Journal of Sol-Gel Science and Technology 22, 23–31, 2001

c 2001 Kluwer Academic Publishers. Manufactured in The Netherlands. 

Preparation of TiO2 Thin Film Photocatalysts by Dip Coating
Using a Highly Viscous Solvent
National Institute for Resources and Environment 16-3 Onogawa, Tsukuba, Ibaraki 305-8569, Japan
Received May 5, 2000; Accepted September 13, 2000

Abstract. TiO2 thin film photocatalysts coated onto soda lime glass were prepared by a dip coating process
using a highly viscous solvent. The source of the TiO2 was tetraisopropyl orthotitanate, and α-terpineol was used
as the solvent. Two types of thin film preparation procedures based on dip coating (a sol-gel system and thermal
decomposition system) were used to prepare the samples. TiO2 thin films were obtained after calcination at 450◦ C
for 1 hour. The film thickness obtained with a single dipping was proportional to the viscosity of the dip coating
solutions. The obtained thin films were transparent with a thickness of 1 µm. The crystal form of the obtained
photocatalyst films was anatase alone. The thin films were formed with aggregated nano-sized TiO2 single crystals
(7–15 nm). The photocatalytic activity of the TiO2 thin films, as evaluated by the photooxidation of NO (1 ppm) in
dry air, was as high as our previous TiO2 thin films prepared by the sol-gel method.
Keywords: TiO2 , photocatalyst, thin film coatings, α-terpineol



Titanium dioxide (TiO2 ) is one of the materials of interest for decomposition of environmental pollutants, i.e.,
to diminish the serious and detrimental effects of air or
water pollution. TiO2 is nontoxic, and is able to oxidize various compounds under irradiation of UV lights
such as part of sunlight wavelength. A lot of studies for
the photo decomposition of pollutants by these photocatalysts are carried out [1–7], and the preparation of
highly active photocatalysts have been studied [8].
In our previous study, TiO2 thin film photocatalysts,
were prepared by dip coating with the sol-gel method
using a water vapor in air for the hydrolysis of metal
alkoxide [8, 9], and these films effectively photooxidized nitrogen oxides [9]. They showed a large specific
surface area, excellent transparency, good crystallinity,
and high photocatalytic activity. However, this preparation method had the following problems. First, more
than 20 dip coatings were needed to obtain 1 µm thick
films. Second, this process could be operated only under low humidity conditions.

In order to obtain sufficient film thickness with only a
few dip coating times, a highly viscous solution would
be required. Previous studies in sol-gel method, control
of the solution viscosity was reported by the control of
the hydrolysis of metal alkoxide and polymerization
of sol [10]. However, it is difficult to control the solution viscosity in the sol-gel reaction, since solution
viscosity is easily changed depending on several complete factors, i.e., temperature, humidity, reaction time,
mixing ratio of reagents, etc. [10, 11]. Viscosity can be
controlled by adding other components to the system, if
they are miscible and do not react with the other components in the system. We have found that a highly
viscous solvent, α-terpineol, successfully gives a fine
and transparent thin film photocatalyst. This solvent is
suitable for dip coating, since it does not react with
metal alkoxide and has a good affinity to glass substrate. Moreover, the thin films do not show any clacks
and flaking off from substrate after calcination in spite
of the thick films obtained. It is expected that the dip
coating time can be reduced significantly by using this

0. EEE and PEG were used as ligand for TPOT. Preparation of the Solutions for Sol-Gel Process The sol-gel solutions were prepared by using acetylacetone (AcAc) as a ligand and acetic acid as an acid catalyst. 1. Thermal analysis (TG/DTA) was carried out using a Shimadzu DTG50 thermal analyzer. and 2. the processes involved in the formation of thin films.34 mol of tetraisopropyl orthotitanate (TPOT).5. EEE). An observation of the surface structure of the thin films was carried out by a Shimadzu SPM-9500 scanning probe microscope (Atomic Force Microscope. 2. 2. The DTA and TG measurements were performed at a heating rate of 2◦ C min−1 in an air flow of 20 ml min−1 using the aluminum cells with α-alumina as the reference material. When the TPOT solution became cloudy. The systems with 0. Compositions of the solution are listed in Table 1. The Brunaur-Emmett-Teller (BET) surface area and pore size distributions of the thin films were measured from N2 gas adsorption-desorption at 77 K using a Quantachrome AS-1 surface area and pore size analyzer. TPOT α-TPO PrOH EEE PEG 0. 2.34 mol of AcAc added were called SG6 and SG34. PEG: polyethyleneglycol 600 Experimental Procedure Sol-gel and thermal decomposition systems were used for the preparation of the thin films.34 2.74 0 0. Table 1. we also examined the thermal decomposition method to obtain thin films using α-terpineol.4. α-TPO: α-terpineol.12 mol of α-terpineol (1. 2.19 0.78 mol of PrOH.12 2.27 1.27 3. The withdrawing speed of the substrate was 1. The dip-coated glass plates were calcined at 450◦ C for 1 hour [9]. 0. and the evaluation of the photocatalytic activity for nitric oxide removal.5 mm/s. Reagent TD TDe TDp TDep Composition of the coating solutions (unit: mol).3.34 0.74 0 0 0. 2. Then.06 and 0.35 mol of acetic acid were mixed continuously during the hydrolysis of TPOT. The above solutions were dried at around 100◦ C after gelation in an aluminum sample cup.1. 0. diethyleneglycol monoethylether ((ethoxyethoxy)ethanol. All images (512 × 512 pixels) were obtained under ambient conditions with low-pass filtering during data acquisitions.34 0. The hydrolysis solution was then dropped into the continuous stirred TPOT solution at a rate of approximately 2 ml min−1 [12]. This dipping process was repeated to obtain a film of approximately 1 µm thickness. Characterization of the Thin Films We have developed a thermal decomposition (TD) process as a novel preparation method for TiO2 thin films. and 0. 2.34 0.12 2. coating. Preparation of the Solutions for Thermal Decomposition Process Preparation of Thin Films The above solutions were coated onto the silica coated soda lime glass plates (100 × 100 × 1 mm) by dip Evaluation of Photocatalytic Activity The photocatalytic activity of the thin films was evaluated from the photo-oxidation of nitric oxide (NO) .96 0 0. The crystal structure of the thin films was identified using a Rotaflex RU-300 (Cu Kα radiation at 40 kV and 80 mA) X-ray diffraction spectrometer. respectively. and polyethylene glycol 600 (PEG). α-TPO. α-TPO) were mixed.19 TPOT: tetraisopropyl orthotitanate. The process was carried out in an ambient air and at room temperature (approximately 25◦ C).57 3. 9]. PrOH: 2-propanol. The film thickness was determined by the interference pattern in the transmittance spectra of the films with a Hitachi 330 spectrophotometer [8. EEE: (ethoxyethoxy)ethanol. The viscosity of these solutions was measured with a Brookfield DV-I+ viscometer.p-menthen-8-ol.34 mol of AcAc was added to the above solution.12 4. the film structure including the porosity. 2propanol (PrOH).11 mol of distilled water.24 Negishi and Takeuchi In this study. In addition to the sol-gel method. The film weight was calculated by subtracting the weight of the substrate from that of TiO2 coated sample. AFM). 1.96 mol of 2-propanol (PrOH) as the solvent. we will discuss the preparation method of TiO2 thin films using α-terpineol as solvent.2.12 2. Chemicals used were TPOT.06 or 0.

80 cP with PEG600 doped system) [9]. ❍: TDep. : TD. Figure 2 shows the increase in film thickness with increasing the number of dipping. : SG6. which agrees with the reports by Takahashi et al. : TDp. [14] and Kozhukharov et al. 3. The other thin films showed almost the same transmittance. The transmittance of the TDep film in the visible region was 75%. 3. Increase in the TiO2 film thickness with the increase in the number of dipping. The viscosity values were higher than that of our previous film (ca.1. Numbers show the number of dipping. 9. The rate of increase in the Figure 3. ❑ : SG34 . . The increase ratio Figure 1. : TDe.Dip Coating Using a Highly Viscous Solvent 25 using a flow type photoreactor. The NO concentration was monitored with a chemiluminescent NOx analyzer (Monitor Labs Model 8840). 4. The film thickness increased linearly. of film thickness using α-terpineol as a solvent is over three times of the previous our method [9]. 3. This result shows that the number of coating can be reduced by α-terpineol used. These wave patterns were attributed to the light interference and the number of peaks was related to the film thickness [8. The UV-A irradiance was approximately 0. Figure 2. [15]. Results and Discussion Preparation of the TiO2 Thin Film Photocatalyst Figure 1 shows the change in the transmittance spectra of the TDep films with an increase in the dip coating times.38 mW cm−2 (λmax = 365 nm) at the photocatalyst surface. as shown in Fig. Transmittance spectra of the TDep system with an increase in the number of dipping. The number of wave pattern peaks increased with the increase in the number of dip coating. The relationship between solution viscosity and film thickness was seen linear. Relationship between the solution viscosity and film thickness. Dry air containing 1ppm of NO was passed through the reactor at a rate of 1. 13].5 L min−1 .

as shown in Fig. In the case of TD film. It is considered that an application of this equation may be accurate for highly viscous solutions. Homogeneous thick films were obtained from the PEG doped system. transparent thin films were successfully obtained using α-terpineol solvent even under the ambient atmosphere. However. the application times for the required film thickness can be easily predicted. These white precipitates are heterogeneously formed with the TiO2 . and as a result. The value 0. The function of EEE in this system is not only to stabilize TPOT by such coordination but also to prevent the segregation of TiO2 while calcination. white precipitates were partially distributed over the film surface with an increase in the number of dip coating. reported that (methoxyethoxy)ethanol (MEE) was coordinated with TPOT [16. it can be assumed that EEE can also be coordinated with TPOT. The TDe film was clear with no precipitates. this case is limited around a low viscous region. As shown in Fig. Figure 4 shows the photographs of the films obtained by the thermal decomposition method on a sheet of section paper. The SG films are not shown here. Appearance of the obtained TiO2 thin films. Figure 4. which leads to a clear and homogeneous thick film.26 Negishi and Takeuchi film thickness was dependent on the solution viscosity. Brinker et al. Equation (1) was obtained from these results t = η n/0. These thin films were preparing under ambient air. Thus. . white spots were formed on the surface of TDp film. 4.0759 (1) where t indicates the film thickness after calcination.0759 is constant under the constant dip coating speed and solution temperature. the TPOT-PEG complex was partially separated from the solution during the thermal treatment process. These result shows that all films were transparent. Since the properties of MEE and EEE are similar. and n indicates the number of dipping. Similarly. Nishide et al. white spots were formed on the TDp film. η is the solution viscosity. These results indicate that the surfaces of thin films have precipitates without EEE. Using this equation. because they are always clear with homogeneous thickness. PEG is also coordinated with TPOT. Both of precipitates (TD film and TDp film) were formed after calcination. only the TDep solution gave a homogeneous and clear thin film. and this point is big difference with our previous film preparation method [9]. but the film thickness was not uniform. reported that the relationship between solution viscosity and film thickness is not linear [11]. 4. It is consid ered that EEE works to increase the TPOTPEG complex dispersion in solution. 17]. However.

Dip Coating Using a Highly Viscous Solvent Figure 5. Similarly. the first decrease in the sample weight was observed between 100 and 150◦ C. 7. The DTA peaks were attributed to the combustion of organic compounds. gels. As shown in Fig. Figure 7 shows the X-ray diffraction pattern of the thin films. Details of this sharp DTA peaks are under investigation [21]. and θ is the diffraction angle. b is the halfwidth. . large peaks with weight loss were observed on the thermal decomposition series. The value of Figure 6. This decrease is attributed to the evaporation of water [19. a sharp DTA peak around 450◦ C was also observed for TD and TDe systems. DTA and TG traces of thermal decomposition system Figure 7. DTA and TG traces of the sol-gel system gels. In the case of the SG systems. no difference in crystallinity could be observed for the films prepared by the sol-gel and thermal decomposition methods. On the other hand. This sharp DTA peak was accompanied by a tiny weight loss and could not be observed in the PEG doped system. Large DTA peaks with a large weight loss were observed at around 27 300◦ C. Figures 5 and 6 show the results of the thermal analysis of the SG and TD systems. However. The sharp DTA peak may be related with the changes of the film structure. 7. such as solvent and alcohol formed from hydrolysis of TPOT. the sharp DTA peak is not directly related with the crystallization of TiO2 . XRD patterns of the TiO2 thin films. A sharp peak at 25◦ is ascribed to anatase (101) and a broad pattern in the same regions attributed to the substrate SiO2 . The crystallite size of TiO2 was calculated from the (101) peak of anatase in Fig. 20]. using r = K λ/b cos θ (2) where λ is the wavelength of the X-ray.

4 depending on many factors including the shape of the primary crystallite.1 7. as shown in Fig. pore volume.9 to 1.9 measured by NO removal (1. and 70 m2 g−1 for the systems without PEG. being a little smaller than those in our previous studies (ca. .28 Negishi and Takeuchi Table 2. the structural features of the films were similar to each other.1 67. 9. The surface of the films was quite flat. 24 h.1 3.7 48. We used the value K = 1. Figure 8.855 4. while the nano-structure of the TD film surface was not obvious compared with the other AFM image of the TiO2 thin films (1 × 1 µm area).740 1.802 1. 100 m2 g−1 ) [9. forming with the aggregated TiO2 particles.0 8.3 4. TDp.9 61.5 10. Table 2 shows the calculated r as well as the specific surface area. reaction time. as reported by Kumar et al. However.3 15. and average pore size of the films.8 3. The specific surface area was around 80 m2 g−1 for the systems with PEG. [20]. with the maximum roughness of 10 nm.1 56. the fine structures between these thin films were remarkably different among these thin films.9 4. 1 µm.38. 18]. and TDep films were also formed along with the TiO2 fine particle.995 1.9 35.8 79 76 71 69 89 76 1. Figure 8 shows the surface structure of the series of TiO2 thin films by AFM observation.8 4. The size of the TiO2 particles constructing SG6 and SG34 films were around 10 to 15 nm. TDe.6 55. As shown in this table.985 2.0 ppm). K varies from 0. film thickness.177 1. Crystallite size (nm) Specific surface area (m2 /g) Pore volume × 10−1 cm3 /g Average pore size (nm) Photocatalytic activitya µmol · 24 h/µm 10.3 14. Samples SG6 SG34 TD TDe TDp TDep a As The texural properties and photocatalyticactivity of TiO2 thin films.

On the other hand. As a result. primary particles generally grow with the linkage of these particles.Dip Coating Using a Highly Viscous Solvent Figure 9. The cloudiness then disappeared. The color change is peculiar to coordination of AcAc to TPOT. the TiO2 particles did not grow as much in the TDp and TDep systems as compared with the TDe system at 450◦ C calcination. as shown in Figs. gelation was observed for the SG6 solution only after 2 days under the same conditions. since TPOT was not thermally stabilized by the ligand. Therefore. The formation mechanism of the TiO2 particles is observed to be different between the sol-gel system and thermal decomposition system. In the case of the TD system. The difference of decomposition temperature shows that PEG is more strongly coordinated with TPOT than EEE. it is considered that the formation temperature of the TiO2 particle may be lower than those of ligand doped systems. It is considered that the TiO2 particles of the films are aggregated single crystals. In our system. In fact. the TiO2 particles may be primary particles formed with the hydrolysis of TPOT. and the solution color changed from white to clear dark brown. the TiO2 particle was sintered among themselves and lost the fine structure with the increase of temperature. 5 and 6. The size of the TiO2 particles in the PEG doped films was at around 7 nm. thin films. although the molar ratio of AcAc and TPOT was equal in the SG34 solution). It is well known that the decomposition temperature of the metal alkoxide complex is affected by the kind of ligand [16]. The reason may be related to the amount of added AcAc (as described in the experimental section. The thermal decomposition temperatures of the TDp and TDep systems (around 330◦ C) were higher than that of TDe system. gelation did not occur for the SG34 solution even after several months when it was hermetically sealed. These TG/DTA results. it is considered that the TiO2 particle was directly formed from thermal decomposition of TPOT complex. In the sol-gel system. In sol-gel reactions. AcAc was added to the precursor solution at the first indication of cloudiness when TPOT was hydrolyzed. These changes show that the growth of the TiO2 particles was disturbed by the addition of AcAc. 29 AFM image of the TiO2 thin films (100 × 100 nm area). . Thermal decomposition of TDe occurred at 264◦ C. The size of these TiO2 particles is in agreement with the results of XRD observations. In the thermal decomposition system.

Peral and D. J. Fujishima. J. Y. It is considered that this result shows that it is possible to develop the high photocatalytic thin films with the thin film preparation method which uses α-TPO. Loddo. and A. catalysts. L. This is attributed to the poisoning of catalyst surface by the NO− 3 ions formed. I. Inui. J. Mater. Kominami. Increasing of UV-irradiation time. since the hydrolysis process of the titanium alkoxide was suppressed. Photochem. leads to excellent transparency and high photocatalytic activity of the TiO2 thin films. Sci.. the photocatalytic activity of TDp and TDep films was almost the same than our previous TiO2 thin film photocatalyst [9]. This result shows that the presence of chemical additives leads to the high photocatalytic activity. and T. i. Photobiol. Kim and M. 94. NO removal by TD (). photocatalytic activity of SG6 and SG34 is lower than that of TDp and TDep regardless of polymer additives. M. 3. Kera. Catal. Watanabe. A: Chem. 3. J.2. V.H. Lett. J. concentration of NO was drastically decreased just after the beginning of UV-irradiation. it is expected that the DTA peak positions of gels may correspond to those of the films. 197 (1996). However. Schiavello. This reason is not still clear. H. The thermal decomposition method was much easier than the sol-gel method. Photocatalytic activity of the thermal decomposition films was relatively higher than that of the sol-gel films except for the TD film. A: Chem. 10 and 11. 32 (1995). References Figure 11. using an appropriate ligand for the thermal decomposition method.F. 4. binary ligand system (EEE and PEG). Augugliaro. Takada. The main difference of film structure was not found between sol-gel films and thermal decomposition films in Table 2. Y. and TDep (❑) film photocatalysts. Anderson. V. 98. Photochem. Yamagiwa. Photocatalytic ability of thin films was evaluated by the oxidation of NO. J. Photocatalytic Activity of the TiO2 Thin Films Figures 10 and 11 show the photocatalytic activity of the TiO2 films. 153. TDe (❍). Hashimoto. Furthermore. K. Photobiol. Inoue. Moreover. H. control of the humidity was not necessary using α-terpineol as solvent. M.. Ollis. Conclusions TiO2 thin film photocatalysts were successfully prepared by the sol-gel and the thermal decomposition method using a highly viscous solvent. However. As shown in Figs. 221 (1996).. and M. Murasawa. Table 2 shows the relationship between the textual properties and photocatalytic activity of the thin films.A. Palmisano.e. 5. The number of dip coating could be reduced compared with the number of previous dip coating to obtain sufficient film thickness. but TiO2 wet gel prepared directly from dip coating solutions. 4. . 15. NO removal by SG6 () and SG34 (❍) film photo- were not obtained from thin films on glass substrates. 136. S. Catal. 1. 2. 79 (1996). However. D. TDp (). Initial concentration of NO was 1 ppm. Sopyan.30 Negishi and Takeuchi Figure 10. results in a little decrease of the photocatalytic activity. 554 (1992).

Sci. L. 17. Sci. Ceram. Trapalis. J. Res. C.J. Paz. 12. Ibusuki. N. Frye. A. Z. Chen. 33. 3(11). Mater. J. Takeuchi. Sci. 86. 1141 (1993).P. Mater. Huang. Negishi. Sci. Chem. S. 15. T. Phys. K. Mater. and C. Hisanaga and K. T. Negishi and K.C. Y. J. L. Mater. K. Takeuchi.J. Hashimoto. 7. 515 (1999). K. Jpn. 28.J. 13. 5789 (1998). Takeuchi. 20. . J. Tanaka. Ibusuki. N. Soc. Negishi and K. H. 89 (1997). N. Iyoda. 8. Ashley. Mater. A. Ch. 23. K. Cohen and G. Gao. Keizer and A. J. J. 18. K. K.R. Terabe. Burggraaf. 1617 (1994). Lett. 31. 100. 29. 1122 (1992). to be published. Matsuoka. J. Imai. R. Mizukami. Kumar. Mater. 38. Rabenberg. Kato. 18. Nishide and F. and D. Chem. T. and A. Mater. T. Hurd. Miyazaki. Mizukami. Nishide and F. V. 2259 (1988). Kozhukharov. and A. H. Mater. Yamaguchi.J. 16. J. Iguchi. Negishi. and A. 10(11). Luo.K. and B. Chem. Thin Solid Films 298. 31 14.Dip Coating Using a Highly Viscous Solvent 6. 1283 (1993). Samuneva. and Y. Denki Kagaku 60. 150 (1999). N. Negishi. 2842 (1995). 3497 (1996). Solids 147/148. Heller. 21. Non-Cryst. 107 (1992). Lett. 3696 (1982). Takahashi and Y. Fujishima. Benedek. 19. Yan. N. Lett. 10. 841 (1995). Sci. P. Sci.B. Schunk. 11. J. Y. and T.S. 9. 424 (1992). J. J. Datye. G. Mater. T. Takeuchi. J. Brinker.