The effect of the crystal planes of metal oxides on their catalytic performance is well known and has generated a significant number of research studies [1, 2, 3, 4, 5]. CeO2 is a particularly promising metal oxide because it readily forms oxygen vacancies and potentially has a wide range of applications, including as an active species, a promoter and a heterogeneous catalyst [6, 7, 8, 9]. The crystal plane effect in CeO2 primarily originates from differences in oxygen vacancies and has been widely studied with regard to gas phase reactions occurring at high temperatures (>200 °C) and involving small molecules such as COx, H2O, H2, CH4 and NOx [5, 10-14]. It has been reported that the oxygen vacancies in this material are mobile at high temperatures [15], which makes it challenging to study the crystal plane effect at elevated temperatures. Therefore, an alternative means of assessing this effect is required, and one such method involves the study of liquid phase organic reactions at low temperatures. In previous research, we have studied the acid properties of the different crystal planes of CeO2 during hydrolysis reactions [16]. However, the redox properties of the various crystal planes during reactions involving large molecules at low temperatures have not yet been examined.
Imines represent an important class of N-based reaction intermediates and are widely used in the synthesis of various biological, agricultural and pharmaceutical compounds [17, 18]. Traditionally, imines are synthesized by the condensation reactions of aldehydes or ketones with amines in the presence of an acid catalyst. More recently, new methodologies allowing the green synthesis of imines have been developed, such as the oxidative dehydrogenation of amines [19], the hydroamination of alkynes with amines [20] and the dehydrogenative or oxidative coupling of alcohols and amines [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40]. The dehydrogenation or oxidative coupling method is one of the most promising approaches, because alcohols are both readily available and inexpensive, and because only hydrogen or water may be produced as by-products. CeO2 has been reported to act as an efficient catalyst for this reaction at room temperature [41]. The high activity of CeO2 for this reaction is derived from the reactive oxygen species in its oxygen vacancies [1], although the redox properties of the various crystal planes of CeO2 remain unclear.
Herein, we report the crystal plane effects in CeO2 for the oxidative coupling of alcohols and amines. The results demonstrate that CeO2(110) possesses a greater concentration of oxygen vacancies than either CeO2(111) or CeO2(100), thus the highest activity was observed over CeO2(110). These findings provide insights that should assist in the tuning of the low-temperature redox properties of CeO2-based catalysts.
All chemicals were of analytical purity and were used as-received without further purification. Ce(NO3)3·6H2O was purchased from Aladdin Chemicals. MgO, Y2O3, Nb2O5, Al2O3, ZrO2, SiO2, TiO2, Cr2O3, CrO3, Fe2O3, CuO, ZnO, MoO3, WO3 and V2O5 were obtained from commercial sources and the majority were purchased from the Aladdin Chemicals.
CeO2 was synthesized by a conventional precipitation method [16]. In this process, Ce(NO3)3·6H2O (5.0 g) was dissolved in 100 mL of deionized water and the pH of the solution was adjusted to 11.0 by the addition of NH4OH (3.4 mol/L) at room temperature with magnetic stirring. The resulting precipitate was filtrated and washed with deionized water, dried in an oven overnight at 120 °C, then calcined at 500 °C for 4 h.
Ce(NO3)3·6H2O was used as the cerium source for the synthesis of different CeO2 morphologies. The preparation method was similar to the hydrothermal process reported by Yan’s group [42]. Briefly, Ce(NO3)3·6H2O (0.868 g) and the requisite amount of NaOH (0.016-15 g) were dissolved in 5 and 35 mL of deionized water, respectively, after which these two solutions were mixed in a Teflon bottle with stirring at room temperature. The mixture was stirred for an additional 30 min, promoting the formation of a milky slurry. The Teflon bottle was subsequently inserted into a tightly sealed stainless steel vessel autoclave and the autoclave was transferred into an oven and subjected to hydrothermal treatment at 100-180 °C for 24 h. After cooling, the precipitate was separated by centrifugation and washed repeatedly with deionized water and ethanol, followed by drying in an oven overnight at 60 °C.
In a typical reaction, the catalyst was weighed into a screw-capped glass pressure vessel containing a stir bar. The required amounts of alcohol (0.5 mmol) and amine (0.6 mmol), pre-dissolved in solvent (2 mL), were added and the vessel was filled with oxygen, sealed, and heated to the desired temperature in an oil bath with stirring. The products of the reaction were analyzed by gas chromatography (GC, Agilent 7890A) and GC-mass spectrometry (MS, Agilent 7890A/5975C).
Powder X-ray diffraction (XRD) patterns were acquired using a PANalytical X-Pert diffractometer, applying Cu-Kα radiation at 40 kV and 40 mA. Continuous scans were collected over the 2θ range of 5° to 80°. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images were recorded on a JEOL-2100F electron microscope operating at 200 kV. Raman spectra were recorded on a micro-Raman spectrometer (Renishaw) equipped with a CCD detector, using a He/Ne laser with a wavelength of 532 nm.
Initially, the synthesis of an imine from benzyl alcohol and aniline was conducted over various solid metal oxides at 60 °C for 12 h (Table 1). The reaction was found not to occur in the absence of the catalyst (Table 1, entry 1). Other metal oxides either showed low benzyl alcohol conversions (Al2O3, V2O5, MgO, Fe2O3, Y2O3, Nb2O5, ZnO and ZrO2; Table 1, entries 3-10) or poor imine selectivities (SiO2, TiO2, Cr2O3, CrO3, CuO, MoO3 and WO3; Table 1, entries 11-17). In contrast, CeO2 exhibited the best performance, giving >99% conversion of benzyl alcohol and 94% selectivity for the imine (3) (Table 1, entry 2). The solvent also had a remarkable effect on the reaction (Table 1, entries 18-23), such that the low polarity solvent p-xylene showed better performance (Table 1, entry 2). More polar solvents, such as acetonitrile, 2-propanol, methanol, dioxane, dimethyl formamide (DMF) and dimethyl sulfoxide (DMSO), gave lower conversions. The reaction did not progress beyond 9% conversion in Ar atmosphere, indicating that the presence of oxygen is necessary to obtain suitable yields (Table 1, entry 24). Following removal of the CeO2 from the reaction mixture by hot filtration, no further conversion of benzyl alcohol was observed (Fig. 1), demonstrating the heterogeneous nature of the catalyst.
a Reaction conditions: solvent (2 mL), catalyst (100 mg), 1 (0.5 mmol), 2 (0.6 mmol), 60 °C, 12 h, O2. b The conversions, based on benzyl alcohol consumption, were determined by GC and the products were identified by GC-MS. c Under Ar.
A remarkable deactivation of the CeO2 was observed during attempts to recycle the catalyst. Following one reaction, the catalyst was filtered from the reaction mixture, washed with ethanol three times, dried and then used for the next cycle. In the subsequent trial, the yield of imine decreased to 35%. In addition, a significant change in the color of the catalyst was observed, from pale yellow to greyish yellow. The catalytic activity could, however, be recovered by calcination of the used catalyst in air (Fig. 2). Therefore, the deactivation of the catalyst may be attributed to poisoning of the active sites by the adsorption of organic species.
After screening for the most suitable catalyst and optimizing the reaction conditions, we focused on the scope of substrates and the functional group tolerance (Table 2). Various alcohols and amines underwent oxidative coupling smoothly to afford the desired products with moderate to excellent yields. The scope of the amines to which the reaction could be applied was extended under the optimized reaction conditions when using benzyl alcohol (3a-3p). A wide range of aniline derivatives (3a-3g) gave the corresponding imines with excellent yields (>95% conversions and >96% selectivities). In the case of N,N′-dimethylbenzene-1,4-diamine (3h), however, the corresponding imine yield was 48% even at elevated temperature (120 °C). Moreover, 2,6-dimethyl-substituted (3i) and 2-chloro-substituted (3j) anilines gave low yields of imines, presumably because of steric hindrance. Different aliphatic amines were also examined in this catalytic system. The reactions of linear or branched alkyl amines, such as n-butyl amine (3k), n-hexyl amine (3l) and tert-butyl amine (3m), proceeded readily to afford the desired imines in moderate yields. Cyclopentyl amine (3n), benzylamine (3o) and o-phenylenediamine (3p) gave moderate yields at 120 °C. Next, a brief examination of various alcohols was conducted. Notably, aromatic alcohols with both electron-donating groups (p-CH3 and p-OCH3) and electron-withdrawing groups (p-Cl, o-Cl,p-Br and p-NO2) could be converted into the corresponding imines with excellent yields (>89% conversions and >90% selectivities) (3q-3v). Additionally, the transformations of 1-naphthalenemethanol (3w) and piperonyl alcohol (3x) produced the corresponding imines with moderate conversions and selectivities greater than 90%.
a Reaction conditions: 1 (0.5 mmol), 2 (0.6 mmol), CeO2 (100 mg), p-xylene (2 mL), 60 °C, for 12 h, O2. The conversions based on alcohol consumption were determined by GC and GC-MS. b 120 °C.
The relationship between the catalytic activity and the shape (rod, cube or octahedron) and crystal plane of the nanostructured CeO2 was investigated during the synthesis of an imine from benzyl alcohol and aniline. CeO2 samples with different morphologies were synthesized according to a procedure previously reported in the literature [42]. The X-ray diffraction (XRD) patterns of the resulting materials (Fig. 3) confirm the formation of ceria based on the diffraction peaks observed at 2θ = 28.7o, 33.2o, 47.7o and 56.6o (JCPDS 65-2975), corresponding to the (111), (200), (220) and (311) planes of a CeO2 fluorite-type structure, respectively, with no significant impurities observed. Transmission electron microscopy (TEM) and high resolution TEM (HRTEM) images (Fig. 4) indicate that these samples had different exposed crystal planes: (110) and (100) for the rods, (100) for the cubes, and (111) for the octahedra. Furthermore, the CeO2 rods consisted of 51% (110) planes and 49% (100) planes. These results are in good agreement with those of Yan’s group [42] and other researchers [5, 10].
Knowing that the synthesis of an imine from benzyl alcohol and aniline involves the oxidation of the alcohol to an aldehyde followed by the condensation of the aldehyde and aniline, we first investigated the alcohol oxidation activity over different CeO2 samples (Fig. 5(a)). Without aniline, the alcohol oxidation can still takes place over CeO2 nanorods, with 9% conversion of benzyl alcohol, while CeO2 nanocubes and nanooctahedra are nearly inert. The addition of aniline increases the conversion for all three CeO2 catalysts (Fig. 5(b)). In this case, the CeO2 nanorods show the highest activity, with the conversion increasing from 9% to > 99%, whereas the CeO2 nanocubes and nanooctahedra give 24% and 2% conversions, respectively. This is most likely because the removal of the benzaldehyde by reaction with aniline promotes the alcohol oxidation reaction. These results indicate the different redox properties of the different crystal planes.
Subsequently, the catalytic performance during the oxidative coupling of benzyl alcohol and aniline was assessed using shorter reaction time spans (Table 3). Under these conditions, CeO2 cubes and octahedra give low imine yields of 5.7% and 3.8%, respectively (Table 3, entries 2 and 3), although the CeO2 rods generate a remarkably high yield of 97% at 60 °C in 2 h (Table 3, entry 1). The imine generation rate over the (110) planes was calculated to be 4.618 mmol/(g·h), a value that is approximately 32 and 49 times faster than the rates obtained from the cubic (100) (0.143 mmol/(g·h)) and octahedral forms (111) (0.095 mmol/(g·h)), respectively.
The above results motivated us to conduct more in-depth investigations of the crystal plane structures. The redox sites of CeO2, which are correlated to the concentrations of oxygen vacancies, may play an important role in promoting the reaction. The oxygen vacancies were thus characterized by Raman spectroscopy (Fig. 6). Two bands are observed in the spectra of these samples; the intense band at 462 cm−1 is associated with the Raman-active F2g vibrational mode of the CeO2 fluorite-type structure [43], whereas the weak band at 595 cm−1 is attributed to Frenkel-type oxygen vacancies [44]. Because Raman spectroscopy using a 532 nm laser provides only surface information for CeO2 [45], the oxygen vacancy sites detected by Raman spectroscopy are primarily those on the CeO2 surface, and so the ratios of the integrated peak areas (A595/A462) were subsequently calculated to quantify the relative surface concentrations of oxygen vacancies. [46] The A595/A462 value for the nanorods is 0.077, while the nanocube and nanooctahedron samples generated ratios that were negligibly low. In the case of a perfect sample of CeO2 nanorods, the surface oxygen concentration should be 9.5×1019/g, a value that is three and seven times greater than those of the cube and octahedron forms, respectively [47], meaning that the probability of oxygen vacancies appearing on the surfaces of the CeO2 nanorods is higher than for the other two morphologies. Theoretical studies have shown that the oxygen vacancy formation energy of different CeO2 surfaces follows the order (110) < (100) < (111) [12, 48], hence oxygen vacancies are more readily formed on the CeO2(110) planes. Our previous work [2] has also proven that CeO2(110) possesses the highest concentration of oxygen vacancies, followed by CeO2(100) and CeO2(111), and this explains why the catalytic activity of CeO2(100) is superior to that of CeO2(100) and CeO2(111). This phenomenon demonstrates that the oxygen vacancies of CeO2 are capable of functioning as redox sites and hence act as active sites for this reaction.
In summary, we have successfully synthesized various imines by the aerobic oxidative coupling of alcohols and amines using pure CeO2 as a non-noble metal heterogeneous catalyst under oxygen. In the absence of a base or additives, a diverse range of substituted imines can be obtained with moderate to excellent yields. We have demonstrated that the CeO2(110) crystal planes are more active than the (100) and (111) planes during the synthesis of an imine from benzyl alcohol and aniline, because these planes possess the greatest concentration of oxygen vacancies. The data obtained in this work suggests that the redox properties of CeO2 are crucial for this reaction.
2015, Vol. 36 Issue (9): 1623-1630
PDF (1075 KB)
2015, Vol. 36 Issue (9): 1623-1630
PDF (1075 KB)