, Lu

: ‘Applications of tunable TiO₂ nanotubes as nanotemplate and photovoltaic device’, Chem. Mater., 2010, 22, 5707–5711.

Dholam

, Patel

, Adami

, Miotello

: ‘Hydrogen production by photocatalytic water-splitting using Cr- or Fe-doped TiO₂ composite thin films photocatalyst’, Int. J. Hydrogen Energy, 2009, 34, 5337–5346.

Songara

, Patra

, Manoth

, Saini

, Gupta

, Gowd

, Vadera

, Kumar

: ‘Synthesis and studies on photochromic properties of vanadium doped TiO₂ nanoparticles’, J. Photochem. Photobiol. A: Chem., 2010, 209A, 68–73.

Takaoka

, Nose

, Kawashita

: ‘Photocatalytic properties of Cr-doped TiO₂ films prepared by oxygen cluster ion beam assisted deposition’, Vacuum, 2009, 83, 679–682.

Chan

KYS

, Goh

GKL

: ‘Hydrothermal growth of ferromagnetic Fe-doped TiO₂ films’, Thin Solid Films, 2008, 516, 5582–5585.

Subramanian

, Vijayalakshmi

, Venkataraj

, Jayavel

: ‘Effect of cobalt doping on the structural and optical properties of TiO₂ films prepared by sol–gel process’, Thin Solid Films, 2008, 516, 3776–3782.

10.

Hou

, Zhao

, Meng

, Jia

, Ye

, Zhou

, Li

: ‘Room-temperature ferromagnetism in Cu-doped TiO₂ thin films’, Thin Solid Films, 2008, 516, 3223–3226.

11.

Zhang

, Zhu

, Li

, Wang

: ‘Photocatalytic Zn-doped TiO₂ films prepared by DC reactive magnetron sputtering’, Vacuum, 2007, 82, 328–335.

12.

Chan

M-H

, Lu

F-H

: ‘Characterization of N-doped TiO₂ films prepared by reactive sputtering using air/Ar mixtures’, Thin Solid Films, 2009, 518, 1369–1372.

13.

Begum

, Farveez Ahmed

, Hussain

: ‘Characterization and photocatalytic activity of boron-doped TiO₂ thin films prepared by liquid phase deposition technique’, Bull. Mater. Sci., 2008, 31, 741–745.

14.

Shannon

: ‘Revised effective ionic radii and systematic studies of interatomic distances in halides and chaleogenides’, Acta Crystallogr. A: Found. Crystallogr., 1976, 32A, 751–767.

15.

Takeda

, Suzuki

, Odaka

, Hosono

: ‘Photocatalytic TiO₂ thin film deposited onto glass by DC magnetron sputtering’, Thin Solid Films, 2001, 392, 338–344.

16.

Bouras

, Stathatos

, Lianos

: ‘Pure versus metal-ion-doped nanocrystalline titania for photocatalysis’, Appl. Catal. B: Environ., 2007, 73B, 51–59.

17.

Hossain

, Biswas

, Takahashi

, Kubota

, Fujishima

: ‘Investigation of sputter-deposited TiO₂ thin film for the fabrication of dye-sensitized solar cells’, Thin Solid Films, 2008, 516, 7149–7154.

18.

Sasaki

, Beck

, Koshizuka

: ‘Preparation of Pt/TiO₂ nanocomposite films by 2-beam pulsed laser deposition’, Appl. Surf. Sci., 2002, 197–198, 619–623.

19.

Baguer

, Neyts

, van Gils

, Bogaert

: ‘Study of atmospheric MOCVD of TiO₂ thin films by means of computational fluid dynamics simulations’, Chem. Vap. Deposition, 2008, 14, 339–346.

20.

Abdullah

, Sorrell

: ‘Preparation and characterisation of TiO₂ thick films fabricated by anodic oxidation’, Mater. Sci. Forum, 2007, 561–565, 2159–2162.

21.

Kim

, Hahn

, Oh

, Kim

: ‘Influence of calcination temperature on structural and optical properties of TiO₂ thin films prepared by sol–gel dip coating’, Mater. Lett., 2002, 57, 355–360.

22.

Mahalingam

, Edirisinghe

: ‘Novel preparation of nitrogen-doped titanium dioxide films’, J. Phys. D: Appl. Phys., 2008, 41D, 215406, 9 pp.

23.

Nakaruk

, Ragazzon

, Sorrell

: ‘Anatase–rutile transformation through high-temperature annealing of titania films produced by ultrasonic spray pyrolysis’, Thin Solid Films, 2010, 518, 3735–3742.

24.

Wang

, Lin

, Yang

: ‘Characteristics and optical properties of iron ion (Fe³⁺)-doped titanium oxide thin films prepared by a sol–gel spin coating’, J. Alloys Compd, 2009, 473, 394–400.

25.

Weill

, Dechenaux

: ‘The spin-coating process mechanism related to polymer solution properties’, Polym. Eng. Sci., 1988, 28, 945–948.

26.

Nakaruk

, Lin

, Perera

, Sorrell

: ‘Effect of annealing temperature on titania thin films prepared by spin coating’, J. Sol-Gel Sci. Technol., 2010, 55, 328–334.

27.

Zhang

, Shen

, Gu

, Wu

, Xie

, Zhang

: ‘Influence of Fe ions in characteristics and optical properties of mesoporous titanium oxide thin films’, Appl. Surf. Sci., 2009, 256, 85–89.

28.

Jeong

J.-H

, Ahn

J.-H

, Kim

B.-H

: ‘Effect of Fe-doping on the crystallinity and the optical properties of TiO₂ thin films’, J. Korean Phys. Soc., 2005, 46, 559–561.

29.

Nakaruk

, Kavei

, Sorrell

: ‘Synthesis of mixed-phase titania films by low-temperature ultrasonic spray pyrolysis’, Mater. Lett., 2010, 64, 1365–1368.

30.

J.-G

, Yu

H.-G

, Cheng

, Zhao

X.-J

, Yu

, Ho

W.-K

: ‘The effect of calcination temperature on the surface microstructure and photocatalytic activity of TiO₂ thin films prepared by liquid phase deposition’, J. Phys. Chem. B, 2003, 107B, 13871–13879.

31.

Morrell

, ‘Handbook of properties of technical & engineering ceramics. Part 1: an introduction for the engineer and designer’; 1989, London, HMSO.

32.

Shand

, ‘Glass engineering and handbook’; 1958, New York, McGraw-Hill Book Company

33.

Thomas

, Dube

: ‘Optical properties of sol–gel processed amorphous and crystalline SrTiO₃ thin films’, Jpn Appl. Phys., 2000, 39, 1771–1775.

34.

, Zhang

M.-S

, Wu

, Kang

, Yang

, Wu

, Yin

: ‘Optical properties of SrTiO₃ thin films by pulsed laser deposition’, Appl. Phys. A, 2003, 76A, 1105–1108.

35.

Kittel

: ‘Introduction to solid state physics’, 8th edn; 2005, New York, John Wiley & Sons.

36.

, Yu

, Ao

, Lee

, Yu

, Ho

: ‘Preparation, characterization and photocatalytic activity of in situ Fe-doped TiO₂ thin films’, Thin Solid Films, 2006, 496, 273–280.

37.

, Xiang

, Zhou

: ‘Preparation, characterization and visible-light-driven photocatalytic activity of Fe-doped titania nanorods and first-principles study for electronic structures’, Appl. Catal. B: Environ., 2009, 90B, 595–602.