Article pubs.acs.org/JPCC
Synergistic Catalysis by Lewis Acid and Base Sites on ZrO2 for Meerwein−Ponndorf−Verley Reduction Tasuku Komanoya,†,‡ Kiyotaka Nakajima,†,§ Masaaki Kitano,∥ and Michikazu Hara*,†,‡,⊥ †
Materials and Structures Laboratory, ‡Frontier Research Center, and ∥Materials Research Center for Element Strategy, Tokyo Institute of Technology, Midori-ku, Yokohama 226-8503, Japan § Precursory Research for Embryonic Science and Technology (PRESTO) and ⊥Advanced Low Carbon Technology Research and Development Program (ALCA), Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi 332-0012, Japan S Supporting Information *
ABSTRACT: Early transition metal oxides, TiO2, ZrO2, and Nb2O5, were studied as heterogeneous catalysts for the Meerwein− Ponndorf−Verley (MPV) reduction of cyclohexanone in 2-propanol. Despite a small amount of Lewis acid sites and weak Lewis acid strength, ZrO2 was clearly superior to TiO2 and Nb2O5 with respect to reaction rate. Fourier transform infrared spectroscopy (FT-IR) and temperature-programmed-desorption (TPD) measurements revealed that ZrO2 has large amounts of base sites that activate the methylene groups in 2-propanol bonded to Lewis acid sites. Various analyses, including experiments using isotopic 2-propanol, suggest that efficient MPV reduction over ZrO2 is due not only to Lewis acid strength and density, but also to a synergistic effect of base and Lewis acid sites.
1. INTRODUCTION Meerwein−Ponndorf−Verley (MPV) reduction is a traditional redox technique through the intermolecular hydride transfer between ketone and alcohol,1,2 which enables carbonyl-selective hydrogenation even in the presence of easily reduced groups, such as halogens, nitriles, alkoxyl groups, and unsaturated carbon moieties. Therefore, inter- and intramolecular MPVtype hydrogen transfers have been widely applied to various organic syntheses for the production of useful chemicals, especially in the fields of asymmetric synthesis3,4 and biomass conversion.5−9 Since the first discovery of the MPV reduction catalyzed by aluminum alkoxides,10−13 the use of solid catalysts has been widely investigated,14−37 due to high durability and reusability, which are necessary for sustainable chemical production.38 To date, an amphoteric oxide, represented by ZrO2,14−25 has been found to function as an efficient heterogeneous catalyst for various MPV reactions. It is generally accepted that the metal center that acts as a Lewis acid site facilitates the formation of a six-membered ring transition state with ketones and alcohols to accomplish the hydride transfer.39 The formation of the six-membered ring intermediate is regarded as the rate-determining step for MPV reduction.39 This concerted pathway is energetically favorable compared to stepwise transfer hydrogenation involving βhydride elimination of alcohols via the formation of metal hydride species, which is typically observed over metallic Ru and Ir catalysts.40,41 The reaction is largely dependent on Lewis acid sites as catalytically active sites. In addition, weak base sites, such as hydroxyl groups on basic and amphoteric oxide © 2015 American Chemical Society
surfaces, have a positive effect to increase the overall reaction rate.18,32−37 The role of base sites has been proposed to assist the formation of alkoxide species with Lewis acid sites by the deprotonation of alcohol molecules17,42 at the initial stage of the reaction. However, no direct evidence has been reported so far. Recently, we have reported that unsaturated coordination Nb and Ti species on their oxide surfaces can act as Lewis acid sites in water.6,9,43,44 While certain amounts of undesirable byproducts are simultaneously evolved over Nb2O5 and TiO2, they function as effective catalysts for the hydride transfer of pyruvaldehyde into lactic acid in water.9 This hydride transfer reaction is considered to proceed on the Lewis acid sites via the MPV reduction mechanism.45 On the other hand, early transition metal oxides, excluding ZrO 2, are generally considered not to be effective for typical intermolecular MPV reduction, such as cyclohexanone with 2-propanol (2-PrOH). However, TiO2 and Nb2O5 have similar Lewis acidity to ZrO2. In this study, the roles of base and Lewis acid sites on heterogeneous catalysts were investigated through the MPV reduction of cyclohexanone with 2-PrOH using ZrO2, TiO2, and Nb2O5. Received: August 27, 2015 Revised: November 8, 2015 Published: November 9, 2015 26540
DOI: 10.1021/acs.jpcc.5b08355 J. Phys. Chem. C 2015, 119, 26540−26546
Article
The Journal of Physical Chemistry C
2. EXPERIMENTAL SECTION
3. RESULTS AND DISCUSSION Figure S1 shows XRD patterns and N2 adsorption−desorption isotherms for the ZrO2, Nb2O5, and TiO2 catalysts. ZrO2 (surface area 42 m2 g−1) is a mixture of monoclinic and tetragonal phases. The XRD pattern for TiO2 (120 m2 g−1) is assignable to anatase TiO2, and that for Nb2O5 (110 m2 g−1) indicates no obvious structure. The N2 adsorption−desorption isotherms for ZrO2, Nb2O5, and TiO2 indicate that TiO2 and Nb2O5 have mesoporous structures, which result in larger surface areas than that for ZrO2 without a porous structure. Table 1 summarizes the activities of the tested catalysts for the MPV reduction of cyclohexanone in 2-PrOH at 393 K for 2
2.1. Catalyst Preparation. A 10 g sample of zirconium tetrapropoxide (70 wt % solution in 1-propanol, Aldrich) was added to 100 mL of distilled water at room temperature. After the solution was stirred for 2 h, a white precipitate was collected by filtration, washed with 2 L of distilled water, and dried at 373 K for several hours. The dried sample was then calcined at 673 K for 8 h in air to obtain ZrO2 powder. TiO2 was also prepared by the sol−gel reaction of Ti(OiPr)4 (Kanto Chemical) as a Ti source in water. Commercially available Nb2O5 (Companhia Brasileira de Metallurgia e Mineraçaõ ) was utilized after calcination of the as-received oxide at 673 K for 8 h in an air atmosphere. 2.2. Catalytic Reaction. A mixture of the catalyst (heterogeneous catalyst, 0.1 g; homogeneous catalyst, 10 μmol), cyclohexanone (0.16 mmol), and 2-PrOH (2 mL, dried with MS4A) was stirred in a glass tube reactor at 363− 393 K for several hours. The products in the resulting solution were analyzed using gas chromatography (GC; Shimadzu GC2014, flame ionization detector) with a DB-FFAP column (0.25 mm × 30 m × 0.25 μm). Decane was used as an internal standard. Four types of deuterium-labeled 2-PrOH ((CH3)2CDOH (CIL), (CH3)2CHOD (Merck), (CD3)2CHOH, and (CD3)2CDOD (Aldrich)) were also used as substrates for the MPV reduction to study the reaction mechanism with the tested oxides. 2.3. Catalyst Characterization. Powder X-ray diffraction (XRD; Rigaku Ultima IV) was conducted using Cu Kα radiation (40 kV, 40 mA). The patterns were compared with references in the Joint Committee for Powder Diffraction Standards (JCPDS) to determine their crystalline structure. N2 adsorption−desorption isotherms (Quantachrome Nova4200e) were measured at 77 K (0.050 ≤ p/p0 ≤ 0.995). Prior to measurements, the catalysts were pretreated at 423 K for 1 h under vacuum to remove physisorbed water and gases. The isotherms were analyzed by the Brunauer−Emmett−Teller (BET) method in the range 0.050 ≤ p/p0 ≤ 0.300 to calculate the specific surface area. Fourier transform infrared spectroscopy (FT-IR; Jasco FT-IR-6100) measurements were conducted with an extended KBr beam splitting device and a mercury cadmium telluride (MCT) detector. Self-supported disks (20 mm diameter, ca. 20 mg) of the oxides were placed into an IR cell attached to a closed circulation system. Prior to measurement, the disks were dried at 573 K for 1 h under vacuum to remove physisorbed water and gases. Pyridine, cyclohexanone, and 2-PrOH were used as probe molecules for the characterization of acid and base sites in the catalysts.9 In the case of cyclohexanone and 2-PrOH adsorption, the sample disk was exposed to saturated vapor (>4 kPa) at room temperature and was then evacuated to remove weakly physisorbed molecules. The obtained spectra were normalized on the basis of the sample weight. CO2 temperatureprogrammed desorption (CO2-TPD; BEL Japan BELCAT-A) was performed at a heating rate of 10 K min−1 from 373 to 1273 K under a He flow (30 mL min−1). After dehydration of the samples at 673 K for 1 h under He flow, CO2 gas (5% CO2/ He, 30 mL min−1) was contacted with the dehydrated samples (50 mg) at 373 K for 0.5 h. Outlet gas was detected with a mass analyzer (BEL Japan BELMass) (m/z = 44) to monitor desorption of CO2 from the samples.
Table 1. Catalytic Conversion of Cyclohexanone to Cyclohexanol in 2-PrOHa entry
catalyst
convb (%)
yieldc (%)
1 2 3 4d 5e 6e
Nb2O5 TiO2 ZrO2 H2SO4 N(CH2CH3)3 Sc(OTf)3
14 12 81 9 99
10 11 80