Comparison of photochemical properties of brookite and anatase TiO2 films

The photocatalytic oxidation activity and photoinduced hydrophilicity of brookite-rich TiO2 film were compared to those of anatase TiO2 film under ultraviolet (UV) light irradiation.

The photocatalytic oxidation activities were evaluated by the initial photodegradation rates of methylene blue and cis-9-octadecenoic acid.

The rates for the methylene blue bleaching were nearly identical (−2.2 × 10−3 Abs min−1) for the brookite-rich and anatase films, and the rate for cis-9-octadecenoic acid decomposition on the brookite-rich film was slightly less than that on the anatase one (−0.08 ± 0.01 μg cm−2 min−1 and −0.10 ± 0.01 μg cm−2 min−1, respectively).

However, the rate constants for photoinduced hydrophilicity, which were evaluated by the changes in water contact angles (θ) under rather weak UV light irradiation (5 μW cm−2), were better for the brookite surface than the anatase one.

Moreover, the brookite-rich film became more hydrophilic (θ = 10°) than the anatase one (θ = 18°) after long time exposure to weak UV light irradiation, and showed a good ability for the photoinduced hydrophilicity.


In recent years, numerous researchers have investigated the photocatalytic activities of TiO2 due to the strong oxidation power of its photogenerated holes under ultraviolet (UV) light irradiation,1 as well as its chemical inertness and nontoxicity.

In addition to the strong oxidation power, we reported that UV light irradiation causes the TiO2 surface to become highly hydrophilic.2

These useful properties have been applied to commercial products, such as anti-fogging mirrors, self-cleaning ceramic tiles, etc. At normal pressure, TiO2 has three different crystal structures, rutile, anatase and brookite.

As photocatalysts, both rutile and anatase phases have been widely investigated, and it is well known that the photocatalytic activities of the anatase phase are mostly superior to those of the rutile one.

In contrast, there are a limited number of reports on brookite phase TiO23–14 and most of these papers are simply on its preparation.

There are a few reports on its photocatalytic and electrochemical properties,10,12–14 but none on photoinduced hydrophilicity of brookite films.

Here, the photocatalytic oxidation activity and the photoinduced hydrophilic property of a brookite-rich film are compared with those of an anatase film.


TiO2 films with a thickness around 270–280 nm were prepared on SiO2-coated glass plates by a spin-coating method of TiO2 sols (1500 rpm for 10 s), followed by a calcination at 500 °C for 30 min.

Two types of films were prepared: film A was from NTB-01 (brookite type TiO2 sol: SHOWA DENKO K. K., Japan) and film B was from STS-01 (anatase sol: Ishihara Sangyo Kaisha, Ltd., Japan).

For comparison, corresponding powders A and B were prepared by desiccating the NTB-01 and STS-01 sols at 500 °C for 30 min, respectively.

The crystal phases of the films were identified by glancing incidence X-ray diffraction (RINT-2100, Rigaku Denki Ltd., Japan) at a constant incident angle of 0.8° using Cu-Kα radiation operated at 40 kV–30 mA.

Conventional θ–2θ scanning was used in the powder XRD measurements.

The optical absorption spectra were recorded on a UV-VIS-NIR scanning spectrophotometer (UV-3100PC, Shimadzu Co., Japan).

The microstructures of these films were observed by atomic force microscopy (AFM, SPA300, Seiko Instruments Inc., Japan) and scanning electron microscopy (SEM, S-4200, Hitachi Co., Japan).

The morphologies of the powders were observed using a transmission electron microscope (TEM, JEM-2010, JEOL, Japan).

Degradation of Methylene Blue (MB) was measured to evaluate the photocatalytic oxidation activity.

MB is a brightly coloured, blue cationic thiazine dye with λmax between 550 and 600 nm, and is not photoblenched by UV light irradiation in the absence of TiO2 under the present experimental conditions.

MB was applied on the TiO2 films by dipping the films in 1 mmol l−1 methylene blue solution for 30 min.

The change of the absorbance peak in the visible region was measured by the spectrophotometer.

It is known that reduction of MB leads to colorless leucoMB, which can be reoxidized to MB by oxygen.15

To exclude the contribution of this bleaching reduction process, the irradiated TiO2 samples were stored in the dark until the absorption remained constant before the measurement.

The photocatalytic oxidation activity of the TiO2 films was also measured by the weight change of cis-9-octadecenoic acid under UV light irradiation.

Approximately 10 mg of cis-9-octadecenoic acid was first applied onto the 6.25 cm2 samples and their weight changes were measured by an electronic balance (AG285, Mettler Toledo, Switzerland, 10 μg precision).

The weight was determined averaging 5 different measurements from the same sample.

Details have been given elsewhere.16

Because cis-9-octadecenoic acid on a glass substrate was stable when irradiating with UV light, we concluded that the weight changes were not due to the evaporation of cis-9-octadecenoic acid but from the photooxidization by the TiO2 films.

The photoinduced hydrophilic property was evaluated by measuring the water contact angle (θ) of the TiO2 film's surfaces by the sessile drop method with a commercial water contact meter (CA-X, Kyowa Interface Science, Japan).

The droplet size used for the measurements was 0.7 μL.

The value of the contact angle was determined averaging at least 5 different measurements from the same sample.

All the above UV light irradiation experiments were done using a cylindrical black-light lamp as a light source.

The UV intensity was measured by a UV radiometer (UVR-2, TOPCON, Japan).

Results and discussion

Fig. 1 shows the XRD patterns of Films A, B and Powders A, B. Although Film B and Powder B showed only anatase peaks, Film A and Powder A displayed characteristic peaks for brookite (JCPDS #29-1360) in addition to weak anatase ones.

In order to estimate the ratio (R) of brookite for Powder A, the following equation was used.17

Note that the equation was obtained from the literature by excluding the contribution of the rutile peak.

(kanatase = 0.886, kbrookite = 2.721)

Here Ianatase101 and Ibrookite121 represent the integrated intensities of the anatase (101) peak and the brookite (121) peak, respectively.

The anatase (101) peak and the brookite (120) and (111) peaks were separated using a numerical deconvolution method.

The R value was estimated to be about 0.75 from Fig. 1 pattern (c), indicating that Powder A consists of mostly brookite phase with some anatase one.

The estimated value agrees well with the previously reported value for the NTB-01 sol,13 suggesting that crystal phases did not change by the thermal treatment at 500 °C for 30 min.

Because Film A was prepared under the same thermal conditions with Powder A, it is reasonably expected that the film also consists of mostly brookite phase with some anatase one.

Hereafter, Film A, Film B, Powder A and Powder B are referred to brookite-rich film, anatase film, brookite-rich powder and anatase powder, respectively.

The SEM and AFM measurements showed that the surface roughnesses of the brookite-rich and anatase films were nearly identical.

The average roughnesses (Ra), which indicate surface roughness estimated from AFM measurements, were 5.2 nm and 4.3 nm for the brookite-rich and anatase films, respectively, with the same grain size around 10–20 nm.

These films were transparent between 400–800 nm.

The UV-VIS spectra showed that the main absorption edges for the brookite-rich and anatase films were located around 380 nm, which corresponds to the photoelectrochemically determined band gap energy values E = 3.26 eV for brookite and E = 3.20 eV for anatase (not shown here).

Fig. 2 is the TEM image of the brookite-rich powders, showing the characteristic crystal shape, a plate-like aspect with a square plane.

Electron diffraction patterns identified the indices of the square surface as brookite (010) and (210).

Now, let us evaluate the photocatalytic oxidation activities of the brookite-rich and anatase films.

Fig. 3 shows the change in the absorbance of the MB peak in visible region under 1 mW cm−2 UV light irradiation.

The continuous UV light irradiation caused the adsorbed MB absorption to decrease.

Here we estimated the degradation rate (k) for MB from the initial slope to be kbrookite = kanatase = −2.2 × 10−3 Abs min−1 under the assumption of a pseudo-zero-order kinetics.16

Fig. 4 shows the weight change of the applied cis-9-octadecenoic acid adsorbed on the TiO2 films.

The weight immediately started decreasing upon UV light irradiation accompanied by CO2 production.

Continuous UV light irradiation decomposed almost all the reactants.

The initial decomposition rates (k′) for the brookite-rich film and the anatase one were estimated to be kbrookite = −0.08 ± 0.01 μg cm−2 min−1 and kanatase = −0.10 ± 0.01 μg cm−2 min−1, respectively.

Irradiating with UV light generates electrons and holes in the bulk TiO2 that diffuse to the surface and then react with organic materials adsorbed on the surface directly or via various active oxygen species, such as ˙OH and ˙HO2.18–21

These organic compounds eventually decompose into CO2 and H2O, due to the strong oxidation power of the generated radicals and holes.

Since these two films have almost identical surface roughness, the difference in the photocatalytic decomposition activities is attributed to that in the amounts of the active species produced by UV light irradiation.

The ratio of the total absorbed photon numbers for the present brookite-rich to anatase films was calculated to be 1.26.

Therefore, the present results suggest that the photocatalytic oxidation activity of the anatase film is roughly the same or rather greater than that of the brookite-rich film.

Ohtani et al. reported that the brookite-type TiO2 showed almost the same activity for photocatalytic oxidation reaction (photocatalytic mineralization of acetic acid) as that of anatase-type TiO2,14 which coincides with our present results.

Next, let us examine the photoinduced hydrophilicity of these films.

Fig. 5 shows the changes in the water contact angle for the brookite-rich and anatase films under the irradiation with sufficient UV light (1 mW cm−2).

Before irradiation, the contact angles of these two TiO2 films were around 30°.

Although a small difference was initially observed between the two films in Fig. 5, the hydrophilicizing behaviors were very similar after 30 min and lead to a decent hydrophilic conversion when θ = approximately 5°.

In contrast, Fig. 6 demonstrates the changes when irradiating with a rather weak UV light (5 μW cm−2).

Under this condition, not enough photons were produced to fully create hydrophilic surfaces and both films were partially hydrophilic, reflecting the sensitivity for the hydrophilicizing reaction.

We previously reported that plots of the reciprocal of the water contact angles versus irradiation time have a linear relationship and its slope can be regarded as the hydrophilicizing rate constant.22,23

From the inset of Fig. 6, it is estimated that the rate constants of the brookite-rich and anatase films for hydrophilic conversion were 7.2 × 10−5 degree−1 min−1 and 2.7 × 10−5 degree−1 min−1, respectively.

In addition, the contact angle of the brookite-rich film became to 10°, although that of the anatase one was only ∼18° after long time exposure to the weak UV light irradiation.

The contact angle of the heterogeneous surface is determined by well-known Cassie equation.24

In the case of brookite-rich film, the contact angle is given bycosθ = fa cosθa + fbcosθbwhere fa, fb are the fractions of anatase and brookite, and θa, θb, θ are the contact angles of anatase, brookite and the mixed surface, respectively.

Since the brookite-rich surface showed lower contact angle than that of the pure anatase film under UV light irradiation, it is reasonable to consider that the brookite phase exhibits a better ability for the photoinduced hydrophilicity than the anatase one.

It is important to note that the brookite-rich film showing the better hydrophilic conversion property is not superior in oxidation ability to the anatase one.

We previously reported that SrTiO3 film having a good oxidation property did not become highly hydrophilic under UV light irradiation.25

These results indicate that photoinduced hydrophilicity is not simply induced by the oxidation of adsorbed organic impurities on the surface.

In fact, nearly zero degrees of water contact angle on the UV light irradiated TiO2 surface increases to about 10 degrees when the TiO2 is sonicated in ultra-pure water.26

Such a deterioration of the wettability is also observed when the highly hydrophilic TiO2 is stored in high vacuum chamber.27

We proposed that the highly hydrophilic state is due to the formation of unstable OH group on TiO2 surface23 and the surface with bridging site oxygen may give a better change for the formation of unstable OH species.28

The (010) face of brookite, which is considered to be one of the main faces of the present brookite particles, has such a two-fold oxygen atoms with rather high surface concentration (7.1 × 1014 site cm−2) as shown in Fig. 7.

This might be one of the reasons for a good hydrophilic property of the present brookite-rich film.

The photoinduced hydrophilicity of TiO2 (mostly anatase) has been already used commercially in various materials.

However, highly hydrophilicity can be obtained only under rather strong UV light irradiation, above 100 μW cm−2, and thus its application field is almost limited in outdoor use.

The present results show that the brookite-rich TiO2 film is more sensitive in hydrophilicity than anatase one especially under weak UV light irradiation, indicating that the application field potentially expanded by using the brookite film.


This study compared the photocatalytic oxidation and photoinduced hydrophilicity of brookite-rich TiO2 to the anatase one.

It is noteworthy that the activity for the photoinduced hydrophilic reaction of the brookite was more sensitive than that of anatase, while the photocatalytic oxidation activities of these two phases were nearly identical or rather lower.

These results showed that the photoinduced hydrophilic conversion was not simply explained by the degradation of organic impurities, and supported the existence of alternative mechanism for it.

Moreover, the current work reveals a high potential for brookite type TiO2 as a coating material.

It suggests that the hydrophilicity may be improved by using the brookite type TiO2.