Catalytic hydrogenolysis of diphenyl ether over Ru supported on amorphous silicon-aluminum-TiO2

CHEN Bo ，LI Lei ，DIAO Zhi-ju ，CAO Rui-dong ，SONG Li-fei ，HUANG Liang-qiu ，WANG Xue

（1. School of Chemical Engineering, Northwest University, Xi'an 710069, China；2. State Key Laboratory of High-efficiency Utilization of Coal and Green Chemical Engineering, Ningxia University,Yinchuan 750021, China；3. Xi 'an Giant Biological Gene Technology Co. LTD, Xi'an 710077, China；4. College of Urban and Environment Science, Northwest University, Xi'an 710127, China）

Abstract: A bifunctional catalyst of Ru5/ASA-TiO2 was prepared by using a novel silicon-aluminum (ASA)-TiO2 amorphous composite, which was synthesized by a steam-assisted method, as the support. X-ray diffraction (XRD), pyridine adsorption infrared (Py-FTIR), ammonia-temperature-programmed desorption (NH3-TPD), scanning electron microscopy (SEM),transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS) and other methods were used to characterize the structure and the acidity of the prepared catalyst. Using diphenyl ether as the lignite-related model compound,the reaction activity of the Ru5/ASA-TiO2 for the catalytic hydrogenolysis of 4-O-5 type ether bonds was investigated under a mild condition. The results show that the weak acid and/or the Lewis acid rather than the strong Brønsted acid mainly contribute to improve the conversion rate and the benzene yield of the catalytic hydrogenolysis of diphenyl ether. The reaction temperature can influence the relative content of various types of acids to significantly affect the selectivity of the hydrogenolysis products of diphenyl ether. The conversion rate of diphenyl ether is greater than 98% while the benzene yield is 67.1%.

Key words: hydrogenolysis；catalyst；lignite；model compound；ether linkage

The lignite reserve is estimated to be more than one trillion tons all over the world[1]. However, the low calorific value and the high content of ash and water of lignite limit its industrial use[2]. Therefore, it is necessary to develop efficient conversion processes to minimize these disadvantages.

The cleave of oxygen-bridge bonds, which is abundant in the organic matter of lignite, is a critical step for converting lignite to clean fuels and valueadded chemicals[3-5]. Hydrogenolysis is one of the methods that can effectively cleave the oxygencontaining bridge bonds in the organic macromolecules of coal.

The oxygen-containing bridge bond in the organic macromolecules of coal is mainly connected in four ways:α-O-4,β-O-4,α-O-γand 4-O-5, among which the 4-O-5 type of ether bond has relatively weak reaction activity. Even for a reaction in 15% formic acid at 315 °C or 10% phosphoric acid at 250 °C for 3 d, it cannot be depolymerized[6]. Therefore, as the simplest compound containing 4-O-5 type ether bonds,diphenyl ether is widely used as a model compound of coal.

Metal Ru is the most promising active metal that can be used in hydrogenolysis reactions. It can not only effectively activate H2but also selectively depolymerize the C-C and C-O bonds[7,8]. TiO2with strong Lewis acidity[9,10]and high hydrodeoxygenation reactivity[11,12]is one of the most widely used catalyst supports. In addition, the surface of TiO2also has the spillover hydrogen effect[13]. According to the reverse Mars-van Krevelen mechanism[14], spillover hydrogen can create oxygen vacancies on the surface of TiO2as active sites for the hydrodeoxygenation reaction.

At the same time, it is well known that the acidity of the support is another important factor affecting the selectivity of the hydrogenolysis reaction. Some researchers[15]found that the strong Brønsted acidity of the molecular sieve can easily cause excessive cracking of the raw materials, resulting in a decrease in the product yield. However, silicon-aluminum (ASA) can effectively inhibit the secondary reaction of products due to its controllability of Brønsted and Lewis acids.

Herein, with the aim to produce arenes from lignite, we synthesized a highly efficient Ru-based catalyst supported by ASA-TiO2for the selective hydrogenolysis of aryl ether bonds under hydrogen atmosphere in aqueous media. The catalyst was characterized and used for the hydrogenolysis of diphenyl ether, which is the simplest compound containing 4-O-5 type ether bond with weak reaction activity and is widely selected as a model compound f or lignite.

1 Experimental

1.1 Materials

Solvents and reagents were purchased from Macklin and were used as received without any further purification. Commercial HZSM-5 molecular sieve,denoted as HZSM-5c, was purchased from Nankai University Catalyst Co., Ltd.

1.2 Synthesis of ASA-TiO2

About 0.10 g aluminum isopropoxide and 2.45 g tetrapropylammonium hydroxide were mixed under magnetic stirring for 1 h at room temperature, after which about 2.04 g tetraethyl orthosilicate was dropwise added to the clear solution to form a mixture, which was further stirred for 24 h at 40 °C. Then, a certain amount of deionized water and anatase TiO2were added to the mixture to form the final sample, which was kept at 110 °C for 6 h before being transferred into a Teflon-lined stainless steel autoclave to be crystallized by a steam-assisted method at 180 °C for 24 h. Finally, the generated solid precipitate was calcined at 550 °C to obtain the composite of ASATiO2, which was used as the support for the catalyst.For comparison, HZSM-5 was prepared through the same procedure but without TiO2.

1.3 Catalyst preparation and characterization

Catalysts were prepared by impregnating the support with 5% of Ru using an aqueous solution of RuCl3·xH2O as the precursor. After impregnation, the catalyst precursor was dried at 110 °C overnight,followed by being reduced at 300 °C for 2 h with a ramp of 5 °C/min under a H2atmosphere (20 mL/min)before finally being passivated under a flow of 2%O2/N2for 0.5 h at room temperature. The black sample using ASA-TiO2as the support is denoted as Ru5/ASATiO2. X-ray diffraction (XRD), pyridine adsorption infrared (Py-FTIR), ammonia-temperature-programmed desorption (NH3-TPD), scanning electron microscopy(SEM), transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS) are used to systematically investigate the effect of the structural characteristics and the changes of the acidity on the performance of the catalyst in the catalytic hydrogenolysis of diphenyl ether.

1.4 Catalytic test

For a typical test, the reactant (0.2 mmol), H2O(5 mL) and the fresh catalyst (0.02 g) were added into a 24 mL stainless steel autoclave reactor, which was sequentially charged with 0.2 MPa H2and 0.6 MPa N2after being purged with hydrogen several times to remove the air. The test was performed at the desired temperature for 1 h with a stirring speed of 1000 r/min.After the test, the reactor was quenched to ambient temperature using cooling water. Then the obtained sample was extracted using ethyl acetate (10 mL) with the addition of 20 μLn-dodecane as the internal standard. The organic liquid products were analyzed with a gas chromatograph (GC) equipped with a flame ionization detector (FID) and a gas chromatographmass spectrometer (GC-MS). The conversion of diphenyl ether and the yield of the liquid product were calculated according to the following equations on the basis of mole balance.

2 Results and discussion

2.1 Characterization of catalysts and supports

As shown in Figure 1, the ASA-TiO2composite as the support presents obvious diffraction peaks of TiO2.In this case, both silicon and aluminum may be distributed on the surface of TiO2in an amorphous form[16]. The existence of anatase-type TiO2does not conducive to the formation of the HZSM-5 molecular sieve while the mechanism needs to be further studied[17].

A series of characterization results of the catalyst of Ru5/ASA-TiO2is shown in Figure 2. The porous surface (Figure 2(a)) of the catalyst provides a larger specific surface area facilitating the dispersion of Ru species. As shown in Figure 2(b), Ru species with an average particle size of (2.1 ± 0.8) nm are uniformly distributed on the surface of the support. In the highresolution TEM image (Figure 2(c)), the interplanar spacing corresponding to Ru (111) and Ru (110) can be observed, indicating that there are a large number of metallic Ru nanoparticles on the Ru5/ASA-TiO2. The lower-left corner of Figure 2(d) presents the interplanar spacing corresponding to TiO2(101). The fast Fourier transform mode of the selected area in the upper-right corner shows obvious amorphous diffraction patterns,in agreement with the XRD results that siliconaluminum oxide is dispersed on the surface of TiO2in an amorphous form.

Figure 2(e) presents the 3p3/2orbital information of the Ru on the surface of the catalyst of Ru5/ASATiO2, which was corrected with the binding energy of C 1sof 284.80 eV to eliminate the effect of charging as suggested by Neimark et al.[18]. The peak at 460.9 eV is attributed to Ru0while another peak at 462.5 is related to RuO2, as shown in Figures 2(e) and 2(f). The oxidized form of Ru may be produced from the partial oxidation of Ru in the air, as indicated by Liu et al.[8].

Figure 1 XRD patterns of different supports

Figure 2 Characterizations of the catalyst of Ru5/ASA-TiO2 with (a) SEM, (b)-(d), (f) TEM and (e) XPS Inset in (d) is the fast Fourier transformation of the selected area

As shown in Figure 3, the ASA-TiO2composite as the support has a Brønsted acid site (BA) and two Lewis acid sites (LA1and LA2) with different strengths and properties. The three absorption peaks are located at 1445, 1454 and 1545 cm-1, respectively, consistent with the work of Shamzhy et al.[19]. Table 1 presents the content of the Lewis and Brønsted sites, which can be calculated according to the areas of the absorption peaks and their corresponding extinction coefficients.The value of the ratio of both BA/LA and LA1/LA2as shown in Table 1 indicates that the Lewis acidity of the support of the ASA-TiO2composite is mainly derived from the LA1acid sites of TiO2. On the other hand,when the desorption temperature of pyridine is increased from 150 to 250 °C, the value of BA/LA and LA1/LA2of ASA-TiO2changes from 0.30 and 1.46 to 0.72 and 0.91, respectively, indicating that an increase in the temperature can significantly reduce the acidity of LA1.

Table 1 Number of BA and LA sites in different supports a

Figure 3 Py-FTIR spectra of the support after desorption at different temperatures

2.2 Catalytic hydrogenolysis performance

As shown in Table 1, the total acid content of the commercial HZSM-5creaches to 1.58 mmol/g.However, the conversion rate of the catalytic hydrogenolysis of diphenyl ether while the benzene yield of Ru5/HZSM-5care the lowest, as shown in Figure 4. Although the total acid content of other supports is significantly lower than that of the commercial HZSM-5c, their content of the Lewis acid,their conversion rate of diphenyl ether and their yield of benzene are all significantly higher than those of the commercial HZSM-5c. This indicates that the Lewis acid of the support can significantly affect the activity of hydrogenolysis reactions, which is consistent with other studies[20,21].

Notably, the relative content of LA1of Ru5/ASATiO2is nearly 10% higher than that of ASA-TiO2(see Table 1) because Ru can promote the formation of more defects (oxygen vacancy and Ti3+as shown in Figure 2(e)) on the surface of anatase TiO2during the reduction process, as indicated by Deng et al.[22]. In addition, Boonyasuwat et al.[12]revealed that these defects are closely related to the acid site of LA1. As mentioned above, although an increase in the temperature can significantly reduce the acidity of LA1in the ASA-TiO2composite as the support under nonreduction conditions, Ru can promote the continuous generation of surface defects on TiO2under the reaction condition in the presence of hydrogen while the concentration of the defects on the surface of TiO2increases with the increase in the reduction temperature,as suggested by Hery et al.[23]. This is significantly positively correlated with the phenomenon that the yield of benzene increases with the increase in the temperature while the content of cyclohexanol and cyclohexanone decreases gradually, as shown in Figure 5. On the other hand, as shown in Figure 6, for a reaction time of 60 min, the maximum benzene yield is 67.1% while the yield of phenol decreases from 10.7%to 0.5%. Further extension of the reaction time can lead to the hydrogenation of a small amount of benzene to cyclohexane.

Figure 4 Hydrogenolysis of diphenyl ether with different catalysts

As shown in Figure 7, the mechanism of the catalytic hydrogenolysis of diphenyl ether over Ru5/ASA-TiO2is proposed as following: First, the aromatic ether bonds of diphenyl ethers are depolymerized directly to generate benzene and phenol,rather than through the depolymerization path via a partially hydrogenated product of (cyclohexyloxy)benzene which cannot be detected during all the reactions. Subsequently, the reaction path of phenol mainly depends on the reaction temperature. A temperature lower than 190 °C is conducive to the hydrogenation reaction while a temperature higher than 190 °C benefits the deoxidation and dehydrogenation to generate benzene as shown in Table 2 and Figures 5-6, which is consistent with the results previously reported[24]. The above reaction process involves two types of Lewis acid centers and Brønsted acid centers.In addition, Nelson et al.[25]revealed that the Lewis acid centers provided by TiO2can be converted to the Brønsted acid centers. These acid centers dynamically change with the reaction conditions and present different reactivities. These characteristics of the catalyst is challenging to understand the hydrogenolysis mechanism of diphenyl ether.

Table 2 Validation tests for the production of benzene over Ru5/ASA-TiO2

Figure 5 Effect of temperature on the distribution of primary products

Figure 6 Effect of reaction time on the distribution of primary products

Figure 7 Possible pathways for the hydrogenolysis of diphenyl ether over Ru5/ASA-TiO2

3 Conclusions

A bifunctional catalyst Ru5/ASA-TiO2was prepared and used in the hydrogenolysis of diphenyl ether, a lignite-related model compound. The prepared catalyst of Ru5/ASA-TiO2presents relatively high reactivity of the depolymerization of diphenyl ether depolymerization and relatively high selectivity of benzene. At 250 °C with a hydrogen pressure of 0.2 MPa, the conversion rate of diphenyl ether is higher than 98% while the yield of benzene is 67.1%. The results reveal that the weak acid and/or the Lewis acid,rather than the stronger Brønsted acid, can improve the conversion rate and the yield of benzene for the hydrogenolysis of diphenyl ether. More importantly,the reaction temperature significantly affects the relative content of various types of acids, thus affecting the selectivity of the product of the hydrogenolysis of diphenyl ether. A lower temperature (＜ 190 °C) is conducive to the hydrogenation reaction while a higher temperature (＞190 °C) promotes both the deoxygenation and dehydrogenation reaction, thus improving the yield of benzene.