Synthesis and Characterization of Nano Titania Powder with High

Nov 10, 2004 - Yuanzhi Li,*,†,‡ Nam-Hee Lee,† Doo-Sun Hwang,† Jae Sung Song,§. Eun Gu Lee,| and Sun-Jae Kim*,†. Sejong Advanced Institute o...
1 downloads 0 Views 356KB Size
10838

Langmuir 2004, 20, 10838-10844

Synthesis and Characterization of Nano Titania Powder with High Photoactivity for Gas-Phase Photo-oxidation of Benzene from TiOCl2 Aqueous Solution at Low Temperatures Yuanzhi Li,*,†,‡ Nam-Hee Lee,† Doo-Sun Hwang,† Jae Sung Song,§ Eun Gu Lee,| and Sun-Jae Kim*,† Sejong Advanced Institute of Nano Technologies, #98 Gunja-Dong, Gwangjin-Gu, Sejong University, Seoul, 143-747, Korea, Department of Chemistry, China Three Gorges University, Yichang, Hubei, 443002, People’s Republic of China, Electric & Magnetic Devices Research Group, Korea Electrotechnology Research Institute, #28-1 Sungju-Dong, Changwon, Kyungnam 641-120, Korea, and Department of Materials Engineering, Chosun University, KwangJu 501-759, Korea Received April 23, 2004. In Final Form: August 3, 2004 Nano rutile, anatase, and bicrystalline (anatase + brookite) titania powders with an average crystal size of below 10 nm are prepared from aqueous TiOCl2 solution at low temperatures by adjusting pH values of the starting solution and adding different additives. Adding a small amount of octyl phenol poly(ethylene oxide) into aqueous TiOCl2 solution leads to the change of particle morphologies of obtained nano titania from needlelike to nano spherical rutile crystals. Amorphous-anatase transformation of titania could proceed in liquid-solid reaction at low temperatures, even at room temperature. A formation mechanism of rutile, anatase, and brookite titania was proposed. It is found that H+ or H3O+ plays a catalytic role in the phase transformation from amorphous to anatase titania and that the presence of a small amount of SO42- ion is unfavorable to the formation of both rutile and brookite. By carefully adjusting preparation conditions, nano pure anatase with higher surface area, good crystallinity, and a lower recombination rate of photoexcited electrons and holes was obtained. This nano pure anatase showed a very good photocatalytic activity for gas-phase photo-oxidation of benzene.

Introduction Nanosized titania has received much research attention because of its unique physico-chemical properties in the applications of pigments, cosmetics, fine ceramics, and photocatalysts for environmental purification, catalyst supports, and dielectric materials.1-3 Titania exists in three main crystallographic forms, i.e., anatase, rutile, and brookite phases. Each crystalline structure exhibits different physical properties, which lead to different applications. The preparation of nano anatase titania has been well documented by sol-gel and hydrothermal methods using TiCl4, TiOSO4, titanium alkoxide, and so forth as starting materials,4 but there have been only a few reports on the preparation of nano pure rutile titania up to now.5-9 The difficulty of preparing nano pure rutile titania may be attributed to the fact that the thermody* To whom correspondence should be addressed. E-mail: sjkim1@ sejong.ac.kr; [email protected]. † Sejong University. ‡ China Three Gorges University. § Korea Electrotechnology Research Institute. | Chosun University. (1) Akihiro, K. Jpn. Kokai Tokkyo Koho 1997, Jp 09301825 A2. (2) Karch, J.; Birriger, R.; Gleiter, H. Nature 1987, 330 (10), 556. (3) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W. Chem. Rev. 1995, 95, 69. (4) (a) Pottier, A. S.; Cassaignon, S.; Chaneac, C.; Villain, F.; Tronc, E.; Jolivet, J. P. J. Mater. Chem. 2003, 13, 877. (b) Iwasaki, M.; Hara, M.; Ito, S. J. Mater. Sci. Lett. 1998, 17, 1769. (c) Arnal, P.; Corriu, R. J. P.; Leclercq, D.; Mutin, P. H.; Vioux, A. Chem. Mater. 1997, 9, 694. (d) Chemseddine, A.; Moritz, T. Eur. J. Inorg. Chem. 1999, 235. (e) Kavan, L.; Rathousky, J.; Graetzel, M.; Shklover, V.; Zukal, A. J. Phys. Chem. B 2000, 104, 12012. (f) Zheng, Y.; Shi, E.; Chen, Z.; Li, W.; Hu, X. J. Mater. Chem. 2001, 11, 1547. (g) Sugimoto, T.; Zhou, X.; Maramatsu, A. J. Colloid Interface Sci. 2002, 252, 339. (5) Wu, M.; Long, J.; Huang, A.; Luo, Y. Langmuir 1999, 15, 8822. (6) Zhang, Q.; Gao, L. Langmuir 2003, 19, 967. (7) Cheng, H.; Ma, J.; Zhou, Z.; Qi, L. Chem. Mater. 1995, 7, 663.

namic stability of phases is particle size dependent, and at particle diameters below ca. 14 nm, anatase is more stable than rutile.10 Therefore, it is very difficult to obtain nano pure rutile titania with particle sizes below 14 nm. One of the most important applications of titania is photodetoxification of water and air pollutants because of its high efficiency, stability, nontoxity, and low cost, and this has been a fast growing research area during recent decades.11-13 It is generally accepted that anatase titania is more efficient as a photocatalyst than rutile and brookite titania. It is widely thought that an anatase powder with both a high surface area and a high degree of crystallinity is desirable to enhance the photocatalytic activity since such a powder will have relatively few disruptions in its electronic band structure.12,13 However, it is difficult to achieve these properties at the same time, and so a powder whose crystallites have fewer flaws would be ideal. To obtain nano titania with high photocatalytic activity, careful and stringent adjustment of experimental conditions is required for the selective production of anatase alone. These experimental conditions involve both the type and concentrations of the precursors, their ratios, order and rate of additions, polarity of the solvent, pH, reaction temperature, the absence or presence of small amounts of dopants, and many other, possibly not yet recognized (8) Kumar, K. N. P.; Keizer, K.; Burggraaf, A. J. J. Mater. Chem. 1993, 3, 923. (9) Li, Y. Z.; Fan, Y.; Chen, Y. J. Mater. Chem. 2002, 12 (5), 1387. (10) Zhang, H.; Banfield, J. F. J. Mater. Chem. 1998, 8, 2073. Zhang, H.; Banfield, J. F. J. Phys. Chem. B 2000, 104, 3481. (11) Fujishima, A.; Rao, T. N.; Tryk, D. A. J. Photochem. Photobiol., C 2000, 1, 1. (12) Linsebigler, A. L.; Lu, G.; Yates, J. J. T. Chem. Rev. 1995, 95, 735. (13) Legrini, O.; Oliveros, E.; Braun, A. M. Chem. Rev. 1993, 93, 671.

10.1021/la0489716 CCC: $27.50 © 2004 American Chemical Society Published on Web 11/10/2004

Nano Titania Powder with High Photoactivity

Langmuir, Vol. 20, No. 25, 2004 10839

Table 1. Properties of As-Prepared TiO2 Prepared from TiOCl2 Aqueous Solution under Different Conditions no.

concentration of TiOCl2 (mol/L)

neutralizing agent

temperature (°C)

pH value

additive

phasea

crystallite sizeb (nm)

surface area (m2/g)

1 2 3 4 5 6 7 8 9 10 11 12 13 14

0.3 0.3 0.3 0.3 0.3 0.3 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15

no no no no no no NH4HCO3 NH4HCO3 NH4HCO3 NH4HCO3 NH4HCO3 NH4HCO3 NH4HCO3 NH4HCO3

40 40 60 40 80 80 20 40 40 40 120c 80 80 120c

0.14 0.14 0.14 0.14 0.14

no OP-10 OP-10 SO42no HCl no no no no no no SO42no

R (needlelike) R (spherical) R (spherical) A R+A R (needlelike) A+B A+B A+B amorphous A A+B A A+B

8.7 6.8 8.3 3.5 R(11.6), A(5.5) 8.9 A(2.5) A(4.7) A(4.6)

116 121 117 184 200 110 214 224 219

8.6 A(5.6) 5.3 A(6.4)

182 213 217 194

a

2.7 2.7 4.0 7.0 7.0 2.7 2.7 2.7

R, rutile; A, anatase; B, brookite. b Calculated by Scherrer formula: L ) 0.89λ/β cos θ. c By hydrothermal method in autoclave.

parameters.14 Many attempts have been done to maximize the photocatalytic activity of anatase titania.15-17 Usually, calcination or hydrothermal methods were involved. In this work, we report a method for preparation of nano rutile, anatase, and bicrystalline (anatase + brookite) titania with an average crystal size of below 10 nm from aqueous TiOCl2 solution at low temperature by adjusting the pH value of the starting solution and adding different additives. Additionally we developed methodologies for the preparation of crystalline anatase nanoparticles with high photoactivity that did not involve hydrothermal synthesis or required calcination. Experimental Section Preparation. TiOCl2 aqueous solutions with a certain concentration were obtained by dropping known amounts of TiCl4 into 100 mL portions of distilled water in an ice-water bath. Then, the aqueous TiOCl2 solutions having various pH values controlled by adding 0.3 mol/L ammonium bicarbonate solution were aged to crystallize TiO2 precipitates at different temperatures in a temperature-controlled oil bath for several days (the detailed conditions are shown in Table 1). After crystallization, the precipitates formed in the solution were filtered, washed thoroughly with distilled water and then with acetone several times, and finally dried at room temperature in air for 24 h. To control the morphology and crystallinity of titania, 0.25 wt % octyl phenol poly(ethylene oxide) or 5.0 wt % ammonium sulfate was added to aqueous TiOCl2 solutions before aging and crystallizing in the temperature-controlled oil bath, respectively. Characterization. X-ray diffraction (XRD) patterns were obtained on a Rigaku Dmax X-ray diffractometer using Cu KR radiation. Scanning electron microscopy (SEM) images were obtained using a Hitachi S-4700. Transmission electron microscopy (TEM) images were obtained using a JEM-100CX electron microscope. The Brunauer-Emmett-Teller (BET) surface area was measured on a KICT-SPA3000 using N2 adsorption at -196 °C for the sample predegassing at 100 °C in vacuum for 10 h. Photoluminescence spectra were recorded at room temperature on a Shimadzu RF-5301 PC spectrometer using 300 nm excitation light. Photocatalytic Activity. The photocatalytic activity of assynthesized titania samples for the gas-phase oxidation of benzene was tested on a homemade recirculating gas-phase photoreactor with a quartz window, which was connected to a ppbRAE meter (RAE System Inc.) to recirculate a mixture of benzene and ambient air without additional drying and measure the concentration of the volatile organic compounds (VOCs). UV

Figure 1. XRD pattern of TiO2 samples prepared at 40 °C from TiOCl2 aqueous solution at different pH values by adding ammonium bicarbonate solution.

Figure 2. TEM image of as-synthesized titania prepared by hydrolysis of TiOCl2 in the absence of surfactant (a) and in the presence of surfactant (b). A (8W F8T5BLB lamp, Philips) black light was used as an irradiation source. First, 0.7000 g of titania powder was put into the reactor, and then a known amount of benzene was injected in the system under dark. After the absorption of benzene on titania reached absorption equilibrium, UV light was turned on.

Results and Discussion (14) Mogyorosi, K.; Dekany, I.; Fendler, J. H. Langmuir 2003, 19, 2938. (15) Kominami, H.; Matsuura, T.; Iwai, K.; Ohtani, B.; Nishimoto, S.; Kera, Y. Chem. Lett. 1995, 11, 693. (16) Hattori, A.; Yamamoto, M.; Tada, H.; Ito, S. Chem. Lett. 1998, 8, 707. (17) Sopyan, I.; Murasawa, S.; Hashimoto, K.; Fujishima, A. Chem. Lett. 1994, 4, 723.

Synthesis and Characterization. Table 1 shows the properties of as-prepared TiO2 prepared from TiOCl2 aqueous solution under various conditions. Figure 1 shows the effect of pH value on the crystallinity of as-synthesized titania. When the pH value was 0.14, nanosized pure rutile titania was obtained. Figures 2 and 3 show TEM and SEM

10840

Langmuir, Vol. 20, No. 25, 2004

Li et al.

Figure 3. SEM image of as-synthesized titania prepared by hydrolysis of TiOCl2 in the absence of surfactant (a) and in the presence of surfactants (b).

morphologies of nano rutile titania. It can be seen from Figure 2 that this nano rutile is needlelike with sizes of 50-70 nm by 5-12 nm. By careful observation, it could be found that the needlelike titania consists of many nano rutile spherical crystals aligned together (Figure 2a). All previous works could also only obtain needlelike or rodlike rutile titania by hydrolysis of TiOCl2 aqueous solution.5-9 It was confirmed theoretically by Zheng et al. that the formation of the nano needle-shaped rutile titania is attributed to strong HCl acid conditions in which the formation of needle-shaped titania with an appropriate ratio of length and diameter is thermodynamically favorable.18 In our work, it was found that adding a small amount of surfactant (0.25 wt % octyl phenol poly(ethylene oxide)) leads to the change of the morphology of the obtained titania sample from needlelike to nano spherical rutile crystals with a crystal size of 5-10 nm (Figure 2b). The SEM photos further confirm the observation by TEM. The obtained titania in the absence of surfactant has an irregularly arranged short rodlike structure by the aggregation of many nano titania particles (Figure 3a), but the obtained rutile titania in the presence of surfactant is aggregates composed of many small nano rutile particles. The possible reason for the morphology change in the presence of surfactants is as follows. Rutile crystals are first formed by hydrolysis of TiOCl2 and following crystallization. In the absence of surfactant, these nano rutile crystals are aligned together along one direction to form the needlelike sample. But in the presence of surfactant, hydrophilic poly(ethylene oxide) (PEO) moieties of surfactant are strongly absorbed on the surface of nano rutile crystals, and the hydrophobic moieties are directed to solution; thus aligning together of nano rutile crystals hinders the formation of needlelike titania samples by aligning together of nano rutile crystals. In the ranges of more than submicron sizes, polycrystalline rutile is thermodynamically stable relative to polycrystalline anatase and brookite. However, Zhang et al. reported that thermodynamic stability is particle size dependent, and at particle diameters below ca. 14 nm, anatase is more stable than rutile.10 This may explain why anatase can be synthesized as ultrafine particles and why there are few results concerning the preparation of nano rutile with crystal size below 14 nm. But by our method, it is very easy to synthesize nano rutile titania with particle size below ca. 14 nm (samples 2 and 3 in Table 1). When the pH value of TiOCl2 solution was adjusted to 2.7 by adding ammonium bicarbonate, the solution rapidly

gelled. The XRD results show that this gel is amorphous. Aging at 40 °C for several days, this amorphous gel is transformed to titania consisting of anatase phase with a small amount of brookite. Our experiments show that even aging for 1 week at ambient temperature, this amorphous TiO2‚xH2O gel could be still transformed to anatase titania with a small amount of brookite. Similar results are obtained when the pH value is between 2.7 and 5.0. However, when the pH value reaches 7.0, this amorphous TiO2‚xH2O gel could not be transformed to anatase phase at 40 °C. Usually, the amorphous-anatase transformation may complete in the temperature range from 250 to 400 °C. Ding et al. reported that the amorphous titania was converted into anatase and rutile titania after aging at room temperature for 1 year.19 Recently, Pottier et al. reported that hydroxylation of TiCl4 at room temperature led instantaneously to an amorphous titanium oxyhydroxide phase which crystallized as anatase upon aging at 60 °C.4 This report shows that the amorphous-anatase transformation could be fulfilled in shorter time even at room temperature. Yanagisawa et al. have made a detailed investigation into crystallization of anatase from amorphous titania by using the hydrothermal technique and realized the transformation of amorphous to anatase at temperature (120-250 °C).20 They found that water has a catalytic effect on the crystallization of amorphous titania, and the chloride ion accelerates the nucleation of the anatase. In our experiment, TiOCl2 aqueous solution is first neutralized by adding ammonium bicarbonate solution to pH ) 7.0; the obtained slurry contains a large amount of chloride ion, but there is no observed anatase titania even after the slurry is aged at 40 °C for 1 week. This observation shows that chloride ion does not accelerate the crystallization of amorphous titania. However, if the pH value of TiOCl2 aqueous solution is controlled in the range of pH ) 2.75.0, after aging at 40 °C for only 4 days, the amorphous titania is transformed to anatase with a small amount of brookite. Moreover, when TiOSO4 solution is used as the starting material instead of TiOCl2, by the same procedure similar results are obtained. Therefore, it is concluded from these observations that it is H+ or H3O+, not anions (e.g., Cl-), that catalyzes transformation of titania from amorphous to anatase. When the slurry was filtered and then dried at room temperature, the obtained dried amorphous titania could not transform to anatase after aging at 40 °C for 1 week. This result shows that water plays an important role in the crystallization of amorphous titania at lower temperature, which is similar to the

(18) Zheng, Y. Q.; Shi, E. W.; Yuan, R. L. Sci. China, Ser. E 1999, 29, 206.

(19) Ding, X. Z.; He, Y. Z. J. Mater. Sci. Lett. 1996, 15, 320. (20) Yanagisawa, K.; Ovenstone, J. J. Phys. Chem. B 1999, 103, 7781.

Nano Titania Powder with High Photoactivity

Figure 4. XRD pattern of TiO2 samples prepared at different aging temperatures from TiOCl2 aqueous solution at pH ) 2.7 or 7.0 by adding ammonium bicarbonate solution.

observation by Yanagisawa et al. Luo et al. thought that the appearance of brookite might be ascribed to the strong acid condition.21 Pottier et al. also thought that hydrothermal treatment in the presence of a strong acid condition benefited the formation of brookite and the high concentration of chloride in the thermolysis medium is necessary for the formation of brookite.22 In our experiment, only rutile titania is obtained by hydrolysis of aqueous TiOCl2 solution under a strong acid condition, but brookite appears under a weak acid condition at lower temperature. Moreover, when the pH value is 7.0, aging at higher temperature (120 °C) leads to amorphousanatase transformation, but there is almost no detected brookite in the as-synthesized anatase sample (Figure 4). This result further confirms that a weak acid condition is favorable to the formation of brookite. It can be seen from Figure 4 that with increasing aging temperature from room temperature to 120 °C the peak intensity at 2θ ) 25.4° increases and becomes sharpened, indicating that the crystallinity of as-synthesized bicrystalline titania does become better. It can be seen from Table 1 that the average crystal size of as-synthesized titania increases from 2.5 to 6.4 nm when the aging temperature increases from room temperature to 120 °C. Figure 5 shows the influence of SO42- on the crystallinity of titania. Our previous work showed that the existence of small amounts of SO42- in the initial TiOCl2 aqueous solution could prevent the formation of rutile titania under a strong acid condition (Figure 5a,b).23 In this work, it was found that adding a small amount of SO42- to the initial TiOCl2 aqueous solution also prevents the formation of brookite titania under weak acid conditions. It can be concluded from these results that the presence of a small amount of SO42- is unfavorable to the formation of both rutile and brookite. By this method, nano pure anatase titania with higher surface area (217 m2/g) and good crystallinity is obtained (sample no. 13 in Table 1). The average crystal size of samples 4 and 7 is smaller than that of other as-synthesized titania from XRD, but the BET surface area of the samples is smaller than the others (Table 1). We suppose that this can be attributed to the pretreatment at 100 °C before the measurement of the BET surface area, which results in easier ag(21) Luo, H.; Wang, C.; Yan, Y. Chem. Mater. 2003, 15, 3841. (22) Pottier, A.; Chaneac, C.; Tronc, E.; Mazerolles, L.; Jolivet, J. P. J. Mater. Chem. 2001, 11 (4), 1116. (23) Kim, S. J.; Park, S. Jpn. J. Appl. Phys. 2001, 40, 6797.

Langmuir, Vol. 20, No. 25, 2004 10841

Figure 5. XRD pattern of TiO2 samples prepared from TiOCl2 aqueous solution in the presence or absence of SO42- at different pH values by adding ammonium bicarbonate solution.

glomerating of titania crystals with smaller crystal size than those with larger size. Figures 6 and 7 show the TEM and SEM images of some of the as-synthesized titania samples. The particles of the obtained nano pure anatase and nano bicrystalline titania of anatase and brookite have a narrow distribution range of particle size. The diameters of these titania particles by TEM and SEM observations are in agreement with the average crystal sizes estimated by XRD. Mechanism of Titania Formation. After having made a detailed investigation into the crystallization of anatase from amorphous titania, Yanagisawa et al. proposed a mechanism of anatase and rutile nucleation.20 Zheng et al. proposed a dissolution-precipitation mechanism of the formation of titania by hydrolysis of TiCl4 under hydrothermal conditions and thought that the concentration of TiCl4 determined the crystallinity of the titania product, and anatase crystallites grew larger and transformed to rutile.24 Jolivet et al. gave a schematic pathway to explain the formation of brookite from the precursor Ti(OH)2Cl2(OH2)2 and thought that chloride ions played an important role in the formation of brookite.4a It seems that their mechanisms are not enough to give a comprehensive and reasonable explanation for our observations; for example, nano anatase cannot recrystallize into rutile, and vice versa, under our experimental conditions. Based on the above-mentioned mechanisms and our results, a mechanism for the crystallization of anatase and rutile is postulated (Figure 8). The key to the differences in rutile, anatase, and brookite formation stems from the structure difference of the three polymorphs: Rutile crystal structure is based on linear chains of TiO6 octahedra that share a pair of opposite edges and which are further linked by shared vertexes to form a 3D structure of 6:3 coordination. Anatase crystal structure is based on spiral chains of edge-sharing TiO6 octahedra that are further linked by sharing edges and corners to form a 3D structure of 6:3 coordination; each TiO6 octahedron in the brookite structure is in contact with 9 other ones (3 through edge sharing and 6 through corner sharing) through corner sharing between zigzag chains resulting in a distorted TiO6 octahedron.14,25 The phase transformation is accomplished by the rearrangement of (24) Zheng, Y. Q.; Shi, E. W.; Chen, Z. Z.; Li, W. J.; Hu, X. F. J. Mater. Chem. 2001, 11, 1547.

10842

Langmuir, Vol. 20, No. 25, 2004

Li et al.

Figure 6. TEM image of as-synthesized titania prepared from TiOCl2 aqueous solution at different conditions: (a) pH ) 2.7, 40 °C; (b) pH ) 2.7, 80 °C, 0.18 mol/L SO42-.

Figure 7. SEM image of as-synthesized titania prepared from TiOCl2 aqueous solution at different conditions: (a) pH ) 2.7, 40 °C; (b) pH ) 2.7, 80 °C, 0.18 mol/L SO42-.

Figure 8. The reaction scheme for crystalline titania formation.

these octahedrons. The rearrangement of these octahedrons at lower temperature proceeds by a liquid or liquidsolid state reaction. We agree with Yanagisawa et al.’s (25) Herry, M. Molecular Tectonics in Sol-Gel Chemistry. In Handbook of Organic-Inorganic Hybrid Materials and Nanocomposites. Vol. 1: Hybrid Materials; Nalwa, H. S., Ed.; American Scientific: Los Angeles, 2003; Chapter 1, pp 15-20.

suggestion that water molecules might form bridges between surface OH groups of different octahedrons that share only one common vertex, using the two lone pairs of electrons on the oxygen. Our experiment confirms that the presence of acid could accelerate the crystallization of anatase and/or brookite while the crystallization proceeds very slowly in the neutral condition. Therefore,

Nano Titania Powder with High Photoactivity

we think that the presence of H+ or H3O+ catalyzes dehydration of surface OH groups of neighboring octahedrons that share only one common vertex, thus linking them by sharing one edge. When another TiO6 octahedra complex attacks the two octahedrons sharing one edge, this H+ or H3O+ catalyzing dehydration process occurs again, and there are two connecting possibilities among the TiO6 octahedrons: one is the formation of linear chains of octahedrons by continuously sharing the opposite edge, which leads to the formation of rutile titania; another is the formation of spinal or zigzag chains of octahedrons, which results in the formation of anatase or brookite. Which possibility is preferred depends on the two factors, kinetic and thermodynamic. If under strong acid conditions the hydroxide groups in neighboring octahedrons sharing one edge are protonated, the electrostatic repulsion between the neighboring octahedrons is favorable to further sharing the opposite edge between neighboring octahedrons. Having thus aligned the octahedrons by sharing the opposite edge between neighboring octahedrons, and with the neighboring linear chains of TiO6 octahedra sharing a pair of opposite edges further dehydrated and linked by shared vertexes to form a 3D structure, formation of rutile will be kinetically favorable. Kinetically favorable formation of rutile under strong acid conditions is further demonstrated by our experiment. In our previous work9 and that of Cheng et al.,7 in the hydrolysis of TiOCl2 solution at higher temperature (above 60 °C), a mixture of anatase and rutile is formed. However, in this work, it is found that increasing the acidity of the TiOCl2 solution by adding hydrochloric acid leads to the formation of nano pure rutile titania even at higher temperature (see samples 5 and 6 in Table 1). The direction along the linear chains of TiO6 octahedra sharing a pair of opposite edges corresponds to the 001 orientation of rutile, and the growth along the [001] direction is kinetically favorable as compared to that along the other direction (e.g., [110]); therefore, rod or needlelike titania is obtained.26 But if some of the surfactant (e.g., octyl phenol poly(ethylene oxide)) exists in TiOCl2 solution, hydrophilic PEO moieties of surfactant could form complexes with TiO6 octahedra in the linear chain with the hydrophobic moieties directed to solution; thus the growth along the [001] direction is hindered. This is a possible reason nano titania particles are formed in the presence of surfactant. Under weak acid conditions, it seems that the formation of both rutile and anatase is possible. However, only anatase with a small amount of brookite is obtained under weak acid conditions (e.g., pH ) 2.7-5.0). Anatase is most thermodynamically stable at sizes less than 11 nm, brookite is most stable for crystal sizes between 11 and 35 nm, and rutile is most stable at sizes greater than 35 nm.10 In our experiment, the results of XRD, SEM, and TEM confirm that the crystal size of as-synthesized titania is below 11 nm. Therefore, we think that the formation of nano anatase with crystal size below 11 nm is thermodynamically favorable under weak acid conditions. In the view of thermodynamics, the more slowly the crystallization proceeds, the more favorable this thermodynamically favorable process. In our experiment, it was found that the formation of anatase from this amorphous gelled TiO2‚nH2O under weak acid conditions proceeds much more slowly than the formation of rutile from TiOCl2 solution under strong acid conditions at lower temperature. The formation of nano anatase with crystal size below 11 nm from this amorphous gelled TiO2‚nH2O under weak (26) Yang, K.; Zhu, J. M.; Zhu, J. J.; Huang, S. S.; Zhu, X. H.; Ma, G. B. Mater. Lett. 2003, 57 (30), 4639.

Langmuir, Vol. 20, No. 25, 2004 10843

Figure 9. The photoluminescence spectra for TiO2 prepared from TiOCl2 aqueous solution at different conditions: (a) pH ) 0.14, 40 °C (sample 1); (b) pH ) 0.14, 40 °C, 0.18 mol/L SO42(sample 4); (c) pH ) 2.7, 80 °C (sample 12); (d) pH ) 2.7, 80 °C, 0.18 mol/L SO42- (sample 13); (e) pH ) 2.7, 120 °C (sample 14).

acid conditions is more thermodynamically favorable than that of nano rutile. Why could rutile not be formed in the presence of SO42under strong acid conditions, and why does the presence of SO42- also prevent the formation of brookite? First, if in the presence of SO42-, SO42- containing two coordinating groups symmetrically chelates with the protonated neighboring octahedrons sharing one edge, when another TiO6 octahedra complex attacks the two octahedrons sharing one edge, only a spinal chain of octahedrons will be formed because of steric hindrance given by the tightly chelated SO4 group. Therefore, the formation of anatase is kinetically favorable in the presence of SO42- under strong acid conditions. Second, we observe that the presence of SO42slows the reaction rate of the crystallization because of its stronger binding to the titania surface, which is in agreement with the observation of Yanagisawa et al.20 It was also observed that prolonging aging time decreases the content of brookite in final titania products. Ding et al. also observed that there was no brookite in their product after the amorphous titania was aged at room temperature for 1 year.19 Therefore, we think that prolonging crystallization in the presence of SO42- is thermodynamically favorable to the formation of nano anatase and the phase transformation of nano anatase with crystal size below 11 nm from nano brookite. Photoluminescence (PL) Spectra. The PL emission mainly results from the recombination of excited electrons and holes, and the lower PL intensity indicates the decrease in recombination rate.27-30 Figure 9 shows the room-temperature photoluminescence spectra for assynthesized TiO2 in the range of 350-550 nm. For nano spherical rutile titania, three emission peaks appear at about 408.5, 433.5, and 465.0 nm wavelengths, which are equivalent to 3.02, 2.85, and 2.66 eV, respectively. The former is ascribed to the emission of the band gap transition of rutile titania (3.0 eV for anatase).29,30 The second and third are emission signals originating from the charge-transfer transition from Ti3+ to oxygen anion (27) Li, X. Z.; Li, F. B.; Yang, C. L.; Ge, W. K. J. Photochem. Photobiol., A 2001, 141, 209. (28) Li, F. B.; Li, X. Z. Appl. Catal., A 2002, 228, 15. (29) Li, F. B.; Li, X. Z. Chemosphere 2002, 48, 1103. (30) Yu, J. G.; Yu, H. G.; Cheng, B.; Zhao, X. J.; Yu, J. C.; Ho, W. K. J. Phys. Chem. B 2003, 107, 13871.

10844

Langmuir, Vol. 20, No. 25, 2004

Figure 10. The amounts of total VOCs with UV A light irradiation time on TiO2 prepared from TiOCl2 aqueous solution at different conditions: (a) pH ) 2.7, 80 °C, 0.18 mol/L SO42(sample 13); (b) TiO2 (P25); (c) pH ) 2.7, 120 °C (sample 14); (d) pH ) 2.7, 80 °C (sample 12); (e) pH ) 0.14, 40 °C, 0.18 mol/L SO42- (sample 4); (f) pH ) 2.7, 40 °C (sample 8); (g) pH ) 0.14, 40 °C (sample 1).

in a TiO68- complex.29 For nano anatase and bicrystalline titania of anatase and brookite, five emission peaks appear at about 365.5, 381.0, 405.0, 433.5, and 465.0 nm wavelengths, which are equivalent to 3.38, 3.24, 3.05, 2.85, and 2.66 eV, respectively. The emission peak at 381.0 nm is ascribed to the emission of the band gap transition of anatase titania (3.2 eV for anatase). The emission peak at 365.5 nm is attributed to the following two aspects: the emission of the band gap transition of brookite titania (3.4 eV)31 and the blue shift (15 nm or so) of the band gap transition of the as-synthesized nano anatase titania with the average crystal size of 4.0-9.0 nm because of the wellknown quantum-size effect for semiconductors.32 The other peaks at higher wavelength are also the emission signal originating from the charge-transfer transition from Ti3+ to oxygen anion in a TiO68- complex. The energy difference between the band gap energy and the other lower emission peak energy is caused by the Stokes shift due to the Franck-Condon effect.33,34 The variation of PL intensity may result from the change of defect state on the shallow level of the TiO2 surface.28,35 The difference of the PL spectra in Figure 9 is due to various preparation conditions leading to different phase structure and surface. The highest PL intensity is observed for as-synthesized nano spherical rutile, which shows that there are many crystalline defects in the nano spherical rutile. For bicrystalline titania of anatase and brookite, with increasing aging temperature, there is an obvious decrease in the intensity of the PL spectra. At the same aging temperature, adding SO42- to TiOCl2 aqueous solution leads to the decrease of PL intensity of the final produced titania. Photocatalytic Activity. Carcinogenic benzene existing in gasoline, paints, and so forth is one of the main air pollutants. In this work, we chose gas-phase photodetoxication of benzene to evaluate the as-synthesized titania photocatalysts. Figure 10 shows the amounts of total VOCs (31) Koelsch, M.; Cassaignon, S.; Guillemoles, J. F.; Jolivet, J. P. Thin Solid Films 2002, 403-404, 312. (32) Anpo, M.; Shima, T.; Kodama S.; Kubokawa, Y. J. Phys. Chem. 1987, 91, 4305. (33) Fujihara, K.; Izumi, S.; Ohno, T.; Matsumura, M. J. Photochem. Photobiol., A 2000, 132, 99. (34) Serpone, N.; Lawless, D.; Khairutdinov, R. J. Phys. Chem. 1995, 99, 16646. (35) Toyoda, T.; Hayakawa, T.; Abe, K.; Shigenari, T.; Shen, Q. J. Lumin. 2000, 87-89, 1237.

Li et al.

with the UV A light irradiation time. The as-synthesized needlelike nano rutile titania has the lowest photoactivity for gas-phase photo-oxidation of benzene. It is also found that there is no obvious difference in photoactivity between needlelike nano rutile and spherical titania. Nanosized pure anatase titania prepared from TiOCl2 aqueous solution in the presence of SO42- at pH ) 2.7 and 80 °C shows the highest photoactivity, even better than Degussa P25 titania, a benchmarking photocatalyst. For nanosized bicrystalline titania of anatase and brookite, increasing aging temperature from room temperature to 120 °C leads to the rapid elevation of photoactivity although surface area decreases. It can be seen from Figure 10 that for as-synthesized titania prepared at same aging temperature, nanosized anatase titania has a better photoactivity than rutile titania and nano bicrystalline titania of anatase. It seems that a small amount of brookite in anatase is unfavorable to the improvement of its photoactivity. It was reported that the lower PL intensity could be ascribed to the lower recombination rate of photogenerated electrons and holes under light irradiation, which leads to the higher photocatalytic activity of the sample.27-30 The lowest photoactivity of nano rutile titania is attributed to its lower surface area and highest PL intensity. Nanosized pure anatase titania prepared from TiOCl2 aqueous solution in the presence of SO42- at pH ) 2.7 and 80 °C has high surface area (217 m2/g) and good crystallinity. The photoluminescence experiment shows that this catalyst has a lower recombination rate of photogenerated electrons and holes. All these factors contribute to its highest photocatalytic activity. For nanosized bicrystalline titania of anatase and brookite, enhanced photocatalytic activity with increasing aging temperature is due to the better crystallinity of assynthesized bicrystalline titania of anatase and brookite, which results in lower lattice defects, at which the photoproduced electron-hole pairs are easy to recombine. Conclusion Nano rutile, anatase, and bicrystalline titania of anatase and brookite with an average crystal size of below 10 nm have been prepared from aqueous TiOCl2 solution at low temperatures by adjusting the pH value of the starting solution and adding different additives. The morphology of nano rutile titania can be controlled by adding a small amount of octyl phenol poly(ethylene oxide) into aqueous TiOCl2 solution, which leads to the change of the morphology of the obtained titania sample from short needlelike to nano spherical rutile crystals. Amorphous-anatase transformation of titania can proceed in liquid-solid reaction at lower temperature, even at room temperature. H+ or H3O+ plays a catalytic role in the transformation of titania from amorphous to anatase. The presence of a small amount of SO42- hinders the formation of both rutile and brookite. Increasing aging temperature is favorable to improve the photoactivity of the final produced titania. By careful adjustment of preparation conditions, nano pure anatase with higher surface area, good crystallinity, and a lower recombination rate of photoexcited electrons and holes is obtained. This nano pure anatase shows a very good photocatalytic activity for gas-phase oxidation of benzene. Acknowledgment. The authors are grateful to the Basic Research Program of the Korea Science & Engineering Foundation (Grant No. R01-2002-000-00338) for financial support. LA0489716