Journal of Sol-Gel Science and Technology 31, 6772, 2004c 2004 Kluwer Academic Publishers. Manufactured in The United States.

Preparation and Characterization of Peroxo Titanic Acid SolutionUsing TiCl3

C.K. LEE, D.K. KIM, J.H. LEE, J.H. SUNG AND I. KIMDepartment of Metallurgical Engineering, Dong-A University, #840 Hadan 2-dong,

Saha-gu, Busan 604-714, Koreaiskim@daunet.donga.ac.kr

K.H. LEESurface Engineering Department, Korea Institute of Machinery & Materials, 66 Sangnam-Dong,

Changwon, Kyungnam 641-010, Korea

J.W. PARKDivision of Materials Science and Engineering, Hanyang University, Seoul 133-791, Korea

Y.K. LEEDivision of Information and Communication Engineering, Science and Engineering Research Institute,

Uiduck University, Kyongju 780-713, Korea

Abstract. The peroxo titanic acid solution was successfully prepared using titanium trichloride as a precursor.The basic properties of the TiO2 film prepared by the solution were investigated in view of phase change, bandgapenergy, crystalline size etc. The film displayed amorphous TiO2 at room temperature, anatase above 281C anda mixture of anatase and rutile at 990C. The crystalline size increases with annealing temperatures, while thebandgap energies decrease due to the quantum size effect and the formation of rutile phase which has low bandgapenergy. As a result of TG-DTA, it was found that annealing treatment at 990C for 2 h formed a mixture of anataseand rutile through three steps: (1) the removal of physically adsorbed water (2) the decomposition of peroxo group(3) amorphous-anatase or anatase-rutile phase transformation.

Keywords: titanium dioxide, titanium trichloride, peroxo titanic acid solution, properties

1. Introduction

Titanium dioxide has attracted much attention becauseof its wide range of industrial applications. Manycoating solutions have been developed. Among these,the peroxo titanic acid solution (PTA) is of great in-terest because it has neutral pH and low materialcost.

To whom all correspondence should be addressed.

The formation mechanism of the titanium per-oxo complex has been previously investigated byMuhlebach et al. [1]. However, a study about the for-mation of TiO2 film using the PTA has been recentlyperformed by Ichinose et al. [2]. They mainly usedTiCl4 as a precursor for the synthesis of the PTA solu-tion, but so far no study performed using TiCl3.

Titanium trichloride can easily dissolve into dis-tilled water (neutral pH), while titanium tetrachlo-ride only dissolves into acid solution. Also, titanium

68 Lee et al.

tetrachloride released large amount of chlorine gaswhen it was exposed in air. At this point, titaniumtrichloride is considered to be a good precursor for thepreparation of the PTA.

In this study, the PTA solution was prepared usingTiCl3 as a precursor and the basic properties of the filmdip-coated on various substrates were investigated.

2. Experimental

A titanium trichloride (Aldrich, 10 wt%, 26 cc) was dis-solved in distilled water. Then NH4OH is added to thesolution until final pH is 8.5. The precipitates obtainedthrough the phase transformation to titanium hydrox-ide were filtered and rinsed with distilled water severaltime to remove impurities. After then, hydrogen perox-ide (10 cc) was slowly added to the precipitates in water(100 cc) to obtain the peroxo titanic acid solution.

The various substrates, such as slide glass, quartzand silicon wafer, were dip-coated using a preparedsolution. Each substrate was cleaned in acetone for5 min and rinsed with distilled water prior to coating. Acoated specimen was dried at room temperature for 6 hand heated to various temperatures up to 990C with aheating rate of 10C/min followed by holding in air for2 h.

The surface morphologies were observed by a FieldEmission Scanning Electron microscope (FESEM) andthe thermogravimetric (TG) and differential thermalanalysis (DTA) were performed for the as-dried pow-

Figure 1. TG-DTA curves for as-dried powders at room temperatures.

ders at room temperature using a TG-DTA analyzer(Shimazu TA-50WSI) with a heating rate of 10C/minin air. The phase change of the film with annealing tem-peratures was investigated by a X-ray diffractometer(Philips PW 3710) and the absorbance of the film wasmeasured by a UV-Visible spectrometer. XPS analysiswas carried out to confirm the presence of impuritiesand identify the chemical states of the film. The bindingenergy values were calibrated with the C1s line of ad-ventitious carbon at 284.60 eV. Before experiments, thesurface of specimen was argon-ion sputtered (energy5 KeV) for 20 s.

3. Results and Discussion

The TG-DTA curves for as-dried powder of PTA so-lution are shown in Fig. 1. The TG-DTA curve showstwo stages of weight loss with an endothermic peakminimum at 102C and exothermic peaks maximum at257C. These two stages of weight loss are attributedto the removal of physically adsorbed water and thedecomposition of peroxo group, respectively. Also, abroad exothermic peak was observed in the range 300370C with a maximum 348C. It is due to the slowconversion of amorphous phase to anatase form.

To investigate the composition and existence of im-purities in the film dip-coated on slide glass, the XPSanalysis was performed as shown in Fig. 2. Some Napeaks were observed in survey scan, which is caused

Preparation and Characterization of Peroxo Titanic Acid Solution Using TiCl3 69

Figure 2. (A) survey and (B) narrow scans for the TiO2 film heat-treated at 500C in air for 2 h.

by the diffusion of Na in slide glass, but the othersare only composed of Ti and O peaks. The Ti(2p3/2)peak and Ti(2p1/2) in Fig. 2(B) are positioned around458.35 eV and 464.05 eV, respectively, and differencebetween two peaks is within 5.7 0.05 eV, which ischaracteristic of Ti in TiO2 [3]. In this study, the ox-ides containing Ti3+, such as Ti2O3, are expected to bepartially formed because titanium trichloride as a pre-cursor is used. However no oxides containing Ti3+ wasobserved. This fact indicates that Ti3+ is fully oxidizedto Ti4+ due to air exposure during reaction in solutionor due to the post heat treatment in air.

Figure 3 shows the phase change of the film dip-coated on silicon wafer in accordance with annealingtemperatures. The samples calcined below 273C dis-played amorphous phase, while those calcined above281C exhibited anatase or a mixture of anatase and ru-tile, indicating that amorphous-anatase transformationoccurred around 280C.

The average crystallite size was estimated from themain peak of the anatase (101) using the DebyeScherrers formula as [4]:

t = (0.9)/( cos )

70 Lee et al.

Figure 3. XRD patterns showing the effects of heat treatment in phase transformation of the TiO2 film dip-coated on Si wafer.

where t is the crystallite size of TiO2 film, is thewavelength of the copper K radiation (1.5406 A), is the full-width at half-maximum (FWHM) of theX-ray line, and is Braggs diffraction angle. It hasbeen well known that the diffraction line width is af-fected by strain, defects, surface tension, and instru-mental broadening effects. In this study, instrumentalbroadening was estimated with a standard silicon sam-ple and taken into account in the crystallite size esti-mation. However, any contributions to line broadeningexcept instrumental line broadening effects were ne-glected. The crystallite size increases with annealingtemperatures as shown in Fig. 4. Figure 5 shows thesurface morphologies of the TiO2 film with annealingtemperatures. The grain size at 710C and at 990Cdisplayed 50 nm and 100 nm, respectively, whilethe average crystallite size measured by the DebyeScherrers formula at same conditions exhibited 43 nmand 55 nm, respectively. This fact indicates that theDebyeScherrers formula induce some errors in mea-suring the absolute values of the crystallite size.

For wavelength close to value where the scatteringlosses are dominated by the fundamental absorption,the absorption coefficient can be calculated using afollowing relationship [5]:

= ln(1/T )/d

where d is the thickness of the film and T is theoptical transmittance. In the vicinity of fundamen-

Figure 4. Change of average crystallite size with annealing tem-peratures.

tal absorption, the dependence of hv on photon en-ergy for indirect transition is given by the expression[6]:

(hv)1/2 = Ai (hv Eg)

where hv is the photon energy, Ai is a constant whichdoes not depend on photon energy and Eg is the bandgap energy. A quantitative evaluation of the bandgapenergy can be performed by plotting (hv)1/2 againsthv and extrapolating the absorption edge to zero asshown in Fig. 6. It was found that the bandgap energies

Preparation and Characterization of Peroxo Titanic Acid Solution Using TiCl3 71

Figure 5. SEM micrographs showing the change of surface morphologies of the TiO2 film with annealing temperatures.

Figure 6. Determination of the bandgap energies of the TiO2 filmheat-treated under different conditions.

of the film decreased with annealing temperatures. Thisis ascribed to the quantum size effects [7] and the for-mation of rutile phase [8] which has lower bandgapenergies than anatase at high temperature as shownin Fig. 3. However, the film formed at 328C exhib-ited lower bandgap energies than those formed at 473and 685C. A similar phenomenon was also observedby Mardare et al. [8]. They formed the TiO2 film onquartz substrate using a DC sputtering technique andinvestigated an influence of heat treatment. The filmheat-treated at 400C displayed lower bandgap ener-gies than that heat-treated at 600C. They reported thatthis phenomenon was cause by the formation of rutilephase at lower temperatures. However, in this study, wedid not observe any rutile pahse at lower temperaturesas shown in Fig. 3. The more research about this willbe performed in near future.

4. Conclusions

The peroxo titanic acid solution was successfully pre-pared using titanium trichloride as a precursor. The

72 Lee et al.

film, undergone annealed process at 500C, exhibitedanatase TiO2, indicating that Ti3+ is fully oxidized toTi4+ due to air exposure during reaction in solution ordue to post heat treatment. TG/DTA results for as-driedpowder at room temperature shows that the film heat-treated at 990C forms a mixture of anatase and rutilephase through three stages: (1) the removal of physi-cally adsorbed water (2) the decomposition of peroxogroup (3) amorphous-anatase or anatase-rutile phasetransformation. The crystallite size increases with an-nealing temperatures, while bandgap energies decreasefrom 3.72 eV at room temperature to 3.17 eV at 990C.This fact is attributed to the quantum size effect and for-mation of rutile phase which has lower bandgap energythan anatase.

Acknowledgment

This work was supported by grant No. R12-2002-004-02-001 from the Center for Advanced Net Shape

Manufacturing and Clean Processes of the Korea Sci-ence & Engineering Foundation. We thank Dr. M.S.Won at the KBSI (Busan) and S.G. Lee at the KBSI(Daegu) for their valuable discussion in XPS and XRDexperiments.

References

1. J. Muhlebach, K. Muller, and G. Schwarzenbach, Inorg. Chem.9(11), 2381 (1970).

2. H. Ichinose, M. Terasaki, and H. Katsuki, J. Ceram. Soc. Jpn.104(8), 715 (1996).

3. P. Babelon, A.S. Dequiedt, H. Mostefa-Sba, S. Bourgeois, P.Sibillot, and M. Sacilotti, Thin Solid Films 322, 63 (1998).

4. L.J. Meng and M.P. Dos Santos, Thin Solid Films 226, 22(1993).

5. N. Martin, C. Rousselot, D. Rondot, F. Palmino, and R. Mercier,Thin Solid Films 300, 113 (1997).

6. D. Mardare and A. Stancu, Mater. Res. Bull. 35, 2017 (2000).7. T.W. Wang, S.K. Zheng, W.C. Hao, and C. Wang, Surface Coating

Technol. 155, 141 (2002).8. D. Mardare and G.I. Rusu, Mater. Lett. 56, 210 (2002).