Characterization of HNbWO6 and HTaWO6 Metal Oxide Nanosheet

---

J. Phys. Chem. C 2009, 113, 7831–7837

7831

Characterization of HNbWO6 and HTaWO6 Metal Oxide Nanosheet Aggregates As Solid Acid Catalysts Caio Tagusagawa,† Atsushi Takagaki,‡ Shigenobu Hayashi,§ and Kazunari Domen*,† Department of Chemical System Engineering, School of Engineering, The UniVersity of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo, 113-7656, Japan, School of Materials Science, Japan AdVanced Institute of Science and Technology (JAIST), 1-1 Asahidai, Nomi, Ishikawa, 923-1292, Japan, and Research Institute of Instrumentation Frontier, National Institute of AdVanced Industrial Science and Technology (AIST), Central 5, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan ReceiVed: January 18, 2009; ReVised Manuscript ReceiVed: March 26, 2009

Nanosheet aggregates prepared from protonated layered tungstates HMWO6 (M ) Nb, Ta) are examined as potential solid acid catalysts. The nanosheet aggregates are formed by soft chemical processing of the layered compound with tetra(n-butylammonium) hydroxide, and the catalytic activity and acid strength of the aggregates are compared with those for HTiNbO5, HNb3O8, and a range of conventional solid acids. The catalytic activity for the Friedel-Crafts alkylation of anisole in the presence of benzyl alcohol increases in the order HTiNbO5 < HNb3O8 < HMWO6 (M ) Nb, Ta), consistent with the acid strengths determined by desorption measurements and nuclear magnetic resonance spectroscopy. Nuclear magnetic resonance spectroscopy indicates that the acid catalytic activity of the nanosheet aggregates is attributable to strong Brønsted acid sites, presumably M(OH)M′ (M ) Ti, Nb, Ta; M′ ) Nb, W). 1. Introduction Growing concerns regarding environmental and human health issues have necessitated the development of new industrial reactions with minimal environmental impact.1,2 The present authors have been developing a low-impact solid-phase replacement for liquid acid catalysts such as sulfuric acid and nitric acid, which are used in a wide variety of industrial chemical reactions. These liquid acids are not reusable, resulting in large volumes of liquid waste requiring neutralization and disposal, and product separation is costly and inefficient. A reusable and nontoxic solid acid with strong acidity is a suitable replacement for such liquid catalysts.3 H+-exchanged forms of cationexchangeable transition metal layered oxides, in which H+ ions are located between two-dimensional (2D) transition metal oxide anion sheets, are potentially useful as strong solid acids. However, as the high charge density of the oxide sheets prevents reactants from penetrating into the interlayer region, unmodified layered transition metal oxides are generally ineffective as solid acid catalysts. Previous studies on layered transition metal oxides have revealed that layered metal oxides can be applied as solid acid catalysts after exfoliation and aggregation of the nanosheets constituting the layered material.4-7 Exfoliation of cationexchangeable layered metal oxides (HTiNbO5, HTi2NbO7, HTiTaO5, and HNb3O8) in aqueous solution affords colloidal single-crystal metal oxide sheets that precipitate under acidic conditions to form aggregates of nanosheets. These nanosheet aggregates possess high specific surface area (ca. 100 m2 g-1) and exhibit high catalytic activity for acid-catalyzed reactions such as the esterification of acetic acid, the hydrolysis of ethyl acetate, and dehydration of D-xylose.4-7 * To whom corresponding author should be addressed. Phone: +81-3-5841-1148. Fax: +81-3-5841-8838. E-mail: domen@ chemsys.t.u-tokyo.ac.jp. † The University of Tokyo. ‡ JAIST. § AIST.

The strong acidity of the HTiNbO5, HTi2NbO7, and HTiTaO5 aggregated-nanosheet catalysts has been shown to be attributable to bridging hydroxyl groups (Ti(OH)M; M ) Nb5+ or Ta5+), representing a novel type of strong Brønsted acid site.4,5 The strong acidity of the HNb3O8 nanosheet catalyst has similarly been attributed to the formation of Nb(OH)Nb hydroxyl groups. HSr2Nb3O10, on the other hand, has a perovskite structure bearing isolated Nb-OH groups and exhibits no or only slight activity for acid-catalyzed reactions even in aggregated nanosheet form.6 These results indicate that the exfoliation and subsequent aggregation procedure for such protonated layered compounds does not always result in high activity for acid catalysis. Rather, the formation of bridging hydroxyl groups, Ti(OH)M (M ) Nb5+ or Ta5+) or Nb(OH)Nb, which are intrinsic to the crystal structure of the two-dimensional sheets, is considered essential for the preparation of nanosheet catalysts with strong acidity. In the present study, nanosheet aggregates prepared from layered metal oxides with the composition HMWO6 (M ) Nb or Ta) are examined as novel solid acid catalysts. The nanosheet catalysts are prepared from the lithium form of the layered metal oxide, LiMWO6 (M ) Nb or Ta), by proton exchange and subsequent exfoliation with tetra(n-butylammonium) hydroxide (TBA+OH-).8-11 HNbWO6 results will be compared with HTiNbO5 and HNb3O8 in order to investigate the effect caused by changing the metal used to form bridging hydroxyl with niobium. HTaWO6 with an analogous structure to HNbWO6 is also examined. The structures of layered LiNbWO6 and LiTaWO6 are compared with those for layered KTiNbO5 and KNb3O8 in Figure 1. Layered LiNbWO6 and LiTaWO6 exhibit the same trirutile-type structure formed by 2D MWO6- (M ) Nb or Ta) anion sheets composed of MO6 (M ) Nb or Ta) and WO6. Layered KTiNbO5 and KNb3O8 are formed by 2D TiNbO5- or Nb3O8- anion sheets composed of NbO6 and TiO6 octahedra. From the crystal structure, HMWO6 is expected to form bridged hydroxyl groups, M(OH)W (M ) Nb5+ or Ta5+), as

10.1021/jp900525a CCC: $40.75  2009 American Chemical Society Published on Web 04/10/2009

7832

J. Phys. Chem. C, Vol. 113, No. 18, 2009

Figure 1. Schematic structures of layered KTiNbO5, KNb3O8, LiNbWO6, and LiTaWO6

strong acid sites, similar to HTiNbO5 and HNb3O8. The acid catalytic activity of the present tungstate-based oxide nanosheets is examined through liquid-phase Friedel-Crafts alkylation, and the results are compared with those for other nanosheet-based solid acids (HTiNbO5 and HNb3O8) and conventional solid acids. The acid proprieties of these solid acids are evaluated by NH3 temperature-programmed desorption (NH3-TPD) and 1H and 31P magic-angle spinning nuclear magnetic resonance (MAS NMR) spectroscopy. 2. Experimental Section 2.1. Preparation of Nanosheet Aggregates. Layered HTiNbO5, HNb3O8, HNbWO6, and HTaWO6 were prepared by proton exchange of the precursors KTiNbO5, KNb3O8, LiNbWO6, and LiTaWO6, which were obtained by calcination according to the procedure described in the literature.8-13 The proton-exchange reaction was performed by shaking 2.0 g of the lithium or potassium form in 150 mL of 1-2 M acid solution (HNO3) at room temperature for 2 weeks, exchanging

Tagusagawa et al. the acid solution twice over that period. The product was then washed with distilled water and dried in air at 343 K. The nanosheets constituting the layered oxides (TiNbO5-, Nb3O8-, NbWO6-, and TaWO6- nanosheets) were separated by adding 15 wt % tetra(n-butylammonium) hydroxide (TBA+OH-) solution to 150 mL of distilled water containing 2.0 g of the protonated compound. The TBA+OH- solution was added to the suspension until the pH reached 9.5-10.0, and the resulting solution was shaken for 2 weeks. The insertion of voluminous and hydrophilic TBA+ cations expands and hydrates the interlayer spaces, resulting in the exfoliation of metal oxide sheets. After shaking, the suspension was centrifuged, and the supernatant solution containing the dispersed nanosheets was collected. The addition of HNO3 aqueous solution (0.1 M, 30 mL) to 150 mL of the nanosheet solution resulted in immediate aggregation of the nanosheets as a precipitate. The aggregated nanosheet samples were then rinsed three times with 150 mL of 0.1 M HNO3 aqueous solution to remove TBA+ and then with 150 mL of distilled water to remove HNO3, the complete removal of which was confirmed by elemental analysis. 2.2. Characterization. Samples were characterized by X-ray diffraction (XRD; RINT-UltimaIII, Rigaku), scanning electron microscopy (SEM; S-4700, Hitachi), and N2 desorption (BEL Japan, BELSORP-miniII). The acid properties of the samples were determined by NH3-TPD, using a TPD-1-AT instrument (BEL Japan) equipped with a quadrupole mass spectrometer. In TPD measurements, 20 mg of the sample was heated at 423 K for 1 h under helium flow, exposed to NH3 at 373 K for adsorption, removed of the excess NH3 by exposing the sample at 373 K for 1 h, and finally heated at 2 deg min-1. 1 H and 31P MAS NMR spectra were measured at room temperature with Bruker ASX400 and MSL400 spectrometers at Larmor frequencies of 400.13 and 162.0 MHz, respectively. Bruker MAS probeheads were used in combination with a 4 mm zirconia rotor operated at a sample spinning rate of 8 or 10 kHz. The 1H chemical shift was referenced to neat tetramethylsilane by using the signal of adamantane (1.87 ppm at a spinning rate of 8 kHz). For 31P NMR, a single-pulse sequence with high-power proton decoupling was employed, and the 31P chemical shift was determined in reference to 85% H3PO4 (at 0.0 ppm). (NH4)2HPO4 was employed as the second reference material (1.33 ppm), and trimethylphosphine oxide (TMPO) was employed as a probe molecule. TMPO-adsorbed samples were prepared by evacuated dehydration at 423 K for 1 h followed by immersion in tetrahydrofuran (THF) solution containing TMPO (0.333 mol L-1 THF solution) at room temperature for 2 days in a glovebox under argon. After evacuation to remove the THF solvent, the samples were packed in a rotor housed in a glovebox under N2. 2.3. Acid-Catalyzed Reaction. The acid catalytic activity of layered HTiNbO5, HNb3O8, HNbWO6, HTaWO6, and the corresponding aggregated-nanosheet catalysts were determined through Friedel-Crafts alkylation, which was performed with 0.2 g of the catalyst, 100 mmol of anisole, 10 mmol of benzyl alcohol, and n-decane as an internal standard. The reaction vessel was placed in an oil bath maintained at 373 K for the duration of the 2 h reaction. The products (benzylanisole and dibenzyl ether) were analyzed by flame ionization gas chromatography (GC-2014, Shimadzu), using a capillary column (J&W ScientificDB-FFAP, length 30 m, i.d. 0.25 mm and film 0.25 µm), helium carrier gas, injector temperature of 443 K, detector temperature of 493 K, and the following temperature program: 308 K (3 min), 15 deg min-1 f 473 K (6 min). The activities of niobic acid (Nb2O5 · nH2O), Nb2O5-WO3, Ta2O5-WO3, ion-exchange

HNbWO6 and HTaWO6 Nanosheets As Solid Acid Catalysts

J. Phys. Chem. C, Vol. 113, No. 18, 2009 7833

Figure 2. SEM images of (a) layered HTiNbO5, (b) HTiNbO5 nanosheets aggregate, (c) layered HNb3O8, (d) HNb3O8 nanosheets aggregate, (e) layered HNbWO6, (f) HNbWO6 nanosheets aggregate, (g) layered HTaWO6, and (h) HTaWO6 nanosheets aggregate

resins (Amberlyst-15 and Nafion NR50), and H-type zeolites (H-Beta: SiO2/Al2O3 ) 25, JRC-Z-HB25; and H-ZSM5: SiO2/ Al2O3 ) 90, JRC-Z-5-90H supplied from Catalysis Society of Japan (Japan Reference Catalyst)) were also determined for comparison. Nb2O5-WO3 and Ta2O5-WO3 were prepared as follows.14 Nb2O5 · nH2O (AD/3512, n ) 3.8) supplied by CBMM and (NH4)10W12O41 · 5H2O were dissolved in oxalic acid solution and distilled water, respectively, at 353 K. The solutions were then mixed and stirred vigorously at 353 K. After drying, the obtained materials were calcined at 773 K for 3 h to yield Nb-W mixed oxides. The tantalum-containing mixed oxides were prepared in a similar manner with Ta2O5 as the starting material. 3. Results and Discussion 3.1. Structures of Layered Metal Oxides and Aggregated Nanosheets. Figure 2 shows SEM images of HTiNbO5,

TABLE 1: BET Surface Area over Metal Oxides SBET/m2 g-1 layered HTiNbO5 layered HNb3O8 layered HNbWO6 layered HTaWO6

1 1 1 2

SBET/m2 g-1 HTiNbO5 nanosheets HNb3O8 nanosheets HNbWO6 nanosheets HTaWO6 nanosheets

153 101 66 47

HNb3O8, HNbWO6, HTaWO6, and the corresponding aggregated-nanosheet catalysts. The as-prepared layered metal oxides consist of tabular particles 3-20 µm in size, with a Brunauer-Emmet-Teller (BET) surface of 1-2 m2 g-1. The layered HNbWO6 and HTaWO6 samples consist of smaller tabular particles than the nontungstate HTiNbO5 and HNb3O8 samples, attributable to the dissimilar calcination conditions employed in the preparation of the precursors (1073 K for 24 h and 1373 K for 30 h, respectively). The SEM images of the aggregated-nanosheet precipitates indicate that the addition of acid (H+) results in the random

7834

J. Phys. Chem. C, Vol. 113, No. 18, 2009

Tagusagawa et al.

Figure 3. XRD patterns for (a) layered HTiNbO5, (b) HTiNbO5 nanosheets aggregate, (c) layered HNb3O8, (d) HNb3O8 nanosheets aggregate, (e) layered HNbWO6, (f) HNbWO6 nanosheets aggregate, (g) layered HTaWO6, and (h) HTaWO6 nanosheets aggregate.

aggregation of nanosheets with the expected composition (confirmed by energy-dispersed X-ray spectroscopy). The BET surface areas of the aggregated-nanosheet precipitates are 153 m2 g-1 for HTiNbO5, 101 m2 g-1 for HNb3O8, 66 m2 g-1 for HNbWO6, and 47 m2 g-1 for HTaWO6 as shown in Table 1, and the precipitate is free of the tabular particles observed in the original layered compounds.4-6 The XRD patterns for the layered metal oxides and the corresponding aggregated-nanosheet precipitates are shown in Figure 3. The XRD pattern for the HTiNbO5 nanosheet precipitate has a relatively weak diffraction peak at low angle (2θ < 10°), indicating that the periodic layered structure of the original HTiNbO5 has been partially destroyed. However, the presence of diffraction peaks due to in-plane diffractions (e.g., (200)) for the HTiNbO5 nanosheet catalyst indicates that the two-dimensional structure is preserved upon precipitation. The HNb3O8 precipitate also has a weaker diffraction peak at low angle (2θ < 10°), and the in-plane diffraction peaks (e.g., (101)) are similarly retained upon exfoliation-aggregation. Layered HNbWO6 and HTaWO6 have the same crystal structure,11 and upon exfoliation-aggregation form the same poorly periodic structure with the partial absence of diffraction peaks at low angle ( HNbWO6 · HNb3O8 > HTiNbO5, which correlates with the trend in maximum temperature of the NH3-TPD curves assigned to M(OH)M′ acid sites. 3.5. Acid Catalytic Activity. The results for the Friedel-Crafts alkylation of anisole in the presence of benzyl alcohol with use of the present catalysts are listed in Table 3. The original layered metal oxides (HTiNbO5, HNb3O8, HNbWO6, and HTaWO6) do not catalyze these reactions, as expected,4-6 with alkylated products absent or afforded at yields of less than 1%. However, the exfoliated and aggregated materials, other than HTiNbO5, exhibit significant activity for this reaction. The HNb3O8 aggregated-nanosheet precipitate has relatively low activity for this reaction at 373 K, but exhibits remarkable activity at 423 K, as previously reported.6 Among the present aggregatednanosheet materials, HTaWO6 exhibits the highest yield at 373 K (58.4%), comparable to that for Amberlyst-15 and Nafion NR50. The catalytic activity of the nanosheet precipitates

HNbWO6 and HTaWO6 Nanosheets As Solid Acid Catalysts decreases in the order HTaWO6 . HNbWO6 > HNb3O8 . HTiNbO5, consistent with the acid strength determined by NH3-TPD. The nanosheet materials consisting of group-5 elements Nb5+ and Ta5+ and the group-6 element W6+ afford higher benzylanisole yields and higher NH3 desorption temperatures than the nanosheet precipitate bearing the group-4 element Ti4+. Reactions were also performed with niobic acid (Nb2O5 · nH2O), Nb2O5-WO3, Ta2O5-WO3, zeolites (H-Beta, H-ZSM5), and ion-exchange resins (Nafion NR50, Amberlyst15). Although the catalytic activity of the HNbWO6 nanosheet precipitate is comparable to that for Nb2O5-WO3,14 the HTaWO6 precipitate and Ta2O5-WO3 exhibit markedly different catalytic activities, suggesting that the homogeneous distribution of metal cations (Ta5+ and W6+), resulting from the twodimensional crystal structure of the nanosheets, contributes to activity. The turnover frequencies (TOFs) of the effective acid sites (strong acid sites) on the nanosheet precipitates and the conventional solid acids are listed in Table 3. The TOF for the HTaWO6 nanosheet precipitate is estimated to be 104 h-1, higher than that for Nafion NR50 (17.5 h-1) and H-ZSM5 (13.1 h-1). Selectivity and yield of benzyl anisole at 2 h tend to follow the same order as that observed for the acid strength of aggregated nanosheets. As mentioned in the Experimental Section, dibenzyl ether forms as a byproduct during the Friedel-Crafts alkylation, when the catalyst possesses Lewis acid sites, 19,20 decreasing the selectivity on benzyl anisole. Table 4 summarizes the results obtained for the acid catalytic activity and acid properties of the present exfoliated-aggregated nanosheet catalysts. The nanosheet catalysts with higher activity for Friedel-Crafts alkylation possess stronger acid sites, as confirmed by NH3-TPD and 31P MAS NMR spectroscopy. It can also be seen from these results that the strength of Brønsted acid sites in the nanosheet aggregates is closely related to the combination of metal cations, increasing in the order Ti4+-(OH)-Nb5+ < Nb5+-(OH)-Nb5+ < Nb5+-(OH)-W6+ < Ta5+-(OH)-W6+. 4. Conclusion Nanosheet aggregates of HNbWO6 and HTaWO6 were examined as solid acid catalysts suitable for catalyzing reactions such as Friedel-Crafts alkylation. The HTaWO6 aggregatednanosheet catalyst was found to function as a strong solid acid catalyst, exceeding the activity of niobic acid, zeolites, and ionexchange resins. The acid strengths determined by NH3-TPD and 31P MAS NMR were found to be consistent with the activities for Friedel-Crafts alkylation, decreasing in the order HTaWO6 > HNbWO6 > HNb3O8 > HTiNbO5. Catalysts with

J. Phys. Chem. C, Vol. 113, No. 18, 2009 7837 the aggregated nanosheet structure were found to have higher acid strength dependent on the combination of metal oxides, decreasing in the order Ta5+(OH)W6+ > Nb5+(OH)W6+ > Nb5+(OH)Nb5+ > Ti4+(OH)Nb5+. Acknowledgment. This work was supported by the Development in a New Interdisciplinary Field Based on Nanotechnology and Materials Science program of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan and the Global Center of Excellence Program for Chemistry. References and Notes (1) Anastas, P. T.; Williamson, T. C. In Green Chemistry; Anastas, P. T., Williamson, T. C., Eds.; American Chemical Society: Washington, D.C., 1996; pp 1-20. (2) Clark, J. H. Green Chem. 1999, 1, 1. (3) (a) Okuhara, T. Chem. ReV. 2002, 102, 3641. (b) Tanabe, K.; Ho¨lderich, W. F. Appl. Catal., A 1999, 181, 399. (4) Takagaki, A.; Sugisawa, M.; Lu, D.; Kondo, J. N.; Hara, M.; Domen, K.; Hayashi, S. J. Am. Chem. Soc. 2003, 125, 5479. (5) Takagaki, A.; Yoshida, T.; Lu, D.; Kondo, J. N.; Hara, M.; Domen, K.; Hayashi, S. J. Phys. Chem. B 2004, 108, 11549. (6) Takagaki, A.; Lu, D.; Kondo, J. N.; Hara, M.; Hayashi, S.; Domen, K. Chem. Mater. 2005, 17, 2487. (7) Dias, A. S.; Lima, S.; Carriazo, D.; Rives, V.; Pillinger, M.; Valente, A. A. J. Catal. 2006, 244, 230. (8) (a) Blasse, G.; de Pauw, A. D. M. J. Inorg. Nucl. Chem. 1970, 32, 3960. (b) Fourquet, J. L.; Le Bail, A.; Gillet, P. A. Mater. Res. Bull. 1988, 23, 1163. (9) (a) Bhat, V.; Gopalakrishnan, J. Solid State Ionics 1988, 26, 25. (b) Bhuvanesh, N. S. P.; Prasad, B. R.; Subramanian, C. K.; Gopalakrishnan, J. Chem. Commun. 1996, 289. (10) Kinomura, N.; Kumada, N. Solid State Ionics 1992, 51, 1. (11) Shaak, R. E.; Mallouk, T. E. Chem. Commun. 2002, 706. (12) (a) Wadsley, A. D. Acta Crystallogr. 1964, 17, 623. (b) Rebbah, H.; Desgardin, G.; Raveau, B. Mater. Res. Bull. 1979, 14, 1125. (13) (a) Gasperin, P. M. Acta Crystallogr. 1982, B38, 2024. (b) Nassau, K.; Shiever, J. W.; Bernstein, J. L. J. Electrochem. Soc. 1969, 116, 348. (c) Nedjar, R.; Borel, M. M.; Raveau, B. Mater. Res. Bull. 1985, 20, 1291. (14) (a) Yamashita, K.; Hirano, M.; Okumura, K.; Niwa, M. Catal. Today 2006, 118, 385. (b) Okumura, K.; Yamashita, K.; Hirano, M.; Niwa, M. Chem. Lett. 2005, 34, 716. (c) Okumura, K.; Yamashita, K.; Hirano, M.; Niwa, M. J. Catal. 2005, 234, 300. (15) Rakiewicz, E. F.; Peters, A. W.; Wormsbecher, R. F. J. Phys. Chem. B 1998, 102, 2890. (16) Kao, H. M.; Yu, C. Y.; Yeh, M. C. Microporous Mesoporous Mater. 2002, 53, 1. (17) Zhao, Q.; Chen, W. H.; Huang, S. J.; Wu, Y. C.; Lee, H. K.; Liu, S. B. J. Phys. Chem. B 2002, 106, 4462. (18) (a) Harmer, M. A.; Sun, Q. Appl. Catal. 2001, 221, 45. (b) Matsuhashi, H.; Arata, K. Phys. Chem. Chem. Phys. 2004, 6, 2529. (19) Cativiela, C.; Garcı´a, J.; Garcı´a-Matres, M.; Mayoral, J. A.; Figueras, F.; Fraile, J.; Cseri, T.; Chiche, B. Appl. Catal., A 1995, 123, 273. (20) Shishido, T.; Kitano, T.; Teramura, K.; Tanaka, T. Catal. Lett. 2009, 129, 383.

JP900525A