Reac Kinet Mech Cat (2014) 111:633–645 DOI 10.1007/s11144-013-0670-2

Selective hydrogenolysis of glycerol to 1,2-propanediol over MgO-nested Raney Cu Chuan-Jun Yue • Li-Ping Gu • Yang Su Shao-Ping Zhu

•

Received: 3 July 2013 / Accepted: 17 December 2013 / Published online: 7 January 2014 Ó Akade´miai Kiado´, Budapest, Hungary 2014

Abstract Cu/MgO and MgO nested on the surface of Raney Cu were prepared by a wet-mixing method and used as catalyst for glycerol hydrogenolysis. The texture of Raney Cu/MgO was characterized by X-ray diffraction and transmission electron microscopy, Brunauer–Emmett–Teller, and CO2-temperature-programmed desorption techniques. Results indicated that MgO was well dispersed in the pores of Raney Cu. Moreover, Raney Cu with large specific surface area had a larger contact surface for MgO, and a new phase of Cu–MgO was produced. Glycerol hydrogenolysis was selectively catalyzed under the synergistic effect of Cu and MgO in Cu/MgO catalyst. Under the moderate conditions of 180 °C and 1 MPa H2, the conversion of glycerol was 36.2 wt% and the selectivity for 1,2-propanediol was *85 wt% when chemical equilibrium. Keywords Raney Cu  Wet-mixing method  Cu/MgO  Selective hydrogenolysis of glycerol  Synergistic catalysis

Introduction With the growing global concern about environmental and sustainable development, methods for the utilization of biomass resources have been changing rapidly. Glycerol is the major by-product of biodiesel production, and the surplus glycerol has been predicted to exceed 1.54 million tons by 2015 [1]. Glycerol is a multifunctional building block [2]. Among the transformations of glycerol via oxidation, reduction, etherification, esterification, and dehydration, glycerol hydrogenolysis to 1,2-propanediol is a significant route [3] because C.-J. Yue (&)  L.-P. Gu  Y. Su  S.-P. Zhu Department of Chemical Engineering, School of Science, Changzhou Institute of Technology, Changzhou 213022, China e-mail: [email protected]

123

634

Reac Kinet Mech Cat (2014) 111:633–645

1,2-propanediol is a green chemical raw material widely used in numerous fields such as resins, cosmetics, medicines, and food. The industrial production of 1,2propanediol is achieved through epoxypropane hydration and catalytic oxidation of propylene based on petrochemical materials [4]. In recent decades, the production of 1,2-propanediol by catalytic hydrogenolysis of glycerol has become popular research subject for the utilization of renewable resources. The study on the core issues related to catalysts has greatly advanced, and heterogeneous catalysis has been applied with interest due to the easy separation of catalyst and products. The catalyst types that have been exploited include various catalyst systems based on noble metals, such as Ru [5], Pt [6], and Rh [7], or other metal systems based on Ni [8], Cu [9], Co [10], and their composites [11–14]. The Cu catalytic system [15–17] is highly valued not only because Cu has a low cost and is nontoxic, but also because Cu breaks the C–O bond without destroying the carbon skeleton of glycerol. The study of the catalytic process of a Cu-based catalyst has revealed that its catalytic hydrogenolysis has the characteristics of a bifunctional catalytic mechanism [18, 19], that is, the catalyst effectively matched Cu with acid or alkali components. Various bifunctional catalytic systems have been developed, including Cu/SiO2 [20], Cu/ZnO [21], and Cu/Al2O3 [22]. These catalysts were prepared by the co-precipitation or impregnation method, which render a few difficulties in controlling composition dispersion and particle size. Furthermore, to achieve desirable catalytic conversion, the catalytic hydrogenolysis of glycerol was usually performed under the high temperature and pressure above 200 °C and 5 MPa. Those methods often resulted in low catalytic conversion and selectivity to the desired product after the next catalytic cycle because the catalyst potential has not been fully developed, which attributed to the texture change of the used catalyst. Therefore, further improvements should be made in terms of catalyst preparation methods to obtain a stable catalyst for glycerol hydrogenolysis, especially for the Cu system. Moreover, the bifunctional catalysts containing Lewis bases such MgO contribute to the improvement in selectivity of glycerol hydogenolysis based on reported literature [11, 23]. In this study, porous MgO-modified Raney Cu through the wet-mixing method, possessing a stable skeleton, was used as the catalyst for glycerol hydrogenolysis. The distribution of MgO achieved by the regulation of Raney Cu and its texture was studied. The effect of reaction conditions on glycerol hydrogenolysis was investigated and the relationship between the catalyst structure and the selectivity of catalytic glycerol hydrogenolysis is discussed.

Experimental Chemicals Raney Cu (with copper content of 98.5 wt%, and average grain size of 300 mesh) was purchased from Dalian Toyounger Chemical Industry Co., Ltd. Glycerol (99.0 %), MgO (AR) ethyl valerate (GR), 1,2-propanediol (GR) and propanol (GR)

123

Reac Kinet Mech Cat (2014) 111:633–645

635

were procured from Aladdin, Shanghai, China. Hydrogen and nitrogen of high purity (99.999 %) were obtained from Sanjing special gas Co., Ltd, China. Catalyst preparation About 2.03 g Raney Cu and 1.27 g MgO (1:1 mol ratio, which value was calculated based on Cu amount) were weighed accurately, and then mixed and placed into a single-necked flask. The deionized water with twice the volume of the mixture was added, and the resulting mixture was stirred evenly for 24 h at room temperature. The mixture was filtered and dried in vacuum in a three-necked flask. Afterwards, the resulting mixture was calcined for 4 h at 400 °C under nitrogen protection, and then cooled to room temperature. The fine powder of the Raney Cu/MgO catalyst was finally obtained. Catalyst characterization Cu/MgO was characterized with Rigaku D/max 2500 PC X-ray diffractometer to analyze the phase composition. Analysis conditions were as follows: scanning was performed in the 2h range of 10°–80°. Cu K radiation at k = 0.1540 nm, pipe pressure = 40 kV, pipe flow = 100 mA. The sample was placed into an aluminum groove for testing after grinding. The particle morphology of Cu/MgO was analyzed by JEM-1230 transmission electron microscope manufactured with JEOL (Japan Electron Optics Laboratory Co., Ltd). Analysis conditions were as follows: the voltage was kept at 40–80 kV for 10 min; and then stopped for 5 min; after which the voltage was kept at 80–100 kV for 10 min, and then stopped for 5 min. The filament current was controlled to automatically reach saturation. The time of exposure was 2 s. The specific surface area and pore-size distribution of the samples were determined using an Autochem 2910 instrument manufactured by Micromeritics (USA). The analysis conditions were as follows: the catalyst sample was vacuum pumped under 623 K. The carbon dioxide adsorption was determined at cryogenic temperature. The cross-sectional area of N2 molecule was taken as 0.162 nm2. CO2-temperature-programmed desorption (CO2-TPD) characterization of the catalyst was performed on a home-made device. The CO2 signals were detected by a chromatographic thermal conductivity detector (TCD) (SP-3420 gas chromatograph manufactured by Beijing Analytical Instrument Factory). Approximately 300 mg of the sample (Cu/MgO or MgO, which were calcined for 4 h at 400 °C before test) was weighed and placed into a reaction tube. The sample was pretreated for 30 min in N2 flow with a rate of 25 mL/min, with increasing temperature from room temperature to 400 °C at a rate of 20 °C/min. The temperature was reduced to room temperature, and then the CO2 adsorption test was carried out at room temperature for 30 min. Afterwards, the temperature was increased to 100 °C at the rate of 10 °C/min. After reaching the stability, the temperature was continuously increased to 600 °C at a rate of 10 °C/min. The CO2TPD spectrogram was obtained by TCD detection.

123

636

Reac Kinet Mech Cat (2014) 111:633–645

Glycerol hydrogenolysis Glycerol hydrogenolysis reactions were performed in a stainless steel autoclave (75 mL) with the magnetic stirrer. In a typical experiment, 0.36 g of Raney Cu/ MgO and 30 mL of aqueous glycerol solution (50 wt% glycerol, 50 wt% H2O) were placed into the reactor, which was then flushed three times with nitrogen to remove the ambient air. Subsequently, the reactive system was pressurized in H2 (1 MPa as the initial pressure), and the mixture was heated up to 180 °C, stirring with a speed of 500 rpm. The reaction was stopped after 6 h. For the workout of the reaction, the gas was collected by a gas bag, and the reaction solution was separated by centrifugation and filtered. The filtrate was taken and analyzed. Liquid components were determined by gas chromatography using PEG-20 M capillary column and FID detector. The column temperature was programmed as follows: the initial temperature 40 °C was kept for 2 min, and raised at a rate of 15 °C/min to reach the final temperature of 200 °C. The temperature of the vaporization chamber was 250 °C. The temperature of the detector was 220 °C. The carrier gas pressure was 0.3 MPa. The air velocity was 60 mL/min. The hydrogen flow rate was 40 mL/min. The split ratio was 50:1. Ethyl valerate was used an internal standard for the calculation. The gas phase with trace and negligible products was determined by a chromatograph equipped with stainless steel chromatographic column (OV-3) and TCD detector. Conversion of the glycerol was calculated based on the following equation: Conversion (wt%Þ ¼ ½ðinitial mass of glycerolfinal mass of glycerolÞ= initial mass of glycerol  100 The selectivity of the products was calculated based on the formula: Selectivity (wt%Þ ¼ ½mass of specific products = ðinitial mass of glycerolfinal mass of glycerolÞ  100 The yield was calculated as the conversion rate of glycerol multiplied by the selectivity for this product: Yield (wt%Þ ¼ Conversion (%Þ  Selectivity (%Þ=100 Results and discussion MgO-modified Raney Cu for glycerol hydrogenolysis Generally, Raney Cu often exhibited higher activity than ordinary Cu powder in catalysis due to Raney Cu of a sponge-like texture, possessing high specific surface area and some catalytically active site, as a precursor for the potential application to multifunctional catalyst preparation. Therefore, the catalytic activity with specified selectivity could be obtained by modifying the surface of Raney Cu. The experiment showed that MgO-modified Raney Cu has excellent catalytic selectivity in the catalytic hydrogenolysis of glycerol.

123

Reac Kinet Mech Cat (2014) 111:633–645

637

The difference in the catalytic effects on glycerol hydrogenolysis is attributed to the varying compositions of Cu/MgO (Table 1). The result of the catalytic hydrogenolysis of glycerol showed that the catalyst of each composition had a certain catalytic activity under moderate conditions. With the increase in Cu content, the conversion of glycerol increases, indicating that Cu exhibits activity in glycerol hydrogenolysis. This catalytic performance is consistent with the high selectivity of Raney Cu in the gas-phase dehydration of glycerol for acetol [24]. On the other hand, the selectivity for 1,2-propanediol first increases, and then decreases. Selectivity is optimal when the mole ratio of Cu to Mg approaches 1, and the catalyst does not exist as precipitate but is dispersed like colloid in the system after the catalytic reaction. Such dispersion is due to the interaction between the hydroxyl group in glycerol and MgO with porosity and low density in Cu/MgO. This process greatly increases the contact of the catalyst with the reaction substrate. The catalysts exhibit optimal selectivity of 83.2 % for the catalytic hydrogenolysis of glycerol, indicating that MgO has a regulatory role in Raney Cu in the process of glycerol hydrogenolysis. Based on the results in Table 1, it suggests that the Cu/MgO favored the formation of acetol while MgO catalyzed mainly the undesirable reactions, resulting in the side products like glycol and polyglycerol. It could also be the reason why the near 1:1 ratio of Cu to MgO was appropriate, most of the side products were gone because there was minimal MgO and Raney Cu alone in the catalyst. Most of the components were in the form of Cu–MgO. On the other hand, the 3-hydroxypropanal intermediate in the glycerol hydrogenolysis catalyzed by Cu in Cu/MgO should be detected, due to the excellent dehydrogenation nature of Cu [24]. So it was speculated that the synergistic catalysis existed in Cu/MgO for glycerol hydrogenolysis. Catalyst properties Fig. 1 shows the X-ray diffraction (XRD) patterns of Cu/MgO (molar ratio 1), Raney Cu, and MgO. Fig. 1a and c shows the characteristic diffraction peaks of MgO (43.1°, 62.6°, and 79.0°) and Cu0 (43.2°, 50.5°, and 73.3°). Fig. 1b shows the XRD pattern of Cu/MgO, wherein new diffraction peaks appeared at double diffraction angles of 18.7°, 37.8°, and 58.1°, apart from the diffraction peaks of MgO and Cu0. This observation indicates that a new phase is produced between Cu and MgO in the process of catalyst preparation. The new phase is marked as Cu– MgO which is formed by the interaction between Raney Cu with MgO. This is because a new phase is easily generated on the interface of active Cu and MgO after the application of high temperature treatment to both mixtures [25], especially the defective sites and MgO (line, plane, and angle defects also exist on the MgO surface [26]). These defects provide the strong driving force for the binding [27]. The disappearance of the diffraction peak of the Cu (220) plane at 73.3° further demonstrates the interaction between Cu and MgO. In addition, the diffraction peak of Cu2O appears at the double diffraction angle of 36.0° in Fig. 1b, which is attributed to Cu oxidation in the process of sample preparation. The comparison of the peak intensity in Fig. 1b and c shows that MgO-modified Raney Cu has finer

123

638

Reac Kinet Mech Cat (2014) 111:633–645

Table 1 Effect of Cu/MgO ratio on the formation of 1,2-propanediol from glycerol Ratio of Mg/Cu (mol)

Conversion (wt%)

Selectivity (wt%)

Only MgO

12.4

5.5

4.0

21.5

63.5

6.5

1.96

30.6

80.3

0.97

48.7

83.2

0.68

49.7

0.49

52.4

1,2Propanediol

Yield (wt%) a

1,2Propanediol

Acetol

Propanol

Others

nd

0.1

94.4

0.7

0.5

29.5

13.7

12.8

1.3

5.6

24.6

12.4

0.3

0.1

40.5

81.3

12.9

4.5

1.3

40.4

79.1

10.7

5.2

5.0

41.5

Reaction conditions: 50 wt% glycerol, 30 mL; catalyst, 0.36 g; time, 6 h; temperature, 180 °C; initial PH2,1. Propanol is the mixture of 1-propanol and/or 2-propanol a

Other products include condensation products and glycol and so on

nd no detection

particles after calcination. The crystallinity declines and the mean particle diameter is *8.3 nm when calculated by half-peak breadth based on the Scherrer equation. Further illustrations on the structure of Cu/MgO are found in the transmission electron microsopy (TEM) image (Fig. 2). The TEM image indicates the framework structure of Raney Cu in the Cu/MgO catalyst. The metal particle Cu (deep black spots) is well dispersed in MgO (Fig. 2a, c). The irregular granular shape of Cu is observed at different positions (Fig. 2b), which is primarily determined by the original structure of Raney Cu from the preparation. The mean particle diameter of Cu is about 2.2 nm. Fig. 2d shows that a phase transition exists between Cu and MgO, which is produced by the interaction of Cu and MgO [28]. The phase change of each component in catalyst preparation is ruled out, and is attributed to the new phase Cu–MgO in the XRD. Apparent differences exist in the catalyst prepared by the wet-mixing method compared with the catalyst prepared by the co-precipitation method [29, 30]. The catalyst Cu/MgO prepared by the former method could regulate the interaction of Cu and MgO, as well as the distribution of MgO, by the microstructure and pore-size distribution of Raney Cu. Further information about the texture of Cu/MgO could be obtained by comparing the data in the Brunauer–Emmett–Teller (BET) analysis of Raney Cu and Cu/MgO (Table 2). The N2 adsorption–desorption profile indicates that Cu/ MgO is endowed with a type IV ring with mesoporous characteristics. The Raney Cu still retains its framework structure after thermal treatment, although the surface area and pore size are reduced. The surface area in the BET analysis was reduced by 21.5 %, from 30.40 to 23.87 m2/g. However, the surface area here is still a dozen units larger than that from Cu/MgO prepared by the co-precipitation method. Meanwhile, the mean pore volume and pore size decreased, indicating that the pore surface of the original Raney Cu is occupied by porous MgO with low density. The mean pore size of the remaining pores is reduced to 25 nm, which is half the original value. These results correspond with the XRD and TEM analysis. The CO2TPD profiles of Raney Cu/MgO and MgO (Fig. 3) shows a significant difference. It

123

Reac Kinet Mech Cat (2014) 111:633–645

639 #

Fig. 1 XRD patterns of MgO (a), Cu/MgO (b), and Raney Cu (c)

# MgO

* Cu

#

^ Cu-MgO #

a) b)

^

^

^

*

*

c)

10

20

30

40

50

60

70

80

90

2Theta/(degree)

is presented that the successive desorption peaks of CO2 around that of 263.7 °C (corresponding to the strong basicity of MgO2- pair and O2- ion) for MgO, which derived from the unevenly dispersed basic sites from the out surface to the inside. While there are three desorption peaks of CO2, including two main peaks and one wide-shoulder peak for Cu/MgO, which illuminated that the total basic strength of 1.35 mmol/g was higher than 0.82 mmol/g in MgO. The desorption peak at 126.1 °C was attributed to the weak physical adsorption with OH, whereas the desorption peak at 221.9 °C was attributed to the strong adsorption with CO2 and MgO (MgO2- pair). Furthermore, the strength changed with the position of the MgO surface. Stronger adsorption appeared at 294 °C with MgO (O2- ion), indicating different basic sites in Raney Cu/MgO, which demonstrated that the copper in Cu/MgO prompted MgO dispersion and improved the basic sites of MgO, and which was substantially consistent with Cu/MgO catalyst prepared the coprecipitation way in terms of basicity [31]. Selective catalytic hydrogenolysis of glycerol For the catalytic hydrogenolysis of glycerol, a high-selectivity catalyst under mild conditions is an ideal target. The temperature and hydrogen pressure are the influencing factors that should be considered. Similarly to other catalytic systems, Raney Cu/MgO catalyst is also sensitive to temperature during the catalytic hydrogenolysis of glycerol (Table 3). The conversion of glycerol increases with the increase in temperature, which corresponds with the general relationship between temperature and reaction. However, the selectivity for 1,2-propanediol declines with an increase in temperature because the active catalyst further catalyzes the conversion of 1,2propanediol in the liquid phase reaction system. With the increase in temperature, the active catalyst not only breaks the C–O bond but also the C–C bond, causing further conversion of 1,2-propanediol to propanol and glycol. This result is

123

640

Reac Kinet Mech Cat (2014) 111:633–645

Fig. 2 TEM images of Cu/MgO

consistent with that in other Cu catalytic systems for glycerol hydrogenolysis [32]. However, the temperature is lower than that reported in the literature reported for glycerol hydrogenolysis with the same degree of conversion because Raney Cu with a relatively large specific surface area increases the contact with glycerol in the catalytic process [24]. The catalysts prepared by this method have excellent activity. From the perspective of total efficiency of the catalytic hydrogenolysis of 1,2propanediol, the yield of glycerol over the catalyst is 40.6 % at about 180 °C. Hydrogen pressure is another important influencing factor in the catalytic hydrogenolysis of glycerol (Table 4). The conversion of glycerol increases with the increase in hydrogen pressure because the increase in hydrogen pressure facilitates the dissolution of hydrogen in the solution, and consequently the H2 adsorption on the catalyst surface. However, when hydrogen pressure increases to 1 MPa or above, the products generated would react further due to the high activity of the

123

Reac Kinet Mech Cat (2014) 111:633–645

641

Table 2 Texture properties of Cu/MgO compared with Raney Cu Sample

Surface area BET (m2/g)

Average pore volume (cm3/g)

Average pore size (nm)

Total base sites (mmol/g)

Raney Cu

30.40

0.24

57

–

Cu/MgO

23.87

0.16

25

1.35a

a

Total base value was investigated based on CO2-TPD; the others were derived from BET determination

Fig. 3 CO2-TPD profiles of Cu/ MgO and MgO

o

o

126.1 C

221.9 C o

Intensity

294.5 C o

263.7 C

Cu/MgO MgO

100

200

300

400

o

T/ C

catalysts in the catalytic reaction system. As a result, the selectivity for 1,2propanediol would be reduced, but the amplitude of such reduction is lower than the degree of the catalytic hydrogenolysis. Thus, the results of the regular dispersion of products from a sequence of reaction steps in the catalytic hydrogenolysis of glycerol are verified in the literature [33]. Moreover, with no hydrogen in the reaction, catalyst deactivation easily occurs. Therefore, hydrogen is important in maintaining catalyst activation, probably by reducing the oxidized copper of uncompleted catalysis cycle or diminishing the adsorbed species (to prevent the cycle of catalytic reaction) in the catalyst surface by hydrogenolysis [34]. The effect of reaction time on the conversion of glycerol and selectivity for 1,2propanediol characterizes the catalytic reaction process. As shown in Table 5, with a prolonged reaction time, the conversion of glycerol is increased all the time. In the first 8 h, the increase in amplitude is apparently higher compared with the later reaction. After 24 h, the conversion reaches 75 %, thereby maintaining the catalytic equilibrium results. However, the selectivity for 1,2-propanediol hovers at about 85 %, and showing a slow increasing trend at 12 h. At this time, 1,2-propanediol is the main selective product of the reaction. In addition, the comparison of the curves for 1,2-propanediol and acetol showed that they change in opposite directions because acetol is the first to be produced, and it is immediately converted to 1,2propanediol by catalytic hydrogenation. Cu-based catalyst catalyzes the glycerol dehydration to acetol under a hydrogen-lacking condition [24, 35].

123

642

Reac Kinet Mech Cat (2014) 111:633–645

Table 3 Effect of reaction temperature on 1,2-propanediol production Temperature (°C)

Conversion (wt%)

Selectivity (wt%)

Yield (wt%)

140

32.6

97.3

31.7

160

41.4

89.2

36.9

180

48.8

83.2

40.6

200

55.2

79.7

44.0

220

61.3

73.1

44.8

Reaction conditions: 50 wt% glycerol, 30 mL; catalyst, 0.36 g; time, 6 h; initial PH2, 1 MPa

Table 4 Effect of hydrogen pressure on 1,2-propanediol production Hydrogen pressure (MPa)

Conversion (wt%)

Selectivity (wt%)

Yield (wt%)

0.5

35.2

80.3

28.3

0.8

42.5

82.2

34.9

1.2

50.4

83.1

41.9

1.5

55.3

82.4

44.6

2.0

57.1

79.2

45.2

Reaction conditions: 50 wt% glycerol, 30 mL; catalyst, 0.36 g; time, 6 h; temperature, 180 °C

Table 5 Effect of reaction time on glycerol hydrogenolysis Reaction time (h)

Conversion (wt%)

Selectivity (wt%)

Yield (wt%)

1,2-Propanediol

Acetol

Propanol

Othersa

1,2-Propanediol

4

26.3

76.2

16.5

1.1

6.2

0.7

5

30.6

80.3

12.8

1.3

5.6

13.7

12

32.1

82.2

12.1

1.6

5.1

24.6

18

35.4

83.5

10.4

2.2

3.9

29.6

24

36.2

84.4

8.1

4.8

2.7

30.6

Reaction conditions: 50 wt% glycerol, 30 mL; catalyst, 0.36 g; temperature, 180 °C; initial PH2,1 MPa. Propanol is the mixture of 1-propanol and/or 2-propanol a

Other products include condensation products and glycol and so on

In addition, the comparative experiments of glycerol hydrogenolysis were carried out using the Cu/MgO catalysts prepared by present method, impregnation [36] and co-precipitation [30] keeping the same 1:1 mol ratio of Cu to MgO under the reaction conditions of 0.36 g catalyst, 30 mL of 50 wt% glycerol solution, 1 MPa of initial hydrogen pressure, 180 °C, and 6 h, the conversions were 48.8, 37.2 and 50.6 wt%, and the selectivities to 1,2-propanediol were 83.2, 64.2 and 77.4 wt%. Compared to the catalytic results, it suggests that the selectivity based on Cu/MgO by wet-mixing takes the advantage over those by the other two methods, while the catalytic conversion is lower than that by co-precipitation but higher than that by

123

Reac Kinet Mech Cat (2014) 111:633–645

643

impregnation. This indicated that the Cu/MgO catalysis to glycerol hydrogenolysis not only depended on the basicity and catalyst surface, also the interaction of Cu and MgO, which was responsible for the improved selectivity to 1,2-propanediol. And this point supported that there was the synergistic effect between Cu and MgO during the catalytic hydrogenolysis of glycerol [34]. Structure and reactivity of catalyst According to the analysis on product distribution in the catalytic hydrogenolysis of glycerol over Cu/MgO and the reported literature on Cu-based catalytic process of glycerol hydrogenolysis [18, 19], the proposed process of catalytic hydrogenolysis over Cu/MgO is shown in Scheme 1. The research indicated that the glycerol conversion depended on the basicity and Cu dispersion in Cu–MgO prepared by the co-precipitation method [30], whereas Raney Cu catalyst was beneficial to transform glycerol into acetol via free-radical process by the elimination of primary hydroxyl group in glycerol in a fixed-bed downflow glass reactor [24]. So MgO cooperated with Raney Cu catalyzed glycerol to improve the selectivity to 1,2-propanediol. Meanwhile, on account of the acidity and basicity of MgO [37], excellent catalytic dehydrogenation property of Cu/MgO [38], and the production of metastable acetol in thermodynamics under mild conditions [39], glycerol dehydration was achieved by the combination of hydrogen spillover of Cu and acid-based sites of MgO in Cu/MgO. The tight connection of the new phase of Cu–MgO formed in the catalyst preparation has a role in selective glycerol dehydration. According to the analysis of the transition state on the Cu surface and glycerol hydrogenolysis under conditions with hydrogen [24], the removal of the secondary hydroxyl group is induced by Mg2? on the Cu–MgO surface in free radical form under synergistic catalysis. If the catalysis is in the form of non-free radical, dehydrogenation first occurs to produce aldehydes, followed by dehydration and hydrogenation to produce acetol. Consequently, acetalization products including cyclic compounds are obtained by the condensation reaction between the aldehyde and hydroxyl groups in the reactants. However, acetalization products are not detected in the system after reaction. Enol is produced by dehydration as intermediate based on the GC qualitative analysis of the products; moreover, the result has been confirmed by other copper-based catalysts during catalytic hydrogenolysis of glycerol [40]. Acetol is obtained by tautomerization of unstable enol, whereas 1,2-propanediol is obtained by catalytic hydrogenation of acetol on the Cu surface. Similarly, H , OH

HO

OH OH

- H2O

OH + H 2

OH

Cu-MgO

OH

O

OH Cu

OH

MgO MeOH, EtOH, glycol, alkane etc.

OH

Scheme 1 Proposed process of the selective hydrogenolysis of glycerol to 1,2-propanediol on Cu/MgO

123

644

Reac Kinet Mech Cat (2014) 111:633–645

2-propanol is obtained by further catalytic dehydration and hydrogenation of 1,2propanediol. However, in-depth investigation is required to confirm this hypothesis. Meanwhile, the products 1,3-propanediol, ethylene glycol, ethanol and methanol are generated along with other catalytic reactions.

Conclusion Cu/MgO, prepared by nesting MgO on Raney Cu with a larger specific surface area, is a catalyst with excellent activity and selectivity in the catalytic hydrogenolysis of glycerol. The characteristic results showed that basic MgO was well dispersed in the surface of Raney Cu. Compared with other preparation methods, Cu/MgO prepared by the current method showed synergetic effect in the catalytic process in addition to the catalytic performance of a bifunctional catalyst. The Cu/MgO exhibited an approximate 85 wt% selectivity for 1,2-propanediol with a conversion of 36.2 wt% under 180 °C and 1 MPa hydrogen pressure, thereby showing high catalytic selectivity under moderate conditions. The selective catalytic performance of Cu/ MgO for glycerol hydrogenolysis contributes to the advantageous paring of Cu with MgO, especially with the large contact surface of Cu and MgO. The improvement of the catalytic performances of Cu/MgO was closed to the skeletal structure of Raney Cu. Further analysis on the catalyst structure and in situ catalytic reaction mechanism can contribute to the design of improved catalysts. Therefore, improved catalytic systems are environment-friendly and cost-effective must be developed. Acknowledgments We thank Nature Science Foundation of Jiangsu Province (Grant No. BK2011235), Guiding Project of Education Department of Jiangsu Province (No. 08kjd30001) and Nature Science Foundation of Changzhou Institute of Technology (No. YN0907) for financial support.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Ganapati DY, Payal AC, Devendra PT (2012) Ind Eng Chem Res 51:1549 Ritter SK (2004) Chem Eng News 82:31 Zhou CH, Beitrmini JN, Fana YX, Lu GQ (2008) Chem Soc Rev 37:527 Balaraju M, Rekha V, Prasad PSS, Devi BLAP, Prasad RBN, Lingaiah N (2009) Appl Catal A 354:82 Lahr DG, Shanks BH (2005) J Catal 232:386 Maris EP, Davis RJ (2007) J Catal 249:328 Shinmi Y, Koso S, Kubota T, Nakagawa Y, Tomishige K (2010) Appl Catal B 94:318 Perosa A, Tundo P (2005) Ind Eng Chem Res 44:8535 Wu Z, Mao Y, Song M, Yin X, Zhang M (2013) Catal Commun 32:52 Anand KA, Anisia KS, Agarwal AK, Kumar A (2010) Can J Chem Eng 88:208 Xia SX, Nie RF, Lu XY, Wang L, Chen P, Hou ZY (2012) J Catal 296:1 Kwak BK, Park DS, Yun YS, Yi J (2012) Catal Commun 24:90 Rodrigues R, Isoda N, Goncalves M, Figueiredo FCA, Mandelli D, Carvalho WA (2012) Chem Eng J 198–199:457 Gandarias I, Requies J, Arias PL, Armbruster U, Martin A (2012) J Catal 290:79 Sa´nchez T, Salagre P, Cesteros Y, Bueno-Lo´pezb A (2012) Chem Eng J 179:302 Dura´n-Martı´n D, Ojeda M, Granados ML, Fierro JLG, Mariscal R (2013) Catal Today 210:98 Vasiliadoub ES, Eggenhuisenc TM, Munnikc P, de Jonghc PE, de Jongc KP, Lemonidou AA (2014) Appl Catal B 145:108 Chaminand J, Djakoviteh L, Gallezot P, Marion P, Pinel C, Rosier C (2004) Green Chem 6:359

123

Reac Kinet Mech Cat (2014) 111:633–645 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

645

Miyazawa T, Koso S, Kunimori K, Tomishige K (2007) Appl Catal A 318:244 Huang Z, Cui F, Xu WJ, Zuo J, Chen J, Xia C (2012) Catal Today 183:42 Wang S, Zhang Y, Liu H (2010) Chem Asian J 5:1100 Akiyama M, Sato S, Takahashi R, Inui K, Yokota M (2009) Appl Catal A 371:60 Dam J, Hanefeld U (2011) ChemSusChem 4:1017 Sato S, Akiyama M, Takahashi R, Hara T, Inui K, Yokota M (2008) Appl Catal A 347:186 Niiria L, Francesc I (1997) J Mol Catal A 119:177 Che MJ, Tench AJ (1982) Adv Catal 31:77 Campbell CT, Starr DE (2002) J Am Chem Soc 124:9213 Lo´pez N, Illas F (1996) J Phys Chem 100:16275 Nagaraja BM, Kumar VS, Shashikala V, Padmasri AH, Reddy SS, Raju BD, Rao KSR (2004) J Mol Catal A 223:339 Balaraju M, Jagadeeswaraiah K, Sai Prasad PS, Lingaiah N (2012) Catal. Sci. Technol. 2:1967 Di Cosimo JI, Dı´ez VK, Xu M, Iglesia E, Apesteguı´a CR (1998) J Catal 178:499 Dasari MA, Kiatsimkul PP, Sutterlin WR, Suppes GJ (2005) Appl Catal A 281:225 Hamzah N, Nordin NM, Nadzri AHA, Nik YA, Kassim MB, Yarmo MA (2012) Appl Catal A 419–420:133 Vasiliadou ES, Lemonidou AA (2011) Appl Catal A 396:177 Niu SS, Wang L, Zheng HY (2011) Chin J Catal 32:345 Yuan ZL, Wang JH, Wang LN, Xie WH, Chen P, Hou ZY, Zheng XM (2010) Bioresour Technol 101:7088 Rekoske JE, Barteau MA (1997) J Catal 165:57 Shi R, Wang F, Mu X, Li Y, Huang X, Shen W (2009) Catal Commun 11:306 Huang L, Zhu YL, Zheng HY, Ding GQ, Li YW (2009) Catal Lett 131:312 Akiyama M, Sato S, Takahashi R, Inui K, Yokota M (2009) Appl Catal A 371:60

123