Raman Study of Vanadium-Doped Titania Nanopowders Synthesized by Sol-Gel Method

International Journal of Modern Physics B Vol. 24, Nos. 6 & 7 (2010) 667–675 World Scientific Publishing Company

667

667

DOI: 10.1142/S0217979210064289

RAMAN STUDY OF VANADIUM-DOPED TITANIA NANOPOWDERS

SYNTHESIZED BY SOL-GEL METHOD

M. ŠĆEPANOVIĆ

Center for Solid State Physics and New Materials, Institute of Physics, Pregrevica 118, Belgrade, Serbia [email protected]

S. AŠKRABIĆ

Center for Solid State Physics and New Materials, Institute of Physics, Pregrevica 118, Belgrade, Serbia sonask@ ipb.ac.rs

M. GRUJIĆ-BROJČIN

Center for Solid State Physics and New Materials, Institute of Physics, Pregrevica 118, Belgrade, Serbia myramyra@ ipb.ac.rs

A. GOLUBOVIĆ

Center for Solid State Physics and New Materials, Institute of Physics, Pregrevica 118, Belgrade, Serbia golubovic@ ipb.ac.rs

Z. DOHČEVIĆ-MITROVIĆ

Center for Solid State Physics and New Materials, Institute of Physics, Pregrevica 118, Belgrade, Serbia zordoh@ ipb.ac.rs

B. MATOVIĆ

Institute of Nuclear Sciences “Vinča”, 11001 Belgrade, Serbia [email protected]

Z. V. POPOVIĆ

Center for Solid State Physics and New Materials, Institute of Physics, Pregrevica 118, Belgrade, Serbia popozor@ ipb.ac.rs

Received 1 September 2009 Revised 30 December 2009

Pure titania (TiO2) nanopowders and TiO2 doped with 10 mol % of vanadium ions (V3+) are synthesized by sol-gel method. The dependence of structural characteristics of nanopowders on synthesis conditions is investigated by X-ray diffraction and Raman spectroscopy. Very intensive modes observed in Raman spectra of all nanopowders are assigned to anatase phase of TiO2. Additional Raman modes of extremely low intensity which can be related to the presence of small

http://dx.doi.org/10.1142/S0217979210064289

668 M. Šćepanović et al.

amount of brookite amorphous phase are observed in pure TiO2 nanopowders. In V-doped nanopowders anatase was the only TiO2 phase detected. The variations in duration and heating rate of calcination influence slightly the Raman spectra of pure TiO2, but have a great impact on Raman modes of anatase, as well as the additional Raman modes related to the presence of vanadium oxides in V-doped samples.

Keywords: Sol-gel syntheses; doped anatase nanopowders; Raman spectroscopy.

1. Introduction

Growing scientific interest for titanium dioxide (TiO2), a material with good performances in photocatalytic oxidation of organic molecules, is mostly due to its application in solar cells, thin-film optical devices, and gas sensors.1 The narrowing of band gap by doping with various transition metals ions and lanthanides1 is the subject of many investigations. Especially, great attention has been dedicated to vanadium (V)-doped titania, with the dramatic impact of preparation procedure on material reactivity.2,3 Although several works confirmed increased photoactivity in V-doped TiO2 with respect to the undoped sample4-6, V-doped TiO2 produced by a co-precipitation method yields opposite performances.7 Besides, considerable attention has been paid to the electronic structure and ferromagnetism of V-doped anatase TiO2, both experimentally and theoretically, for reasons of fundamental interest and potential technological applications in spintronic devices.8

Exploration of the variations in anatase structure, induced by doping nanopowders with vanadium under different synthesis conditions, is the first step in the investigation of reactivity and magnetic properties of V-doped TiO2 nanopowders. In the present work, V-doped TiO2 nanopowders synthesized by sol-gel method are characterized by Raman scattering measurements.

2. Experimental Details

Sol-gel method was used for the synthesis of pure (TiO2) and V-doped anatase nanopowders with TiCl4 as precursor. The Ti(OH)4 hydrogel was obtained by hydrolysis at 0°C with controlled addition of 2.5 wt. % ammonia solution at the pH value of 9.4. Alcogel (obtained after conversion by anhydrous ethanol) was dried in the furnace at 280°C and subsequently calcined at 550°C. The heating rate and duration of calcination for pure (G97-G102) and V-doped (G91-G96) TiO2 samples are specified in the Table 1. The samples G91-G96 were doped with 10 mol % of vanadium ions (V3+) using VCl3 as a dopant.

Powder X-ray diffraction (XRD) was used to identify crystalline phases, to obtain lattice parameters and to estimate crystallite size and lattice strain. The XRD patterns of pure TiO2 were collected on a Philips diffractometer (PW1710) employing CuKα1,2 radiation. Crystal structure of V-doped TiO2 nanopowders was identified by X-ray diffraction (XRD) using filtered CuKα radiation (Siemens D5000). Before measurements, the angular correction was performed by high quality Si standard. Lattice parameters were refined from the data using the least square procedure. Standard deviation was about

Raman Study of Vanadium-Doped Titania Nanopowders Synthesized by Sol-Gel Method 669

1%. Williamson-Hall plots were used to separate the effect of the size and strain in the nanocrystals.

Table 1. Calcination heating rate and duration for pure and V-doped TiO2 samples.

Pure TiO2

samples

V-doped TiO2

samples

Calcination

heating rate

[°C/h]

Calcination

duration

[h]

G97 G91 33.75 7

G98 G92 33.75 24

G99 G93 67.5 7

G100 G94 67.5 24

G101 G95 135 7

G102 G96 135 24

Raman scattering measurements were performed using 514-nm laser line of mixed

Ar+/Kr+ laser and Jobin Yvon T64000 triple spectrometer system, equipped with confocal microscope and a nitrogen-cooled CCD detector. All Raman scattering spectra were recorded at room temperature in air.

3. Results and Discussion

Relevant and the most intensive diffraction peaks in the XRD patterns of all samples belong to anatase crystal structure of TiO2 (JCPDS card 78-2486). The patterns of a pure TiO2 sample and selected TiO2 samples doped with 10 mol % of vanadium ions (V3+) are presented in Fig. 1. The results obtained by Rietveld analysis of X-ray diffractograms (Table 2) show that values of the parameters a and c vary around their reference values (a0 = 0.378479(3) nm, c0 = 0.951237(12) nm). In the XRD patterns of pure TiO2 the presence of low-intensity diffraction peak at 2θ ≈ 30.8°, that can be ascribed to the brookite phase of TiO2 (JCPDS card 29-1360), is also observed. The greatest difference between XRD pattern of pure and doped samples lies in the width of main diffraction peaks of anatase. Large broadening of these peaks implies very small crystallite size (dRD) and high strain value (σ) in pure TiO2 samples, which is confirmed by the results of Rietveld analysis listed in Table 2. In addition, the diffraction peak of extremely low intensity, observed in the diffraction patterns of V-doped samples, at about 20.3° can be ascribed to V2O5.

9 Broadening of this peak in the diffractogram of G96 sample indicates that V2O5 in this sample is highly disordered and partially amorphized, though in order to make a final conclusion, a high resolution XRD would be necessary.


Table 2. Lattice parameters (a, c), crystallite dimension dRD and microstrain σ obtained by Rietveld analysis of X-ray diffractograms of selected pure and V-doped anatase powders.

Lattice parameters The sample

a [nm] c [nm] dRD [nm] σ [%]

Pure TiO2 0.37856 0.94860 12 0.42

G91 0.37851 0.95201 76 0.35

G92 0.37781 0.94873 48 0.36

G96 0.37719 0.95266 42 0.31

The Raman spectra of pure and doped samples are shown in Fig. 2. The dominant modes in the Raman spectra of pure TiO2 sample at about 144 (Eg(1)), 197 (Eg(2)), 397 (B1g(1)), 517 (A1g, B1g(2)) and 639 cm−1 (Eg(3)) can be assigned to the Raman active modes of the anatase crystal. 10 The shift and the asymmetrical broadening of the most intensive anatase Raman mode (Eg(1)), analyzed by phonon confinement model11-14, predict the particle size in pure TiO2 samples in the range of 10-12 nm. The additional Raman modes of extremely low intensities at about 245, 297, 324 and 367 cm-1 (shown in the inset of Fig. 3) can be ascribed to the highly disordered brookite phase of titania.15 Raman spectra from Fig. 3 show that the modes originated both from anatase and brookite phase in pure TiO2 are slightly influenced by heating time and duration of calcination in the range investigated here. However, the anatase Raman modes go through severe changes, whereas brookite modes disappear after doping of TiO2 with vanadium, as can be seen in Fig. 4. The frequencies of anatase modes in V-doped TiO2, shift more with different synthesis condition, in comparison to their pure TiO2 counterpart, as can be seen from

Fig. 1. XRD diffractograms of pure and V-doped TiO2 samples. The features denoted by circles correspond to the reflections of the V2O5.


data listed in Table 3. This is also obvious from Fig. 5, which presents frequency and linewidth of Eg(1) Raman mode for pure and doped samples. Larger blueshift and broadening of Eg(1) mode in V-doped samples point out to increased nonstichiometry16, as these changes cannot be ascribed to phonon confinement effect, due to relatively great crystallite size (>40 nm) in these TiO2 nanopowders. The greatest nonstoichiometry was observed in the sample G96 with the greatest heating rate and duration of calcination.

Fig. 2. Raman spectra of pure and V-doped TiO2 samples.

Fig. 3. Raman spectra of selected pure TiO2 fitted to a sum of Lorentzian shape functions.


Table 3. Frequencies of anatase Raman modes in pure and V-doped TiO2 samples.

Raman frequencies of anatase modes [cm-1] The sample

Eg(1) Eg(2) B1g(1) A1g(1)+B1g(2) Eg(3)

G91 146.0 198.6 398.5 517.0 636.7

G92 145.3 198.2 396.6 516.0 637.4

G93 145.3 198.4 397.0 516.5 637.7

G94 145.8 198.3 398.3 516.4 636.4

G95 145.4 198.0 397.0 517.0 637.7

G96 148.9 201.9 394.6 514.2 633.4

G97 143.9 197.7 396.8 519.0 639.8

G98 143.8 197.9 396.9 519.0 639.7

G99 143.7 197.7 396.8 518.7 639.6

G100 143.5 197.6 396.8 518.7 639.7

G101 143.9 198.0 396.8 519.0 639.7

G102 143.8 197.8 397.0 518.8 639.7

In addition to Raman modes of anatase TiO2, the Raman spectra of V-doped samples

feature twelve well resolved bands at about 101, 143, 284, 305, 406, 478, 531, 699, 810, 926, 996 and 1020 cm-1, numbered from 1 to 12 in Figure 4. The frequencies of these additional Raman modes, which can be related to the presence of vanadium are listed in Table 4. The peaks around 101, 143, 284, 305, 406, 478, 531, 699, and 996 cm-1 may be

Fig. 4. Raman spectra of V-doped TiO2 fitted to a sum of Lorentzian shape functions.


attributed to crystalline V2O5.17,18 Strong bands in the region ~100–300 cm–1 are

attributed to external modes due to the relative motions of long range ordered oxide units, which is why they lie at low-frequencies (101, 305 cm-1).18 The Raman band at about 143 cm−1, due to the vibrations in a V–O–V atomic chain, also belongs to this region18,19, but the position and the width of this band cannot be precisely estimated in the spectra of V-doped TiO2 because of its overlapping with very intensive Eg(1) mode of anatase. The V-O-V stretching and bending modes are assigned to the frequencies from 280 cm–1 to 720 cm-1.18 The sharp and intense band, located at 997 cm-l, indicates the symmetric V=O stretching mode.17,19 There is no doubt that relatively weak band at about 1020 cm−1 corresponds to stretching vibrations of the V=O bond, but maybe not in the configuration of an isolated monovanadate, normally characterized by a band at ~1030 cm−1.20 Note that the band at ~1030 cm-1 has been also attributed to monomeric vanadyl (V4+) species bound directly to the TiO2 support.21 In addition, two broad features in the frequency range of 750–950 cm–1 have been classified as short-range order vibrations for the typical vanadium coordination in vanadate structures.18 The Raman band near 810 cm-1 seems to belong to V-O-V bending vibrations in polyvanadate species.22 The other broad band at ~930 cm-1 also belongs to the polymeric vanadate groups: it is formed by the superposition of two modes corresponding to the vibrations of the bonds V=O internal to the vanadates in the region 960 cm-1 and the terminal V=O bonds between 915 and 955 cm-1.20

In addition, very low intensity of the Raman modes related to crystalline V2O5 and the appearance of well resolved mode at about 1027 cm-1 which can be ascribed to amorphous V2O5

23 in the Raman spectrum of G96 sample confirm the amorphization of V2O5 in this sample, already implied by XRD patterns.

The presence of additional Raman modes in the spectra of doped samples unambiguously shows that vanadium ions form vanadium oxides (mostly V2O5) and some other vanadate structures in V-doped nanopowders. This confirms that higher concentration of V in TiO2 tends to stabilize V in the 5+ state predominantly.3 However, the change in the anatase Raman modes in those samples reveals that certain amount of vanadium ions is introduced in TiO2 crystal lattice, which strongly depends on conditions of synthesis. Although it is obvious that none of investigated V-doped nanotitania powders is the single-phasic V-Ti mixed oxide, it seems that the sample G96 with the

Fig. 5. Raman shift and linewidth of anatase Eg(1) Raman mode in pure (a) and V-doped (b) TiO2 nanopowders.


greatest heating rate and duration of calcination is the closest to the Ti0.9V0.1O2-δ structure,

with δ≈0.03 (estimated by comparison of nonstoichiometry induced blueshift and broadening of Eg(1) anatase Raman mode to the literature data16).

Table 4. Frequencies of V-related Raman modes in V-doped anatase nanopowders.

Raman frequencies [cm-1] Sample

1 2 3 4 5 6 7 8 9 10 11 12

V2O5 100 143 283 303 407 477 526 696 - - 995 -

G91 101 143 284 304 406 477 531 699 808 931 996 1017

G92 101 143 284 305 406 478 531 699 808 924 996 1017

G93 101 143 284 305 406 479 532 700 808 924 996 1017

G94 101 143 284 304 406 479 531 699 808 928 995 1017

G95 102 143 284 305 406 480 531 700 812 926 996 1020

G96 - 143 284 308 - - - 699 810 934 996 1027

4. Conclusion

A detailed Raman study of pure TiO2 nanopowders and TiO2 nanopowders doped with 10 mol % of vanadium ions (V3+), synthesized under different conditions by sol-gel method is presented in this paper. Very intensive modes observed in the Raman spectra of all nanopowders are assigned to anatase phase of TiO2. It is demonstrated that the frequency shift and broadening of the most intensive anatase Eg(1) Raman mode in pure TiO2 nanopowders are almost independent on selected heating rate and duration of calcination. However, Eg(1) mode as well as the other anatase Raman modes in V-doped samples are strongly influenced by doping with vanadium ions. Observed changes, ascribed both to the nonstoichiometry induced by doping and the incorporation of vanadium ions into TiO2 lattice, are most pronounced in the sample with the greatest calcination heating rate and duration. The Raman spectroscopy measurements have also shown presence of amorphous V2O5 in this sample, as well as the existence of crystalline V2O5 in all V-doped nanopowders. This study allows us to investigate the structural variations of V-doped TiO2 nanopowders synthesized by sol-gel method arose from the change in calcination conditions.

Acknowledgment

This work is supported by the Serbian Ministry of Science and Technological Development under Project No. 141047, SASA project F-134 and the Swiss National Science Foundation through grant IZ73Z0-128169.

References

1. M. Bettinelli, V. Dallacasa, D. Falcomer, P. Fornasiero, V. Gombac, T. Montini, L. Romanò, A. Speghini, Journal of Hazardous Materials 146, 529 (2007).

2. K. Bhattacharyya, A.K. Patrab, P.U. Sastryb, A.K. Tyagi, J. Alloys Comp, 482, 256 (2009).


3. K. Bhattacharyya, S. Varma, A. K. Tripathi, S. R. Bharadwaj, and A. K. Tyagi, J. Phys. Chem,

112, 19102 (2008). 4. S. M. Karvinen, Ind. Eng. Chem. Res. 42, 1035 (2003). 5. J. J. Shyue, M. R. De Guire, J. Am. Chem. Soc. 127, 12736 (2005). 6. M. Anpo, Y. Ichihashi, M. Takeuchi, H. Yamashita, Res. Chem. Intermed. 24, 143 (1998). 7. S. T. Martin, C. L. Morrison, M. R. Hoffmann, J. Phys. Chem. 98, 13695 (1994). 8. K. H. He, G. Zheng, G. Chen, T. Lü, M. Wan, G. F. Ji, Solid State Commun. 144, 54 (2007). 9. J. B. de Paiva Jr, W. R. Monteiro, M. A. Zacharias, J. A. J. Rodrigues and G. G. Cortez, Braz.

J. Chem. Eng. 23(04), 517 (2006). 10. T. Ohsaka, F. Izumi, Y. Fujiki, J. Raman Spectrosc. 7, 321 (1978). 11. M. Šćepanović, S. Aškrabić, V. Berec, A. Golubović, Z. Dohčević-Mitrović, A. Kremenović,

and Z. V. Popović, Acta Phys. Polon. A 115(4) 771 (2009). 12. M. J. Šćepanović, M. U. Grujić-Brojčin, Z. D. Dohčević-Mitrović, Z. V. Popović, Mater. Sci.

Forum 518, 101 (2006). 13. M. J. Šćepanović, M. Grujić-Brojčin, Z. D. Dohčević-Mitrović, Z. V. Popović, Appl. Phys. A

86, 365 (2007). 14. M. Šćepanović, S. Aškrabić, M. Grujić-Brojčin, A. Golubović, Z. Dohčević-Mitrović,

A. Kremenović and Z.V. Popović, Acta Phys. Polon. A 116(1), 99 (2009). 15. S. Yin, K. Ihara, B. Liu, Y. Wang, R. Li, T. Sato, Phys. Scr. T129, 268 (2007). 16. J. C. Parker, R. W. Siegel, Appl. Phys. Lett. 57(9), 943 (1990). 17. G. T. Went, S. T. Oyama, A. T. Bell, J. Phys. Chem. 94 (10), 4240 (1990). 18. J. P. Schreckenbach, K. Witke, D. Butte, G. Marx, Fresenius J. Anal. Chem. 363, 211 (1999). 19. S. P. Lim, J. D. Long, S. Xu, K. Ostrikov, J. Phys. D: Appl. Phys. 40, 1085 (2007). 20. F. Chiker, J. Ph. Nogier, J. L. Bonardet, Catalysis Today 78, 139 (2003). 21. C. B. Rodella, P. A. P. Nascente, R. W. A. Franco, C. J. Magon, V. R. Mastelaro, A. O.

Florentino, Phys. Stat. Sol. (a) 187(1), 161 (2001). 22. Z. Luan, P. A. Meloni, R. S. Czernuszewicz, L. Kevan, J. Phys. Chem. B, 101(44), 9046

(1997). 23. S.-H. Lee, H. M. Cheong, M. Je Seong, P. Liu, C. Edwin Tracy, A. Mascarenhas, J. Roland

Pitts, and S. K. Deb, J. Appl. Phys. 92(4), 1893 (2002).

Raman Study of Vanadium-Doped Titania Nanopowders Synthesized by Sol-Gel Method

Documents