Phase-pure TiO2 nanoparticles: anatase, brookite

IOP PUBLISHING

NANOTECHNOLOGY

Nanotechnology 19 (2008) 145605 (10pp)

doi:10.1088/0957-4484/19/14/145605

Phase-pure TiO2 nanoparticles: anatase, brookite and rutile D Reyes-Coronado1, G Rodr´ıguez-Gattorno1, M E Espinosa-Pesqueira2, C Cab1 , R de Coss1 and G Oskam1,3 1

Departamento de F´ısica Aplicada, CINVESTAV-I. P. N., Mérida, Yucatán, 97310, México 2 Departamento de Tecnolog´ıa de Materiales, Instituto Nacional de Investigaciones Nucleares (ININ), Ocoyoacac, México. 52750, México E-mail: [email protected]

Received 22 October 2007, in final form 12 January 2008 Published 5 March 2008 Online at stacks.iop.org/Nano/19/145605 Abstract We report on the synthesis of phase-pure TiO2 nanoparticles in anatase, rutile and brookite structures, using amorphous titania as a common starting material. Phase formation was achieved by hydrothermal treatment at elevated temperatures with the appropriate reactants. Anatase nanoparticles were obtained using acetic acid, while phase-pure rutile and brookite nanoparticles were obtained with hydrochloric acid at a different concentration. The nanomaterials were characterized using x-ray diffraction, UV–visible reflectance spectroscopy, dynamic light scattering, and transmission electron microscopy. We propose that anatase formation is dominated by surface energy effects, and that rutile and brookite formation follows a dissolution–precipitation mechanism, where chains of sixfold-coordinated titanium complexes arrange into different crystal structures depending on the reactant chemistry. The particle growth kinetics under hydrothermal conditions are determined by coarsening and aggregation–recrystallization processes, allowing control over the average nanoparticle size. S Supplementary data are available from stacks.iop.org/Nano/19/145605

transform to rutile [19, 21]. Secondly, the crystal structure stability has been explained on the basis of a molecular picture, where the nucleation and growth of the different polymorphs of TiO2 are determined by the precursor chemistry, which depends on the reactants used [23–29]. A complicating factor in the understanding of nanoparticle formation is the multitude of experimental conditions used for synthesis of the different TiO2 phases, making it difficult to compare mechanisms. Synthesis of anatase nanoparticles by solution-phase methods in aqueous environments has been reported for a large variety of experimental conditions: using an alkoxide precursor, anatase is formed by reaction in aqueous solutions of a variety of acids or bases. Phase-pure anatase nanoparticles with diameters ranging from 6–30 nm are generally prepared from titanium (IV) isopropoxide and acetic acid [16]. When stronger acids are used, a fraction of the product usually consists of brookite nanoparticles [18, 30]. Larger anatase particles are difficult to synthesize due to transformation to rutile upon increasing treatment times and/or temperature. Phase-pure brookite particles (0.3–1 μm) have

1. Introduction Titanium dioxide (TiO2 ) nanomaterials are used in a wide range of applications such as (photo)catalysis, separations, sensor devices, paints, and dye-sensitized solar cells [1–4]. The material properties of TiO2 nanoparticles are a function of the crystal structure, nanoparticle size, and morphology and, hence, are strongly dependent on the method of synthesis [5–15]. TiO2 exists in three main phases: anatase, brookite and rutile. As a bulk material, rutile is the stable phase; however, solution-phase preparation methods for TiO2 generally favour the anatase structure [8, 14–18]. These observations are attributed to two main effects: surface energy and precursor chemistry. At very small particle dimensions, the surface energy is an important part of the total energy and it has been found that the surface energy of anatase is lower than those of rutile and brookite [19–22]. Surface energy considerations accurately describe the observation of a crossover size of about 30 nm where anatase nanoparticles 3 Author to whom any correspondence should be addressed.

0957-4484/08/145605+10$30.00

1

© 2008 IOP Publishing Ltd Printed in the UK

Nanotechnology 19 (2008) 145605

D Reyes-Coronado et al

been prepared using amorphous titania as the starting material and hydrothermal treatment with NaOH [31]. The mechanism is related to the formation of a sodium titanate, which subsequently is transformed to pure TiO2 with the brookite structure. The synthesis of brookite nanoparticles of 5–10 nm has been reported by thermolysis of TiCl4 in aqueous HCl solution [29]. The composition of the reaction product was found to be strongly dependent on the Ti:Cl concentration ratio, and up to 80% pure brookite was reported at Ti:Cl = 17– 35. The brookite nanoparticles could be separated by selective precipitation of rutile. At higher Ti:Cl ratio, pure rutile particles were obtained which were generally much larger and rod-shaped. Phase-pure rutile nanoparticles have been prepared from TiCl4 or TiCl3 in HCl solution or from titanium (IV) isopropoxide in nitric acid at pH = 0.5 [26, 32–35]. Recently, several authors have compared synthesis methods for the three phases, in order to determine the effect of crystal structure on the physical properties [27, 36, 37]. An advanced understanding of the parameters that determine phase formation is crucial for the successful technological application of nanomaterials. In this work, we determine the influence of the reactant chemistry in solutionphase synthesis of TiO2 on the formation of crystalline nanomaterials using amorphous titania as a starting material, with as two main goals: (i) to gain insight into the phase formation mechanisms, and (ii) to provide experimental methods for the controlled synthesis of phase-pure anatase, rutile, and brookite nanoparticles with control over the nanoparticle dimensions.

Figure 1. Schematic flow chart illustrating the steps in the synthesis pathway of phase-pure TiO2 nanomaterials.

The synthesis products were characterized using x-ray diffraction (XRD) with a D-5000 Siemens diffractometer with Cu Kα radiation and a Ni filter. The particle radius, r , was obtained using the Scherrer equation for the peak width, B , corrected for instrument broadening according to

2. Experimental details

r=

Figure 1 shows a flow chart of the proposed synthesis method for anatase, rutile and brookite nanomaterials. As a first step amorphous titania was synthesized; this was subsequently used as a starting material for the formation of the three crystal structures by hydrothermal treatment under different experimental conditions, such as the acid used, the pH, and temperature. Amorphous titania was prepared by dropwise addition of a solution of 1.14 ml of water (Labconco WaterPro PS; 18 M cm) in 105 ml of 2-propanol (J.T. Baker, 99.9%) to a solution of 5 ml of titanium (IV) isopropoxide (Aldrich, 97%) in 105 ml of 2-propanol (J.T. Baker 99.9%) under stirring at 0 ◦ C. The solution was stirred for 24 h at room temperature, and filtered to yield a white paste of amorphous titania. Water was added to the clear filtrate, resulting in a second precipitation product, which was filtered off to give a second batch of amorphous titania. The second batch was generally used for the synthesis of pure brookite nanoparticles. The amorphous titania paste was diluted to 0.3 M in aqueous solutions of acetic acid (Reasol 99.75%) or HCl (Aldrich, 37%) and sealed in a Teflon cup, contained in a pressure vessel (Parr Instruments). The experiments were performed for different acid concentrations and temperatures in order to establish the conditions for obtaining phase-pure products. The duration of the hydrothermal treatment at the optimal temperature in each system was varied in order to determine the kinetics of particle growth.

0. 9λ , 2 B cos θ

(1)

where λ = 0.154 nm, and θ is the reflection angle [38]. Transmission electron microscopy (TEM) samples were prepared by allowing a drop of a dilute colloid on HC-coated Cu grids to evaporate, and high resolution TEM (HRTEM) and high angle angular dark field (HAADF) images were obtained on a JEM 2010 microscope. Colloidal samples were evaluated using an Agilent 8453 UV–vis diode array spectrophotometer, and dynamic light scattering (DLS) was performed using a Malvern HPPS-Z sizer. Samples for UV–vis reflectance spectroscopy were prepared by evaporation of solvent and by-products at low temperature (40 ◦ C), and collecting the pure titania powder. An Ocean Optics (USB2000) fibre optics spectrophotometer was used for the measurements. The reflectance spectra were analysed using the Kubelka– Munk formalism to convert the reflectance into the equivalent absorption coefficient, αKM , according to [39–41]

αKM =

(1 − R ∞ )2 , 2 R∞

(2)

where R∞ is the reflectance of an infinitely thick sample with respect to a reference at each wavelength. Ab initio calculations were implemented to determine the electronic structure of the three phases, and the optical properties were obtained in terms of the absorption 2



and brookite can be obtained under the following conditions: anatase was prepared using 1.5 M acetic acid at 200 ◦ C, rutile was obtained using 4 M HCl at 200 ◦ C, and brookite was obtained with 3 M HCl at 175 ◦ C. Under these conditions the phase purity was over 95% as determined from the integrated intensity of the XRD peaks for the phases [20]. Figure 2 shows an example of x-ray diffraction spectra for the synthesized products of each system. For anatase and rutile, no other phases were observed, while in the case of brookite a small fraction of rutile ( 1 where the amorphous titania is insoluble (see for example figure S4 of the supplementary data available at stacks.iop.org/Nano/19/145605); hence, dissolution– precipitation processes are expected to be negligible. These results suggest that anatase is formed due to the lower surface energy as compared to the other phases. The lower surface energy affects phase formation in terms of thermodynamics and kinetics. Thermodynamically, anatase is the structure with the lowest total energy if the nanoparticles are sufficiently small. According to classical nucleation theory the homogeneous nucleation rate, JN , is given by the following expression [67, 68]:

Figure 12 shows TEM images of the brookite phase after hydrothermal treatment (175 ◦ C for 6.5 h) illustrating that nanoparticles are oblate and elongated along the [010] axis with an aspect ratio ranging from 1.5 to 2.5. The mean particle size is 14 nm (see figure S3(c) in the supplementary data (available at stacks.iop.org/Nano/19/145605)), which is very close to the size estimated from XRD. Higher magnification images show several nanoparticles consisting of a crystalline brookite core and thin amorphous-like shell (highlighted with arrows). The shell consists of a short-range ordered phase with ˚ which coincides with a mean interplanar distance of 5.4 A, the distance between octahedron chains in brookite along the [001] axis, as illustrated in figure 13. The observation of a short-range ordering in amorphous material in the vicinity of brookite nanoparticles obtained by a very different synthesis route has been reported before [66]. These observations suggest that brookite grows from an amorphous-like phase that consists of small oligomers having a high grade of condensation along the edge-sharing octahedron chains. 3.5. Mechanisms for phase formation

Anatase is easily synthesized by transformation of amorphous TiO2 ; in fact, it is obtained by hydrothermal treatment of amorphous titania with water without any additional reactant. Water

JN = J0 exp 8

−G c RT

= J0 exp

−16π Vm γ 3 3(RT )3 (ln S)2

,

(6)



where J0 is a pre-factor determined by the frequency of collisions and the type of reaction, G c is the energy needed to form a nucleus of critical size, and S is the supersaturation. Hence, a lower surface energy significantly accelerates the nucleation kinetics, thus favouring the formation of anatase. At high hydrochloric acid concentrations and pH < 0, titania becomes soluble, suggesting that dissolution–precipitation processes can become important. The partial dissolution of the amorphous titania in 3–5 M HCl is the first step in a reaction pathway for the dissolution–recrystallization processes previously proposed in the literature [24]. Interestingly, synthesis of rutile and brookite nanoparticles is achieved under very similar conditions (high acid concentration and similar temperatures) suggesting that recrystallization processes involve different precursor chemistry for each phase. With increasing temperature the solubility of titania tends to decrease in relation to the increasing condensation rate [69, 70]; therefore, amorphous solid is expected to be present in equilibrium with soluble species of titania (e.g., TiO2+ and chlorocomplexes) prior to crystallization. Speciation diagrams for titania chloro-complexes show that in 3 M HCl the complex [Ti(OH)2 Cl(OH2 )3 ]+ with only one coordinated chloride ion predominates at room temperature (see figure S4 in the supplementary data (available at stacks.iop.org/Nano/19/145605)). At higher acid concentration this complex is displaced by another complex containing four coordinated chloride ions: [Ti(OH)2 Cl4 ]2− . Homogeneous nucleation is usually assumed to involve zero-charge complexes [67]; hence, it is likely that the eventual precursors for nucleation consist of two or more complexes, combined with water as a ligand. With the increase in the number of coordinated chloride ions with higher HCl concentration, the titania chloro-complex becomes more symmetric. Hence, chlorolysis–oxolation processes lead to brookite in the presence of an asymmetric chloro-complex, [Ti(OH)2 Cl(OH2 )3 ]+ , while rutile is formed in the presence of the more symmetric [Ti(OH)2 Cl4 ]2− . This reaction pathway could explain the high sensitivity of the formation of different titania phases to the HCl concentration. Structural details of brookite and rutile favour this inference. Figure 13 shows that the rutile structure is essentially very similar to that of brookite, with both phases formed by straight 2μ2 polyhedron chains linked through corner sharing. An important difference between the two phases is that the symmetry of octahedra in rutile is higher than that in brookite, and that the linking of 2μ2 chains occurs through trans-oxo bridges, and in brookite through cis-oxo bridges. In summary, we report the synthesis of anatase, rutile, and brookite nanoparticles from amorphous titania by hydrothermal treatment at elevated temperatures with the appropriate reactants. We propose that anatase formation is dominated by surface energy effects, while the formation of rutile and brookite follows a dissolution–precipitation mechanism, where chains of sixfold-coordinated titanium complexes arrange into different crystal structures. The particle growth kinetics under hydrothermal conditions are determined by coarsening and aggregation–recrystallization processes, allowing control over the average nanoparticle

size. Rutile synthesis tends to result in relatively low surface area material due to the formation of compact aggregates. The anatase and brookite nanomaterials consist of porous aggregates, and are suitable for application in photocatalysis, sensors, and dye-sensitized solar cells.

Acknowledgment Financial support from the Consejo Nacional de Ciencia y Tecnolog´ıa (CONACYT, Mexico) under grant number 43828Y is gratefully acknowledged.

References [1] Hagfeldt A and Grätzel M 1995 Chem. Rev. 95 49 [2] Hoffmann M R, Martin S T, Choi W and Bahnemann D W 1995 Chem. Rev. 95 69 [3] Linsebigler A L, Lu G and Yates J T Jr 1995 Chem. Rev. 95 735 [4] Chen X and Mao S S 2007 Chem. Rev. 107 2891 [5] Matijevic E 1986 Langmuir 2 12 [6] Sugimoto T 1987 Adv. Colloid Interface Sci. 28 65 [7] Livage J, Henry M and Sanchez C 1988 Prog. Solid State Chem. 18 259 [8] Anderson M A, Gieselmann M J and Xu Q 1988 J. Membr. Sci. 39 243 [9] Barringer E A and Bowen H K 1985 Langmuir 1 414 Barringer E A and Bowen H K 1985 Langmuir 1 420 [10] Jean J H and Ring T A 1986 Langmuir 2 251 [11] Look J-L and Zukoski C F 1992 J. Am. Ceram. Soc. 75 1587 Look J-L and Zukoski C F 1995 J. Am. Ceram. Soc. 78 21 [12] Bokhimi X, Morales A, Novaro O, Lopez T, Sanchez E and Gomez R 1995 J. Mater. Res. 10 2788 [13] Vorkapic D and Matsoukas T 1999 J. Colloid Interface Sci. 214 283 [14] Penn R L and Banfield J F 1999 Geochim. Cosmochim. Acta 63 1549 [15] Barbé C J, Arendse F, Comte P, Jirousek M, Lenzmann F, Shklover V and Grätzel M 1997 J. Am. Ceram. Soc. 80 3157 [16] Zaban A, Aruna S T, Tirosh B A, Gregg B A and Mastai Y 2000 J. Phys. Chem. B 104 4130 [17] Zhang Z, Wang C-C, Zakaria R and Ying J Y 1998 J. Phys. Chem. B 102 10871 [18] Oskam G, Nellore A, Penn R L and Searson P 2003 J. Phys. Chem. B 107 1734 [19] Zhang H and Banfield J 1998 J. Mater. Chem. 8 2073 [20] Zhang H and Banfield J 2000 J. Phys. Chem. B 104 3481 [21] Navrotsky A 2003 Geochem. Trans. 4 34 [22] Naicker P K, Cummings P T, Zhang H and Banfield J F 2005 J. Phys. Chem. B 109 15243 [23] Cheng H, Ma J, Zhao Z and Qi L 1995 Chem. Mater. 7 663 [24] Yanagisawa K, Yamamoto Y, Feng Q and Yamasaki N 1998 J. Mater. Res. 13 825 [25] Yanagisawa K and Ovenstone J 1999 J. Phys. Chem. B 103 7781 [26] Aruna S T, Tirosh S and Zaban A 2000 J. Mater. Chem. 10 2388 [27] Li J-G, Ishigaki T and Sun X 2007 J. Phys. Chem. C 111 4969 [28] Yin H, Wada Y, Kitamura T, Kambe S, Murasawa S, Mori H, Sakata T and Yanagida S 2001 J. Mater. Chem. 11 1694 [29] Pottier A, Chaneac C, Tronc E, Mazerolles L and Jolivet J-P 2001 J. Mater. Chem. 11 1116 [30] Wu M, Lin G, Chen D, Wang G, He D, Feng S and Xu R 2002 Chem. Mater. 14 1974 [31] Nagase T, Ebina T, Iwasaki T, Hayashi H, Onodera Y and Chatterjee M 1999 Chem. Lett. 9 911

9



[32] Wang C C and Ying J Y 1999 Chem. Mater. 11 3113 [33] Yin H, Wada Y, Kitamura T, Kambe S, Murasawa S, Mori H, Sakata T and Yanagida S 2001 J. Mater. Chem. 11 1694 [34] Park N-G, Van de Lagemaat J and Frank A J 2000 J. Phys. Chem. B 104 8989 [35] Wang W, Gu B, Liang L, Hamilton W and Wesolowski D 2004 J. Phys. Chem. B 108 14789 [36] Koelsch M, Cassaignon S, Thanh Minh C T, Guillemoles J-F and Jolivet J-P 2004 Thin Solid Films 451 86 [37] Tomita K, Petrykin V, Kobayashi M, Shiro M, Yoshimura M and Kakihana M 2006 Angew. Chem. Int. Edn 45 2378 [38] Cullity B D 1978 Elements of X-ray Diffraction (Reading, MA: Addison-Wesley) [39] Kubelka P 1948 J. Opt. Soc. Am. 38 448 [40] Yang L and Kruse B 2004 J. Opt. Soc. Am. A 21 1933 [41] Yang L and Miklavcic S J 2005 J. Opt. Soc. Am. A 22 1866 [42] Sanchez-Portal D, Ordejon P, Artacho E and Soler J M 1997 Int. J. Quantum Chem. 65 453 [43] Hohenberg P and Kohn W 1964 Phys. Rev. 136 B864 [44] Kohn W and Sham L J 1965 Phys. Rev. 140 1133 [45] Perdew J P, Burke K and Ernzerhof M 1996 Phys. Rev. Lett. 77 3865 [46] Yu P Y and Cardona M 2005 Fundamentals of Semiconductors (Berlin: Springer) p 250 [47] Luccarini V, Saarinen J, Peiponen K and Vartiainen E 2005 Kramers–Kronig Relations in Optical Materials Research (Berlin: Springer) [48] Pascual J, Camassel J and Mathieu H 1978 Phys. Rev. B 18 5606 [49] Amtout A and Leonelli R 1995 Phys. Rev. B 51 6842 [50] Glassford K M and Chelikowsky J R 1992 Phys. Rev. B 45 3874 [51] Kavan L, Grätzel M, Gilbert S E, Klemenz C and Scheel H J 1996 J. Am. Chem. Soc. 118 6716

[52] Asahi R, Taga Y, Mannstadt W and Freeman A J 2000 Phys. Rev. B 61 7459 [53] Hosaka N, Sekiya T, Aatoko C and Kurita S 1997 J. Phys. Soc. Japan 66 877 [54] Hosaka N, Sekiya T, Fujisawa M, Satokob C and Kurita S 1996 J. Electron Spectrosc. Relat. Phenom. 78 75 [55] Serpone N, Lawless D and Khairutdinovt R 1995 J. Phys. Chem. 99 16646 [56] Posternak M, Baldereschi A, Walter E J and Krakauer H 2006 Phys. Rev. B 74 125113 [57] Zhang Y-F, Lin W, Li Y, Ding K-N and Li J-Q 2005 J. Phys. Chem. B 109 19270 [58] Mo S-D and Ching W Y 1995 Phys. Rev. B 51 13023 [59] Lifshitz I M and Slyozov V V 1961 J. Phys. Chem. Solids 19 35 [60] Wagner C 1961 Z. Elektrochem. 65 581 [61] Talapin D V, Rogach A L, Haase M and Weller H 2001 J. Phys. Chem. B 105 12278 [62] Penn R L, Oskam G, Strathmann T J, Searson P C, Stone A and Veblen D R 2001 J. Phys. Chem. B 105 2177 [63] Maira A J, Yeung K L, Lee C Y, Yue P L and Chan C K 2000 J. Catal. 192 185 [64] Banfield J F and Veblen D R 1992 Am. Mineral. 77 545 [65] Ribeiro C, Vila C, Stroppa D B, Mastelaro V R, Bettini J, Longo E and Leite E R 2007 J. Phys. Chem. C 111 5871 [66] Bakardjieva S, Stengl V, Szatmary L, Subrt J, Lukac J, Murafa N, Niznansky D, Cizek K, Jirkovsky J and Petrova N 2006 J. Mater. Chem. 16 1709 [67] Jolivet J-P 2003 Metal Oxide Chemistry and Synthesis—From Solution to Solid State (New York: Wiley) [68] Oskam G 2006 J. Sol-Gel Sci. Technol. 37 161 [69] Narita E, Takeuchi H, Horiguchi N and Okabe T 1984 Bull. Chem. Soc. Japan 57 1388 [70] El-Hazek N, Lasheen T A, El-Sheikh R and Zaki S A 2007 Hydrometallurgy 87 45

10