Insight into the catalytic mechanism of glycerol hydrogenolysis using basal spacing of hydrotalcite as a tool

Abstract

Information on the catalytic mechanism is crucial for understanding heterogeneous catalysis of glycerol hydrogenolysis. In this work, we proposed a simple and effective method to study the catalytic mechanism of glycerol hydrogenolysis with Cu/Mg–Al mixed-oxide as catalyst. This method is based on the memory effect of the hydrotalcite-derived Mg–Al mixed-oxide and the tunable property of hydrotalcite interlayer distance. The experimental results showed that the basal spacing of hydrotalcite could work as an effective tool to reflect the relationship between desorption of products and reaction conditions. With this tool, the effect of H₂ pressure on glycerol hydrogenolysis was elucidated. It was found that H₂ pressure affected desorption and yield of products.

---

1. Introduction

Increasing energy costs and environmental concerns have emphasized the need to produce renewable fuels and chemicals. It is generally accepted that the primary source of renewable organic fuels, chemicals, and materials will be from plant derived biomass.¹ To date, major advances have been made in the transesterification of vegetable oil-based fatty acids to biodiesel, which is already at an advanced stage of commercialization. The utilization of the relatively large amount of glycerol side products (10 wt%) formed in biodiesel manufacturing is a major factor in the overall cost effectiveness of the process. Therefore, use of this glycerol as a renewable feedstock to produce value added chemicals is crucial to biorefinery economics.^2–6 Among the different possible transformations of glycerol, the selective conversion to 1,2-propanediol, which is usually produced from petroleum derivatives, presents a special interest (Scheme 1).^7–18

Scheme 1 Hydrogenolysis of glycerol.

Hydrogenolysis of glycerol to 1,2-propanediol has been reported previously using several supported transition metal catalysts such as Ru,^7–9 Pt,^8–12 Cu,^13–15 Ni,¹⁶ and some bimetallic catalysts consisting of Pt–Ru, Au–Ru¹⁷ and Ru–Re.¹⁸ In this heterogeneous catalysis as to glycerol hydrogenolysis, the usual reported data are conversions obtained at different process conditions and with different catalysts. Little information is available on the catalytic mechanism of glycerol hydrogenolysis.^19,20 However, information on the catalytic mechanism is crucial for understanding a reaction system of heterogeneous catalysis.

In recent years, hydrotalcite-like compounds (hereafter indicated as HT) have many applications, including as catalyst supports for glycerol hydrogenolysis.^{14,15,21–28} HT are layered double hydroxides belonging to the class of anionic clays found in Nature, in which cationic layers and anionic interlayers show a basal spacing of 7.8 Å (Scheme 2). The interactions between the metal hydroxide layers are much weaker than the intralayer bonding, leading to swelling of the unit cell c-parameter (perpendicular to the layers, basal spacing) with adsorption of energetically favorable species. HT interlayers thus serve as excellent hosts for a broad range of organic and inorganic charge-balancing anions and a variety of neutral chemical species, forming the structure of guest-intercalated HT (hereafter indicated as HT-guest). The basal spacing of HT-guest changes with the size of guest species. The formation of HT can be identified easily by X-ray diffraction studies. The (003) reflection can be used to calculate the basal spacing, d, between the layers.^24–26 Thermal decomposition of HT leads to the formation of mixed oxides. These mixed oxides show a memory effect, a property by which they can recover the original lamellar structure if they come into contact with water.^27,28

Scheme 2 Structure of hydrotalcite.

Water, as solvent and by-product, is one component of the reaction system for aqueous glycerol hydrogenolysis (Scheme 1).^9,10 In the reaction system, the mixed-metal oxides, which are originated from HT, could in situ reconstruct into the structure of HT. Furthermore, the basal spacing of the reconstructed HT may be enhanced by the intercalation of guest species such as reactants, reactive intermediates, or products. The basal spacing of reconstructed HT could supply effective information about the catalytic mechanism of glycerol hydrogenolysis. In this work, metal Cu was loaded on the hydrotalcite-derived Mg–Al mixed oxides to prepare Cu/Mg–Al mixed-oxide catalysts, and the Cu/Mg–Al mixed-oxide catalysts were used in glycerol hydrogenolysis. We determined the structural change of the catalysts to study the catalytic mechanism based on the basal spacing of HT. Using the basal spacing of the reconstructed hydrotalcite as an effective tool, the effect of H₂ pressure on glycerol hydrogenolysis was elucidated.

2. Experimental section

2.1 Preparation of catalysts

2.1.1 Preparation of CO₃²⁻-intercalated HT. CO₃²⁻-intercalated HT with an M²⁺ : M³⁺atomic ratio of 3 : 1 were prepared using a standard aqueous co-precipitation method, as described in our previous report.²⁹ The M²⁺ used was Mg²⁺, or the mixture of Cu²⁺ and Mg²⁺; and the M³⁺ used was Al³⁺. An aqueous solution (300 mL) of the metal nitrates in a desired M²⁺ : M³⁺ molar ratio with a total concentration of 1.5 M was mixed slowly with continuous stirring with an alkaline solution of Na₂CO₃–NaOH. The molar amount of Na₂CO₃ was twice that of M³⁺. The pH of the mixture was kept constant, typically at values between 9 and 10, by adjusting the flow rate of the alkaline solution. The temperature was maintained at 60 °C. Following this addition, which resulted in the formation of a heavy slurry, the mixture was aged at 60 °C for 18 h with stirring, to facilitate the selective growth of the precipitated HT phase. The slurry was then cooled to 25 °C, filtered, and washed with water until the pH value of the filtrate was near 7. The precipitate was dried at 90 °C for 16 h. The resulting material was CO₃²⁻-intercalated HT (hereafter indicated as HT-CO₃²⁻). Following this method, the prepared HT-CO₃²⁻ were Cu/Mg–Al HT-CO₃²⁻ and Mg–Al HT-CO₃²⁻.

2.1.2 Preparation of OH⁻-intercalated HT. The OH⁻-intercalated HT was prepared according to the procedure described in the literature.²⁸ The Mg–Al HT-CO₃²⁻ was calcined at 500 °C for 3 h in static air, to form the corresponding mixed oxides (hereafter indicated as Mg(Al)O). The Mg(Al)O was immersed in decarbonated water at room temperature for 1 h, leading to OH⁻-intercalated Mg–Al HT (hereafter indicated as Mg–Al HT-OH⁻).

2.1.3 Preparation of supported Cu catalysts. Two methods (co-precipitation and impregnation) were used for the preparation of Cu catalysts.

The Cu/Mg–Al HT-CO₃²⁻ samples, which contained Cu²⁺ and were prepared by co-precipitation, were calcined at 500 °C for 3 h in static air, to form the corresponding mixed oxides containing Cu²⁺. After reduction in hydrogen, the supported Cu catalysts, which were prepared by a co-precipitation method, were obtained.

The calcined sample of Mg–Al HT-CO₃²⁻ (Mg(Al)O) was used for impregnation of copper nitrate by a wet impregnation method using an excess amount of water. The solution of copper nitrate (0.54 g in 20 mL of deionized water) was added slowly, under stirring, to the calcined sample of Mg–Al HT-CO₃²⁻ (2.0 g). The mixture was then dried at 90 °C for 12 h and calcined at 500 °C for 3 h. After reduction in hydrogen, the supported Cu catalysts were obtained.

All the supported Cu catalysts used in this study were reduced prior to the reaction in a tube furnace by passing a stream of hydrogen over the catalyst (mixed oxides) bed at 300 °C for 3 h. The reduced Cu catalysts were labeled Cu/Mg(Al)O.

2.2 Catalyst characterization

The surface area and pore character of the catalysts were determined using a Micromeritics ASAP 2020 instrument. The sample was degassed at 250 °C for 4 hours in N₂ prior to surface area measurement. The nitrogen adsorption and desorption isotherms were measured at −196 °C. Specific surface areas of the catalysts were determined by nitrogen adsorption data in the relative pressure range from 0.06 to 0.30 using the BET (Brunauer–Emmett–Teller) equation. Total pore volumes were estimated from the amount of nitrogen adsorbed at a relative pressure of 0.995. Pore volume and pore-size distribution curves were obtained from the analysis of the desorption branches of the nitrogen isotherms using the BJH (Barrett–Joyner–Halenda) method.

CO₂ temperature programmed desorption (TPD) experiments were performed using a Micromeritics ASAP 2020 instrument loaded with 100 mg of sample. The sample was pre-treated under He (20 mL min⁻¹) ramping at 15 °C min⁻¹ up to 600 °C where it was held for 30 min, and then cooled to 40 °C prior to the adsorption of CO₂. After the adsorption of CO₂ (25 mL min⁻¹) for 60 min, the catalyst was flushed with He (20 mL min⁻¹) for 60 min at 40 °C to remove the physisorbed gas from the surface of the catalyst, and the desorption profile was recorded employing a heating ramp of 10 °C min⁻¹ between 40 to 600 °C where it was held for 30 min.

X-ray diffraction patterns were recorded on a D/Max-3C X-ray powder diffractometer (Rigaku Co., Japan), using a Cu-Kα source fitted with an Inel CPS 120 hemispherical detector.

2.3 Catalytic hydrogenolysis of glycerol

The reaction of glycerol hydrogenolysis was carried out in a 75 mL stainless steel autoclave (Parr Instrument Co., USA) equipped with a temperature controller. The standard activity test was conducted under the following conditions: 210 °C reaction temperature, 9 MPa hydrogen pressure, 24 h reaction time, 80% (volume ratio of glycerol/(glycerol + water)) glycerol aqueous solution (15 mL), and 0.54 g of catalysts (catalyst/glycerol weight ratio, 3.6 wt%). The stirring speed was set constant at 600 rpm throughout the reaction. Reaction temperature and reacting atmosphere were changed to investigate the dependence of conversion and selectivity on the operating conditions. In addition, the concentration of glycerol aqueous solution was changed to investigate the effect of water content.

After reaction, the product sample was transferred to a 100 mL volumetric flask. The samples were prepared for analysis by adding a known amount of internal standard (n-butanol) into the volumetric flask. Ethanol was used as solvent to dilute the product sample. These samples were analyzed using a gas chromatograph (Shimadzu 2014) equipped with a flame ionization detector and a Rtx-Wax capillary column (30 m × 0.32 mm × 0.5 μm).

3. Results and discussion

3.1 BET surface area and pore size

As shown in Table 1, Cu/Mg(Al)O had high surface area (128–140 m² g⁻¹), which was not affected by the metal ratio of Cu : Mg : Al. Fig. 1 shows nitrogen physisorption isotherms and BJH pore size distribution for Cu/Mg(Al)O. The isotherm of Cu/Mg(Al)O has the characteristic Type IV shape. The Type IV isotherm and pore size distribution suggested that Cu/Mg(Al)O was a mesoporous material. The hysteresis loop was in the shape of Type A, suggesting that Cu/Mg(Al)O contained the open-ended cylindrical pores.

Table 1 Nitrogen physisorption data of the catalysts^a

Catalysts	Metal ratio		Surface area (m^2 g^−1)	Pore volume (cm^3 g^−1)	Average pore width (Å)
	Prepared	Measured^a
a Determined by EDS.
Cu/Mg(Al)O	1.5 : 1.5 : 1	1.81 : 1.64 : 1	140	0.59	183
	1 : 2 : 1	0.8 : 2.16 : 1	128	0.44	158
	2 : 1.5 : 1	1.85 : 1.68 : 1	138	0.53	159

---

Fig. 1 Nitrogen physisorption isotherms (□, adsorption; ○, desorption) and BJH pore size distribution for Cu/Mg(Al)O catalyst.

3.2 The basic strengths and basicity of the catalysts

It was already reported that the activity of Cu catalysts was affected by their basic strength and basicity.⁸ For this reason, the basic strengths and basicity of prepared catalysts were determined by a CO₂ TPD method. Fig. 2 shows the rate of CO₂ desorption, normalized to the sample loading, as a function of temperature for Cu/Mg(Al)O catalysts. The desorption profiles for the Cu/Mg(Al)O catalyst showed two apparent desorption peaks with the maximum at 100 °C and 250 °C, respectively. On this basis, Cu/Mg(Al)O had the basic site of weak and medium strength.

Fig. 2 CO₂-TPD pattern of Cu/Mg(Al)O catalyst.

3.3 Effect of reaction variables

Five reaction variables were tested. The details about the effect of reaction time and catalyst amount are given in the ESI† (Fig. S1 and S2). The data about other reaction variables are given below.

3.3.1 Effect of reaction temperature. Fig. 3 shows the effect of reaction temperature on glycerol conversion and 1,2-propanediol yield. The glycerol conversion increased with reaction temperature, and reached >95% when the reaction temperature was above 210 °C. The 1,2-propanediol yield also increased with reaction temperature. The yield reached a maximum when reaction temperature was 210 °C (Fig. 3). Above 210 °C, the yield decreased with increase in the reaction temperature.

Fig. 3 Effect of reaction temperature on glycerol hydrogenolysis. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalyst, 80% glycerol aqueous solution, reaction time 24 h, hydrogen pressure 4 MPa.

3.3.2 Effect of hydrogen pressure. Fig. 4 shows the effect of hydrogen pressure on glycerol conversion and 1,2-propanediol yield. The conversion of glycerol increased with hydrogen pressure, and reached >98% when hydrogen pressure was above 7 MPa. The yield of 1,2-propanediol also increased with hydrogen pressure. The yield reached a maximum (>82%) when hydrogen pressure was above 7 MPa (Fig. 4). In contrast to the effect of hydrogen pressure on conversion, hydrogen pressure has a stronger effect on the yield of 1,2-propanediol.

Fig. 4 Effect of hydrogen pressure on glycerol hydrogenolysis. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalyst, 80% glycerol aqueous solution, reaction time 24 h, reaction temperature 210 °C.

3.3.3 Effect of glycerol concentration. Fig. 5 shows the effect of glycerol concentration on the activity of Cu/Mg(Al)O catalysts. Under conditions of H₂ pressure 9 MPa and reaction temperature 210 °C, both the conversion of glycerol and the yield of 1,2-propanediol increased with the glycerol concentration (Fig. 5A). Under conditions of H₂ pressure 4 MPa and reaction temperature 200 °C, a similar phenomenon was observed (Fig. 5B).

Fig. 5 The conversion/yield change versus glycerol concentration in the presence of Cu/MgAl(O) catalyst. (A) H₂ pressure 9 MPa and reaction temperature 210 °C; (B) H₂ pressure 4 MPa and reaction temperature 200 °C. Reaction conditions: 3 wt% of catalyst, reaction time 24 h, glycerol aqueous solution (15 mL).

3.4 Reusability of catalyst

We attempted to reuse the Cu/Mg(Al)O catalyst by the method as follows. The used Cu/Mg(Al)O catalyst was separated from the reaction solution by centrifugation. It was taken as the catalyst for the repeated reactions without any further treatment. The repeated reactions were carried out under the same reaction conditions as those of the previous reaction. The catalyst was investigated for eight successive uses and the results are shown in Fig. 6. The Cu/Mg(Al)O catalyst can be repeatedly used 8 times with no apparent loss of activity (glycerol conversion, >95%). The yield of 1,2-propanediol was also unchanged when the catalyst was used less than 5 times. However, the yield decreased slightly after the catalyst was used more than 6 times.

Fig. 6 Reusability of catalysts. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalyst, 100% glycerol (15 mL), reaction time 24 h, reaction temperature 210 °C, hydrogen pressure 9 MPa.

3.5 Catalytic mechanism

3.5.1 XRD analysis of Cu catalysts before and after reaction of glycerol. Fig. 7 presents the XRD patterns of Cu/Mg–Al HT-CO₃²⁻, Cu/Mg–Al HT-CO₃²⁻ after calcination, Cu/Mg(Al)O catalyst before reaction, and Cu/Mg(Al)O catalyst after reaction. The X-ray diffractograms of Cu/Mg–Al HT-CO₃²⁻ showed the typical X-ray diffractograms of HT, with characteristic diffraction peaks at 11.4°, 23.0°, 34.9° (Fig. 7a).^24–26 The thermal pre-treatment of HT results in changed XRD patterns, caused by the structural changes associated with the loss of CO₂ and H₂O from the starting material. The Cu–Mg–Al HT-CO₃²⁻ after calcination gave a mixed oxide phase of CuO, MgO and Al₂O₃, with diffraction lines very similar to that of CuO (Fig. 7b). The prepared Cu/Mg(Al)O catalyst showed reflections characteristic of Cu (Fig. 7c). The XRD patterns of the Cu/Mg(Al)O catalyst after reaction were different from the patterns before reaction. After reaction, the XRD patterns of Cu/Mg(Al)O catalysts showed, in addition to Cu, the diffraction peaks characteristic of HT (Fig. 7d).^24–26 It means that Mg(Al)O supports in situ reconstructed into the structure of HT. For the case of the reconstructed HT, the intensities of the original (003) and (006) reflections decreased substantially and additional reflections at 6.2° and 18.3° were now observed after reaction. The interlayer space of HT was increased from 7.8 to 15.0 Å. HT has the property of tunable basal spacing. The basal spacing includes two parts, the thickness of the hydroxide layer and the interlayer space. The thickness of the hydroxide layer does not change. The interlayer space changes with the size of intercalated guest species. The basal spacing of small inorganic species (CO₃²⁻ and OH⁻) intercalated HT is 7.8 Å. After reaction of glycerol hydrogenolysis, the inorganic anionic species OH⁻ existed in the interlayer space of HT for balancing the residual positive charge of the hydroxide layer. Without the existence of other guest species, the basal spacing would keep the size of 7.8 Å. If other species intercalated into the interlayer space, the basal spacing would increase with the size of guest species because the guest species occupy a certain space.^24–26 The basal spacing of catalyst after reaction increased from 7.8 to 15 Å. This implied that guest species besides OH⁻ existed in the interlayer space. The intercalated guest species expanded the interlayer space. In the reaction system of glycerol hydrogenolysis, the possible guest species are glycerol (reactant), reactive intermediates, or 1,2-propanediol (product).

Fig. 7 XRD patterns of (a) Cu–Mg–Al HT-CO₃²⁻, (b) Cu–Mg–Al HT-CO₃²⁻ after calcination, (c) Cu/Mg(Al)O catalyst before reaction, (d) Cu/Mg(Al)O catalyst after reaction. Reaction conditions: 6 wt% of catalyst, 80% glycerol aqueous solution (15 mL), reaction temperature 200 °C, hydrogen pressure 4 MPa.

Cu/Mg(Al)O catalysts, containing different ratios of Cu : Mg : Al and prepared by a co-precipitation method or an impregnation method, were tested. The XRD patterns of the various Cu/Mg(Al)O catalysts after reaction are shown in Fig. S3 (ESI†). All the used catalysts displayed, in addition to Cu, the structure of HT-guest. This means that the formation of HT-guest was not affected by the preparation methods (i.e., co-precipitation method or impregnation method) and ratio of Cu : Mg : Al.

Fig. 8 shows the XRD patterns of Cu/Mg(Al)O catalysts after reaction of neat glycerol under conditions of H₂ pressure 9 MPa and reaction temperature 210 °C. The XRD patterns of Cu/Mg(Al)O after the first run, 3rd run and 8th run display, in addition to Cu, the diffraction peaks characteristic of HT-guest.

Fig. 8 XRD patterns of Cu/Mg(Al)O catalyst after (a) the first run, (b) the third run, and (c) the eighth run. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalyst, neat glycerol (15 mL), reaction time 24 h, reaction temperature 210 °C, hydrogen pressure 9 MPa.

Fig. 9 shows the XRD patterns of Cu/Mg(Al)O catalysts after reaction over different reaction temperatures. In the reaction temperature range of 180–230 °C, all the used catalysts showed the structure of HT-guest. This indicated that the formation of HT-guest was not related to reaction temperature.

Fig. 9 XRD patterns of Cu/Mg(Al)O catalysts after reaction run under different reaction temperatures: (a) 180 °C, (b) 200 °C, (c) 210 °C, (d) 230 °C. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalyst, 80% glycerol aqueous solution, reaction time 24 h, hydrogen pressure 4 MPa.

Fig. 10 displays the XRD patterns of Cu/Mg(Al)O catalysts after reaction over different hydrogen pressures. Below 4 MPa, the used catalyst showed the structure of HT-guest; while above 7 MPa, the used catalyst showed the structure of HT-OH⁻. This indicated that the formation of HT-guest was affected by hydrogen pressure.

Fig. 10 XRD patterns of Cu/Mg(Al)O catalysts after reaction run under different hydrogen pressures: (a) 2 MPa, (b) 4 MPa, (c) 7 MPa, (d) 9 MPa, (e) 11 MPa. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalyst, 80% glycerol aqueous solution, reaction time 24 h, reaction temperature 210 °C.

Fig. 11 shows the effect of glycerol concentration on structure of Cu/Mg(Al)O catalysts. Below the concentration of 60% glycerol aqueous solution, the used catalyst displayed diffraction peaks of reconstructed HT which were similar to those of HT-OH⁻ (Fig. 11 a–c). Above 80% concentration of glycerol aqueous solution, the used catalyst displayed apparent diffraction peaks at 6.2° and 18.3° and a weak diffraction peak at 11.4° (Fig. 11d and e). The reflections at 6.2° and 18.3° were characteristic of HT-guest, while that at 11.4° was associated with HT-OH⁻. This indicates that the used catalyst contained HT-guest as the major phase and HT-OH⁻ as the minor phase. On this basis, it is assumed that the structure of used catalyst was affected by reactant concentrations.

Fig. 11 XRD patterns of Cu/Mg(Al)O catalysts after reaction run under different concentrations of glycerol solution: (a) 20%, (b) 40%, (c) 60%, (d) 80%, (e) 100%. Reaction conditions: 3 wt% of catalyst, reaction time 24 h, H₂ pressure 4 MPa, reaction temperature 200 °C.

3.5.2 XRD analysis of Cu catalysts before and after reaction of 1,2-propanediol. 1,2-Propanediol is the main product of glycerol hydrogenation. In order to investigate the possibility of 1,2-propanediol being intercalated into HT, the used Cu/Mg(Al)O after reaction of 1,2-propanediol hydrogenolysis was analyzed by an XRD method. The intensity of diffraction peaks at 6.2° and 18.3° was used to estimate the crystallinity of HT-guest, and the intensity of diffraction peak at 11.4° was used to estimate the crystallinity of HT-OH⁻. The results are shown in Fig. 12. In this reaction, 1,2-propanediol was taken as a reactant, and the reaction was run under similar conditions to those of glycerol hydrogenation. The reaction conditions were: H₂ pressure 0.1–9 MPa, reaction temperature 180–210 °C and 1,2-propanediol concentration between 80% and 5%. Under the reaction conditions, almost each used catalyst contained both the crystalline phases of HT-guest and HT-OH⁻. Their crystallinity was affected by 1,2-propanediol concentration. Under H₂ pressure 9 MPa, reaction temperature 210 °C and 1,2-propanediol concentration between 80% and 50%, the crystallinity of HT-guest phase decreased with decrease in the concentration of 1,2-propanediol (Fig. 12a–c). Under H₂ pressure 4 MPa, reaction temperature 200 °C and 1,2-propanediol concentration between 80% and 5%, the crystallinity of HT-guest phase also decreased with decrease in the concentration of 1,2-propanediol (Fig. 12e–i). However, the crystallinity of the HT-OH⁻ phase increased with decrease in the concentration of 1,2-propanediol (Fig. 12e–i). The crystalline structure of used catalysts was also affected by the reaction pressure. Under the same reaction temperature (210 °C) and concentration of 1,2-propanediol (50%), the crystallinity of the HT-guest phase in used catalysts under 2 MPa H₂ pressure (Fig. 12d) was higher than that under 9 MPa H₂ pressure (Fig. 12c). The other example is the reaction which was run under the same reaction temperature (180 °C) and concentration of 1,2-propanediol (5%), the crystallinity of the HT-guest phase in used catalysts under 0.1 MPa H₂ pressure (Fig. 12k) was higher than that under 4 MPa (Fig. 12j).

Fig. 12 XRD patterns of Cu/Mg(Al)O catalysts after reaction of 1,2-propanediol hydrogenolysis which was run under different conditions: (a)–(c) H₂ pressure 9 MPa, reaction temperature 210 °C, and different concentrations of 1,2-propanediol aqueous solution (a, 80%; b, 67%; c, 50%); (d) H₂ pressure 2 MPa, reaction temperature 210 °C, and 1,2-propanediol aqueous solution (50%); (e)–(i) H₂ pressure 4 MPa, reaction temperature 200 °C and different concentrations of 1,2-propanediol aqueous solution (e, 80%; f, 67%; g, 50%; h, 20%; i, 5%); (j and k) reaction temperature 180 °C, 1,2-propanediol aqueous solution (5%), and different H₂ pressure (j, 4 MPa; k, 0.1 MPa). Reaction conditions: 6 wt% of Cu/Mg(Al)O catalysts, reaction time 24 h.

For the reaction of 1,2-propanediol, the reaction solution was just aqueous solution of 1,2-propanediol. No reaction products were identified at the end of either reaction, indicating that 1,2-propanediol did not react under the reaction conditions. Since the bulk solution was 1,2-propanediol aqueous solution throughout the whole reaction process, it is very possible that 1,2-propanediol is intercalated into the interlayer galleries of HT-guest. Furthermore, the formation of HT-guest was related to the reaction conditions. The higher the 1,2-propanediol concentration and the lower the H₂ pressure, the more the possibility of 1,2-propanediol being intercalated into the interlayer galleries of HT-guest.

3.5.3 XRD analysis of Cu catalysts before and after reaction of 1,3-propanediol. 1,3-Propanediol is the other possible product of glycerol hydrogenation although it was not detected on Cu doped catalysts in this work.⁴ The possibility of 1,3-propanediol being intercalated into HT was investigated by a similar method to that of 1,2-propanediol. The result is shown in Fig. 13. Under H₂ pressure of 0.1 MPa and 1 MPa, the used Cu/Mg(Al)O catalyst contained HT-guest as the major phase and HT-OH⁻ as the minor phase (Fig. 13a and b). However, under H₂ pressure of 4 MPa, the used Cu/Mg(Al)O catalyst showed the structure of HT-OH⁻, suggesting that no chemical was intercalated into the interlayer galleries to expand the layers (Fig. 13c). This indicated that the formation of the HT-guest phase was affected by H₂ pressure, which is consistent with the results regarding 1,2-propanediol hydrogenation.

Fig. 13 XRD patterns of Cu/Mg(Al)O catalysts after reaction run under different H₂ pressures: (a) 0.1 MPa, (b) 1 MPa, (c) 4 MPa. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalysts, reaction time 24 h, reaction temperature 200 °C, and 80% 1,3-propanediol aqueous solution.

The reaction mechanism of 1,3-propanediol hydrogenolysis was similar to that of glycerol hydrogenolysis.³⁰ The product analysis showed that part of 1,3-propanediol was converted to n-propanol and H₂O (Scheme 3). Under 0.1 MPa, the conversion of 1,3-propanediol was 66% and the yield of n-propanol was 11%; under 4 MPa, the conversion of 1,3-propanediol was 76% and the yield of n-propanol was 59% (Fig. 14). Therefore, in contrast to the effect of hydrogen pressure on conversion, hydrogen pressure has a stronger effect on the yield of n-propanol.

Scheme 3 Hydrogenolysis of 1,3-propanediol.

Fig. 14 Effect of hydrogen pressure on 1,3-propanediol hydrogenolysis. Reaction conditions: 6 wt% of Cu/Mg(Al)O catalysts, reaction time 24 h, reaction temperature 200 °C, and 80% 1,3-propanediol aqueous solution.

3.5.4 XRD analysis of Mg(Al)O before and after reaction. Mg(Al)O was the support of the Cu/Mg(Al)O catalyst. The structure of the Mg(Al)O support after reaction was tested. As shown in Fig. 15, Mg(Al)O could rehydrate into HT-OH⁻ in the aqueous solution of 1,2-propanediol (concentration, 80%) or glycerol (concentration, <60%). However, 1,2-propanediol or glycerol could not intercalate into the interlayer galleries to form the structure of HT-guest since the interlayer distance of rehydrated HT did not increase.

Fig. 15 XRD patterns of (a) Mg(Al)O, (b) Mg(Al)O catalysts after reaction of 80% 1,2-propanediol aqueous solution, (c)–(e) Mg(Al)O catalysts after hydrogenolysis reaction of glycerol with different concentrations (c, 80%; d, 60%; e, 10%). Reaction conditions: 6 wt% of Mg(Al)O materials, reaction time 24 h, H₂ pressure 4 MPa, reaction temperature 200 °C.

3.5.5 XRD analysis of Mg–Al HT before and after reaction. Fig. 16 shows the XRD patterns of Mg–Al HT-CO₃²⁻ before and after reaction. It was found that the used Mg–Al HT-CO₃²⁻ still kept the original structure of HT-CO₃²⁻, although its crystallinity was affected by the concentration of glycerol. A similar phenomenon was found for Mg–Al HT–OH⁻ (Fig. 17). This indicates that, in the absence of Cu, 1,2-propanediol or glycerol could not intercalate into the interlayer galleries of HT.

Fig. 16 XRD patterns of (a) Mg–Al HT-CO₃²⁻, (b–d) Mg–Al HT-CO₃²⁻ catalyst after hydrogenolysis reaction (b, 80% glycerol aqueous solution; c, 80% 1,2-propanediol aqueous solution; d, neat glycerol). Reaction conditions: 6 wt% of Mg–Al HT-CO₃²⁻ catalysts, reaction time 24 h, H₂ pressure 4 MPa, reaction temperature 200 °C.

Fig. 17 XRD patterns of (a) Mg–Al HT-OH⁻, (b–d) Mg–Al HT-OH⁻ catalyst after hydrogenolysis reaction (b, 80% glycerol aqueous solution; c, 80% 1,2-propanediol aqueous solution; d, neat glycerol). Reaction conditions: 6 wt% of Mg–Al HT-OH⁻ catalyst, reaction time 24 h, H₂ pressure 4 MPa, reaction temperature 200 °C.

3.5.6 Relation between reaction conditions and desorption of product. The structure analysis using XRD patterns showed that pure supports (Mg(Al)O, HT-CO₃²⁻, HT-OH⁻) after reaction did not form the structure of HT-guest with enhanced basal spacing (Scheme 4, Fig. 15 and 16). This indicated that guest species could not intercalate into the interlayer gallery of pure supports in the absence of Cu. However, as shown in Fig. 7, the Mg(Al)O-supported Cu catalyst could form the structure of HT-guest with enhanced basal spacing (d = 15.0 Å). This suggested that the presence of Cu may play a key role in the intercalation of guest species into the interlayer space of HT. Since catalysts could chemisorb the reactants, reactive intermediates, or products, the role of Cu in the formation of HT-guest may be due to its interaction with guest species. On this basis a further question can be posed as to the structure of the guest species.

Scheme 4 The structure of pure supports (i.e. Mg(Al)O, HT-CO₃²⁻, HT–OH⁻) after reaction.

The intercalation possibility of glycerol (reactant) and 1,2-propanediol (product) is discussed below. Fig. 11 indicates that the formation of HT-guest was under the condition of glycerol concentration >80%. Therefore glycerol seems to be the guest species. However, Fig. 5 shows that the conversion of glycerol changed with its concentration and the glycerol concentration was not constant. Table 2 shows the concentration of glycerol and 1,2-propanediol pre-reaction and post-reaction. The glycerol concentration after reaction did not increase apparently when its initial concentration increased. When the initial concentration of glycerol was 40%, the concentration after reaction was 24% (Table 2, entry 2). When the initial concentration of glycerol was increased from 40% to 100%, the concentration of glycerol after reaction increased from 24% to 29% (Table 2, entry 5). However, the 1,2-propanediol concentration post-reaction increased apparently when the initial concentration of glycerol increased. When the initial concentration of glycerol was 40%, the 1,2-propanediol concentration post-reaction was 11%. When the initial concentration of glycerol was increased to 100%, concentration of 1,2-propanediol post-reaction increased to 51%. This implied that, compared with glycerol, 1,2-propanediol had more possibility of being the guest species. Besides, the results in Section 3.5.2 had proved that 1,2-propanediol could be intercalated into the interlayer space of HT.

Table 2 The concentration of glycerol and 1,2-propanediol pre-reaction and post-reaction^a

Entry	Glycerol		1,2-Propanediol		Ethylene glycol^b		The structure of used catalyst
	Pre-	Post-	Pre-	Post-	Pre-	Post-
a Derived from the data of Fig. 5 and Fig. 11. b By-product.
1	20%	14%	0	4%	0	0.93%	HT-OH^−
2	40%	24%	0	11%	0	1.5%	HT-OH^−
3	60%	27%	0	22%	0	1.8%	HT-OH^−
4	80%	27%	0	36%	0	1.9%	HT-guest
5	100%	29%	0	50%	0	1.6%	HT-guest

---

Table 3 also shows the concentration of glycerol and 1,2-propanediol pre-reaction and post-reaction. In Table 3, the initial concentration of glycerol was set constant at 80% and the H₂ pressure was a variable. The concentration of glycerol post-reaction decreased with the increase in H₂ pressure, and went down to the lowest value (<1.4%) when the H₂ pressure was above 7 MPa. However, the concentration of 1,2-propanediol post-reaction increased with the increase in H₂ pressure, and went up to the maximum value (62%) when the H₂ pressure was above 7 MPa. When the H₂ pressure was above 7 MPa, the crystalline phase of HT-OH⁻ was formed (Table 3, entries 3–5). This revealed that the crystalline phase of HT-OH⁻ was formed when 1,2-propanediol reached its maximum concentration (62%) and glycerol went down to its lowest concentration (<1.4%). Why was the HT-OH⁻ formed when 1,2-propanediol reached its maximum concentration? This could be ascribed to the effect of H₂ pressure. The concentration of H₂ dissolved in the bulk solution may increase with H₂ pressure, which accordingly increased the amount of H₂ chemisorbed on the active sites of Cu and decreased the amount of the chemisorbed 1,2-propanediol. This implied that the chemisorption of 1,2-propanediol on Cu was affected by H₂ pressure. The discussion in the previous section suggested that the formation of HT-guest was ascribed to the chemisorption of 1,2-propanediol on Cu. Therefore, the effect of H₂ pressure on the formation of HT-guest may result from its effect on chemisorption of 1,2-propanediol on Cu. This proposal was corroborated by the results in Section 3.5.2, where, for the same concentration of 1,2-propanediol, HT-guest was more easily formed at low H₂ pressure.

Table 3 The concentration of glycerol and 1,2-propanediol versus H₂ pressure^a

Entry	H2 pressure (MPa)	Glycerol		1,2-Propanediol		Ethylene glycol^b		The structure of used catalyst
		Pre-	Post-	Pre-	Post-	Pre-	Post-
a Derived from the data of Fig. 4 and Fig. 10. b By-product.
1	2	80%	12%	0	36.51%	0	2.3%	HT-guest
2	4	80%	3.9%	0	55.45%	0	2.9%	HT-guest
3	7	80%	0.73%	0	59.93%	0	3.3%	HT-OH^−
4	9	80%	1.4%	0	60.72%	0	3.7%	HT-OH^−
5	11	80%	0.72%	0	60.22%	0	3.3%	HT-OH^−

---

The other possible guest species were by-products. The by-product in the liquid phase, which was detected in this system, was ethylene glycol. As shown in Table 3, the concentration of ethylene glycol was low and similar whether HT-guest was formed or not. This implied that there was very low possibility for ethylene glycol to be a guest species.

The proposed mechanism of glycerol hydrogenolysis on reconstructed HT is depicted in Scheme 5. The Mg(Al)O shows a memory effect, recovering the original lamellar structure when immersed in the reactive system of glycerol hydrogenolysis. We suspect that guest species (i.e., 1,2-propanediol) were intercalated into the interlayer galleries, thus expanding the layers. The intercalation of glycerol or 1,2-propanediol was due to a driving force, i.e. the interaction between copper and glycerol or 1,2-propanediol. The size of glycerol or 1,2-propanediol was calculated, employing the quantum chemical method of Gaussian 09.³¹ The result showed that the size of glycerol or 1,2-propanediol was in the range of 4.7–6.3 Å (Table S1, ESI†). This implied that glycerol or 1,2-propanediol occupied a certain space. Accordingly, this explained why the intercalation of 1,2-propanediol would increase the basal spacing of HT. The formation of HT-guest was affected by the reaction conditions. In the case of glycerol hydrogenolysis with low H₂ pressure and high concentration of 1,2-propanediol, the used catalyst showed the structure of HT-guest with enhanced basal spacing (basal spacing, d = 15.0 Å), as seen from the results shown in Fig. 10a and b and Fig. 11d and e; however, under the reverse reaction conditions, it showed the structure of HT-OH⁻ (basal spacing, d = 7.8 Å), as seen from the results shown in Fig. 10c–e and Fig. 11a–c.

Scheme 5 Mechanism of glycerol hydrogenolysis on reconstructed HT.

The basal spacing of HT could work as an effective tool to explore the effect of pressure on glycerol hydrogenolysis. The formation of HT-guest means that the product 1,2-propanediol is strongly adsorbed on the active site of Cu. This chemisorbed 1,2-propanediol may crack into the gas phase product,⁹ which decreased the yield of 1,2-propanediol. One proof is the experimental results shown in Fig. 4, where the conversion of glycerol was slightly affected by the hydrogen pressure, the yield of 1,2-propanediol, however, increased apparently with hydrogen pressure, and reached >82% when hydrogen pressure was above 7 MPa. Under the condition of hydrogen pressure below 4 MPa, the structure analysis using XRD showed that the catalyst had the structure of HT-guest (Fig. 10a and b); while above 7 MPa, the catalyst had the structure of HT-OH⁻ (Fig. 10c–e). Therefore, the lower yield of 1,2-propanediol below 4 MPa of hydrogen pressure may be due to the strong chemisorption of 1,2-propanediol on catalysts; while the higher yield of 1,2-propanediol above 4 MPa was ascribed to the easy desorption of 1,2-propanediol from catalysts.

The proposal above was also fit to explain the hydrogenolysis of 1,3-propanediol. The formation of HT-guest showed little influence on the reaction conversion of 1,3-propanediol. However, the yield with respect to n-propanol was affected by it (Fig. 14). When the used catalyst showed the structure of HT-guest (Fig. 13a), the yield of n-propanol was only 11% (Fig. 14); however, when it had the structure of HT-OH⁻ (Fig. 13c), the yield of n-propanol was increased to 59% (Fig. 14).

4. Conclusions

HT has the property of memory effect and tunable basal spacing. The formation of HT-guest was affected by the reaction conditions. In the case of glycerol hydrogenolysis at low H₂ pressure and high concentration of 1,2-propanediol, the used catalyst showed the structure of HT-guest with enhanced basal spacing (basal spacing, d = 15.0 Å); however, under the reverse reaction conditions, it showed the structure of HT-OH⁻ (basal spacing, d = 7.8 Å). It was found that the basal spacing of HT could work as an effective tool, reflecting whether or not the guest species (i.e., 1,2-propanediol) was strongly adsorbed on the active sites of Cu. With this tool, the effect of H₂ pressure on glycerol hydrogenolysis was elucidated. H₂ pressure affected the adsorption of 1,2-propanediol on catalysts. The higher the H₂ pressure, the easier the desorption of 1,2-propanediol from catalysts. The lower yield of 1,2-propanediol below 4 MPa of hydrogen pressure was due to the strong chemisorption of 1,2-propanediol on catalysts; while the higher yield of 1,2-propanediol above 4 MPa was ascribed to the easy desorption of 1,2-propanediol from catalysts. In addition, this method may be used to study the catalytic mechanism in other reactions with HT as a catalyst or a catalyst support.

Acknowledgements

This work was supported by the project funded by Natural Science Basic Research Plan in Shaanxi Province of China (Program No. 2010JM2003) and by the Fundamental Research Funds for the Central Universities (Program No. GK200902006).

References

1. E. J. Steen, Y. Kang, G. Bokinsky, Z. Hu, A. Schirmer, A. McClure, S. B. Cardayre and J. D. Keasling, Nature, 2010, 463, 559.
2. M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi and C. D. Pina, Angew. Chem., Int. Ed., 2007, 46, 4434.
3. C. H. Zhou, J. N. Beltramini, Y. X. Fan and G. Q. Lu, Chem. Soc. Rev., 2008, 37, 527.
4. Y. Nakagawa, Y. Shinmi, S. Koso and K. Tomishige, J. Catal., 2010, 272, 191.
5. M. Ayoub and A. Z. Abdullah, Renewable Sustainable Energy Rev., 2012, 16, 2671.
6. M. A. Medeiros, C. M. M. Leite and R. M. Lago, Chem. Eng. J., 2012, 180, 364.
7. M. Balaraju, V. Rekha, P. S. S. Prasad, B. L. A. P. Devi, R. B. N. Prasad and N. Lingaiah, Appl. Catal., A, 2009, 354, 82.
8. E. P. Maris and R. J. Davis, J. Catal., 2007, 249, 328.
9. D. Roy, B. Subramaniam and R. V. Chaudhari, Catal. Today, 2010, 156, 31.
10. A. Wawrzetz, B. Peng, A. Hrabar, A. Jentys, A. A. Lemonidou and J. A. Lercher, J. Catal., 2010, 269, 411.
11. E. D' Hondt, S. V. d. Vyver, B. F. Sels and P. A. Jacobs, Chem. Commun., 2008, 6011.
12. I. Gandarias, P. L. Arias, J. Requies, M. B. Güemez and J. L. G. Fierro, Appl. Catal., B, 2010, 97, 48.
13. Z. Huang, F. Cui, H. Kang, J. Chen, X. Zhang and C. Xia, Chem. Mater., 2008, 20, 5090.
14. L. C. Meher, R. Gopinath, S. N. Naik and A. K. Dalai, Ind. Eng. Chem. Res., 2009, 48, 1840.
15. Z. Yuan, L. Wang, J. Wang, S. Xia, P. Chen, Z. Hou and X. Zheng, Appl. Catal., B, 2011, 101, 431.
16. A. Perosa and P. Tundo, Ind. Eng. Chem. Res., 2005, 44, 8535.
17. E. P. Maris, W. C. Ketchie, M. Murayama and R. J. Davis, J. Catal., 2007, 251, 281.
18. L. Ma and D. He, Top. Catal., 2009, 52, 834.
19. D. Coll, F. Delbecq, Y. Aray and P. Sautet, Phys. Chem. Chem. Phys., 2011, 13, 1448.
20. F. Auneau, C. Michel, F. Delbecq, C. Pinel and P. Sautet, Chem.–Eur. J., 2011, 17, 14288.
21. T. W. Kim, M. Sahimi and T. T. Tsotsis, Ind. Eng. Chem. Res., 2009, 48, 5794.
22. Y. Zhao, S. Zhang, B. Li, H. Yan, S. He, L. Tian, W. Shi, J. Ma, M. Wei, D. G. Evans and X. Duan, Chem.–Eur. J., 2011, 17, 13175.
23. I. Gualandi, E. Scavetta, S. Zappoli and D. Tonelli, Biosens. Bioelectron., 2011, 26, 3200.
24. G. Centi and S. Perathoner, Microporous Mesoporous Mater., 2008, 107, 3.
25. S. K. Jana, Y. Kubota and T. Tatsumi, J. Catal., 2008, 255, 40.
26. J. Liang, R. Ma, N. Iyi, Y. Ebina, K. Takada and T. Sasaki, Chem. Mater., 2010, 22, 371.
27. D. P. Debecker, E. M. Gaigneaux and G. Busca, Chem.–Eur. J., 2009, 15, 3920.
28. S. Abelló, F. Medina, D. Tichit, J. Pérez-Ramírez, J. C. Groen, J. E. Sueiras, P. Salagre and Y. Cesteros, Chem.–Eur. J., 2005, 11, 728.
29. C. Xu, J. Sun, B. Zhao and Q. Liu, Appl. Catal., B, 2010, 99, 111.
30. K. Wang, M. C. Hawley and T. D. Furney, Ind. Eng. Chem. Res., 1995, 34, 3766.
31. S. Hayase, D. A. Hrovat and W. T. Borden, J. Am. Chem. Soc., 2004, 126, 10028.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c2cy20144e

---