Hydrogen Production by Methanol Steam Reforming over CuO/ZnO/Al₂O₃ Catalysts Based on Oxalate-Precursor Synthesis

1. Introduction

Hydrogen energy has been called "the ultimate energy of the 21st century", and there is a huge development space for hydrogen energy. Among various hydrogen production methods, hydrogen production via methanol is increasingly favored by researchers, methanol as a raw material is easy to store and transport [1]. Currently, there are four main ways to produce hydrogen from methanol: methanol decomposition, (MD, equation (1)), partial oxidative reforming of methanol (POM，equation (2)) methanol steam reforming (MSR, equation (3)), and oxidized methanol steam reforming (OSRM, equation (4)), respectively [2].

CH₃OH → 2H₂ + CO ΔH = +128.0 kJ·mol^-1

(1)

CH₃OH + 0.5O₂ → 2H₂ + CO₂ ΔH = -192.2 kJ·mol^-1

(2)

CH₃OH + H₂O → 3H₂ + CO₂ ΔH = +49.4 kJ·mol^-1

(3)

CH₃OH + (1-n)H₂O + 0.5nO₂ → (3-n)H₂ + CO₂ ΔH= +49(1-n)-192.2n kJ·mol^-1

(4)

Among the four methods of hydrogen production, MSR is the most widely studied and the most in-depth method. MSR produce the highest amount of hydrogen per mole of methanol, and has the advantages of high purity of hydrogen and low content of CO [3]. Besides, MSR technology has low reaction temperature, low energy consumption and low investment [4].

At present, there are two types of catalysts used in MSR for hydrogen production [5,6]: one type is precious metal catalysts (such as Pd/ZnO, etc.); the other type is non-precious metal catalysts, including non-copper based catalysts (such as Zn-Cr, etc.) and copper based catalysts (such as CuO/ZnO/Al₂O₃, etc.).

Cu-based catalysts can produce H₂ with high selectively at low temperatures and have low CO selectivity, so Cu-based catalysts are widely used. The Cu-based catalysts for MSR have been widely studied, the effect of synergy between CuO and ZnO and the surface structure of CuO/ZnO/Al₂O₃ catalysts on the catalytic activities are critical [7,8].

Co-precipitation is a common method for preparing CuO/ZnO/Al₂O₃ catalysts, the precipitation process have a profound impact on the structure and performance of the prepared catalysts. Inui et al. [9] studied the effects of pH and temperature on catalyst precursors in the precipitation process. It indicated that the formation of Cu₂(NO₃)(OH)₃ is advantageous when pH ≤ 6, while (Cu,Zn)₂CO₃(OH)₂ is dominant when pH ≥ 7. The effect of temperature on the precursor is mainly to change the reaction rate, it has almost no effect on its phase composition. Spencer et al. [10] studied the phase transition process in the mother liquor, it indicated that amorphous Cu₂CO₃(OH)₂ first generated, which gradually transformed into (Cu,Zn)₂CO₃(OH)₂ during the aging process. Fang et al. [11] studied the effects of different feeding methods. Cu₂(NO₃)(OH)₃ mainly formed in the forward addition method, while amorphous Cu₂CO₃(OH)₂ mainly formed in the concurrent flow method, which interacts with Zn₅(CO₃)₂(OH)₆ and transforms into (Cu,Zn)₂CO₃(OH)₂ and (Cu,Zn)₅(CO₃)₂(OH)₆, respectively, a CuO-ZnO solid solution formed after decomposition, which is the active phase of MSR reaction.

The solvents and heating methods are the main factors in the precipitation process. Ma et al. [12] prepared CuO/ZnO/Al₂O₃ catalyst using ethanol and diethylene glycol as solvent, which possessed larger superficial area and exhibit higher catalytic performance. Zhang et al. [13,14] synthesized the smaller particle CuO/ZnO/Al₂O₃ catalyst by oxalate co-precipitation using ethanol as solvent, the catalyst showed better catalytic performance for MSR. Dai et al. [15] investigated the surface property of CuO/ZnO/Al₂O₃ catalysts prepared by oxalate co-precipitation, and explained that isomorphous substitution promoted synergy between CuO and ZnO and increased the superficial content of CuO.

Microwave irradiation heating was rapid and even in the preparation of catalysts, the active components was well-distributed on the support. Besides, microwave irradiation could control the micro structure of materials and enhance the selectivity of target product [16]. It is reported that microwave irradiation had obvious effects on preparation of ZnO and Al₂O₃ nanoparticles [17]. Zhang et al. [18] treated CuO/ZnO/Al₂O₃ catalysts with microwave irritation (200W) for 3 ~ 10 min, the catalyst micro structure was significantly improved and the catalytic activity of MSR increased by 7%. Fernández et al. [19] synthesized CuO/ZnO precursor mainly containing aurichalcite and CuO/ZnO/Al₂O₃ precursor only containing hydrotalcite-like, respectively, under microwave irradiation. The aurichalcite was burned into Cu-O-Zn solid-solution, which exhibited strong synergy and exhibited excellent activity and stability in MSR reaction.

At present, researchers generally believe that the strong synergy effect between Cu and Zn is beneficial for hydrogen production via MSR, many direct evidences for the evolution of the structure, morphology and coordination status of Cu-O-Zn solid solution were provided, which lay a foundation for identifying the active sites and studying the interface effect in the catalytic process [20,21,22].

In present work, the carbonate precursor is replaced by the oxalate precursor, and ethanol replaced water as a solvent, microwave replaced conventional heating as a heating method, the synergy effect between CuO and ZnO was further strengthened. The effects of solvents and heating methods on the composition of oxalate precursors, the structure, properties, and final MSR reaction performance of the calcined catalysts were studied from atomic to nano-scales.

2. Experimental

2.1. Catalyst preparation

The Cu-O-Zn/Al₂O₃ catalyst precursors were prepared by dropping simultaneously 1 mol/L Cu(NO₃)₂-Zn(NO₃)₂-Al(NO₃)₃ (Cu²⁺/Zn²⁺/Al³⁺ = 16/8/1 (molar ratio)) solution and 1 mol/L H₂C₂O₄ solution into a beaker with stirring constantly and keeping in water bath at 70 °C. Then the suspension was aged in microwave oven or water bath with circulating cooling equipment after co-precipitation, the aging process was conducted at 80 °C for 1 h. The precipitate was filtered and washed with distilled water or ethanol, then the precursor was obtained after drying at 110 °C for 12 h, and the catalyst was obtained after calcining the corresponding precursor at 350 °C for 4 h in air atmosphere. The precursor was designated as XYP and the catalyst was designated as XYC. Table 1 shows the summary the preparation conditions of the catalysts.

2.2. Catalyst characterization

X-ray diffraction (XRD) patterns of solid samples were recorded using a Rigaku D/max 2500 power diffractometer with Cu Kα radiation at 40 kV and 100 mA with a scanning rate of 8°/min in the 2θ ranges from 10° to 40°.

Temperature-programmed reduction (H₂-TPR) was performed in an Autochem Ⅱ 2920. About 20 mg catalyst sample was set in a U-mode quartz tube, pretreated in a Helium at 50 °C for 30 min, then heated to 300 °C at a rate of 10 °C /min, under a mixture of 10 vol% H₂/Ar (50 ml/min), the sample was then heated to 600 °C at a rate of 10 °C /min, The consumption of hydrogen was monitored by a thermal conductivity detector.

Differential thermal gravity (DTG) measurement was executed in a STA409C thermal analyzer. 30 mg sample was heated to 600 °C at a rate of 8 °C/min in a gas mixture of 20 vol% O₂/N₂ (50 ml/min).

JSM-6700F cold-field scanning electron microscope (SEM) was used to characterize the size and morphology of the samples.

BET specific surface areas and pore distribution of catalyst was measured with a SORPTMATIC 1990 automatic adsorption instrument employing N₂ as the adsorbent. BET specific surface areas were calculated by applying the Brunauer-Emmett-Teller (BET) method.

X-ray photoelectron spectroscopy (XPS) spectra of the samples were collected on an ESCAL-ab 220i-XL electron spectrometer using Al Kα radiation at 300 W. The samples were compressed into a pellet of 2 mm thickness and then mounted on a sample holder. The chamber was maintained at lower than 10⁻¹⁰ Torr. The binding energies were calibrated by C1s as the reference energy (C1s = 284.6 eV).

2.3. Catalytic performance of MSR

The performance evaluation of MSR was conducted on a continuous flow fixed bed device with a catalyst loading of 2 g, using industrial refined methanol as raw material. Evaluation conditions: raw material methanol aqueous solution (the molar ratio of water to methanol is 1.5), reaction temperature 260 °C, pressure 0.5 MPa, WHSV = 1.0 h^-1. After cooling, gas and liquid samples were analyzed using chromatographs equipped with PorpakT columns and TDX-01 columns, respectively, and thermal conductivity cell detectors. The methanol conversion rate and product distribution were calculated.

3. Results and discussion

3.1. XRD characterization of precursors

Figure 1 shows the XRD patterns of all catalyst precursors. It is seen that the phases of CuC₂O₄·xH₂O (2θ = 18.5°, 23.1°, 31.2°, 36°, 38.5°), α-ZnC₂O₄·2H₂O (2θ = 18.9°, 35.1°), and β-ZnC₂O₄ (2θ = 24°, 25.1°, 29.1°, 34.1°, 36.9°) were observed in precursor of WWP prepared by water solvent and water bath heating. But only diffraction peaks of CuC₂O₄·xH₂O (2θ = 23.1°, 36°, 38.5°) and weak peaks of β-ZnC₂O₄ (2θ = 24°, 36.9°) existed in precursor of EWP prepared by ethanol solvent and water bath heating, no α-ZnC₂O₄·2H₂O phase was observed. It indicated that ethanol solvent restrained the formation of α-ZnC₂O₄·2H₂O, and enhanced the phase selectivity of product. Microwave irradiation promoted the isomorphous substitution between Cu²⁺ of CuC₂O₄·xH₂O and Zn²⁺ of ZnC₂O₄ in mother liquid, thus, WMP and EMP mainly contained (Cu,Zn)C₂O₄ and partial CuC₂O₄ failing to be substituted during aging, so diffraction peak of CuC₂O₄·xH₂O (2θ = 23.1°, 36°) and β-ZnC₂O₄ (2θ = 24°, 36.9°) overlapped each other respectively and deviated original position. Compared with WWP and EWP, WMP and EMP didn’t contain diffraction peaks of α-ZnC₂O₄·2H₂O and β-ZnC₂O₄, and the crystal degree declined and failed to be detected owning to isomorphous substitution, it indicated that microwave irradiation have strong selectivity on the formation of crystal phase [23,24].

CuC₂O₄·xH₂O, α-ZnC₂O₄·2H₂O and β-ZnC₂O₄ generated in mother liquid in co-precipitation, as equation (1), (2) and (3). The isomorphous substitution mainly occurred in the aging process, it means that the Cu²⁺ of CuC₂O₄ entered into ZnC₂O₄ and the Zn²⁺ of ZnC₂O₄ entered into CuC₂O₄ formed (Cu,Zn)C₂O₄, as equation (4) and (5). However, the concentration of Cu²⁺ was higher than that of Zn²⁺, it primarily produced CuC₂O₄ as equation (1) in mother liquid, and the ZnC₂O₄ was little. In other words, the reaction (4) and (5) were promoted simultaneously by microwave irradiation, while CuC₂O₄ was not substituted completely.

Cu²⁺ + C₂O₄^2- + xH₂O → CuC₂O₄·xH₂O

(1)

Zn²⁺ + C₂O₄^2- + 2H₂O → α-ZnC₂O₄·2H₂O

(2)

Zn²⁺ + C₂O₄^2- → β-ZnC₂O₄

(3)

CuC₂O₄ + xZn²⁺→ (Cu_1-x,Zn_x)C₂O₄

(4)

ZnC₂O₄ + xCu²⁺→ (Cu_x,Zn_1-x)C₂O₄

(5)

3.2. DTG characterization of precursors

DTG curves of different catalyst precursors are shown in Figure 2. Three weight loss peaks for WWP were observed at 118 °C, 308 °C and 343 °C respectively, as shown in equation (6) ~ (9), the peak at 118 °C was due to the desorption of physically absorbed water of α-ZnC₂O₄·2H₂O, the peak around 308 °C was ascribed to the decomposition of CuC₂O₄·xH₂O or (Cu,Zn)C₂O₄, and the peak at about 343 °C was attributed to decomposition of β-ZnC₂O₄ and further decomposition of α-ZnC₂O₄ [13]. The weight loss peak for CuC₂O₄·xH₂O at 307 °C and weak peak for β-ZnC₂O₄ at 343 °C were observed from curve of EWP, which verified the XRD analysis of its precursor.

α-ZnC₂O₄·2H₂O → α-ZnC₂O₄ + 2H₂O t ≈ 118 °C

(6)

CuC₂O₄·xH₂O → CuO + CO₂ + H₂O t ≈ 308 °C

(7)

α-ZnC₂O₄ → ZnO + CO₂ t ≈ 343 °C

(8)

β-ZnC₂O₄ → ZnO + CO₂ t ≈ 343 °C

(9)

(Cu,Zn)C₂O₄ →(Cu,Zn)O + CO₂ t ≈ 296 °C

(10)

WMP and EMP prepared under microwave irradiation only included weight loss peak for (Cu,Zn)C₂O₄ or CuC₂O₄·xH₂O at about 300 °C, it can be speculated that microwave irradiation accelerated the isomorphous substitution between Cu²⁺ and Zn²⁺, the lesser Zn²⁺ was incorporated into CuC₂O₄, so there is little content of α-ZnC₂O₄·2H₂O or β-ZnC₂O₄ and there were no corresponding decomposition peaks. Especially, EMP had a large quantity of substitution and produced more (Cu,Zn)C₂O₄, the crystal phase trended to uniformity, therefore, the decomposition peak of EMP became narrower and the decomposition temperature decreased by 12 °C [23], the peak was mainly ascribe to (Cu,Zn)C₂O₄ and partial CuC₂O₄ unable to be substituted as equation (7) and (10).

3.3. SEM images of precursors and catalysts

SEM images of different precursors are shown in Figure 3, SEM images of different catalysts after calcining the corresponding precursors ar shown in Figure 4. The appearance of catalysts kept coherence with precursors in some extent by comparing Figure 3 and Figure 4. WWP presented as spherical particles, its diameter was 400 nm or so, it exhibited wafer-shaped and agglomerated each other after calcination, its particle size reduced to 300 nm about. EWC was consisted of irregular bar-shaped and block-shaped, 400 nm in size, it elucidated that ethanol solvent had effects on the shaping process of precursor in mother liquid. EWC didn’t change a lot on appearance or size compared with EWP, its size was about 300~400 nm, however, there were many minor porous channel in its superficial structure.

After introducing microwave irradiation, irregular uniwafer could be observed from morphology of WMP, its size was in the range of 300~500 nm. WMC was needle-like and flakiness owning to not be broke down completely, its diameter of section was about 50~100 nm; when ethanol was selected as solvent further, EMP mainly contained unidirectional ordered fibrous nanoparticles, its diameter of section was about 50 nm, which exhibited fine fibrous or flocculent and dispersed evenly after calcination [25].

After combining SEM photographies of precursors and catalysts, it can be concluded that precursors formed towards low dimensions under bulk heating effects of microwave. Precursors prepared by water bath exhibited spherical and massive, while precursors prepared by microwave irradiation presented single fibrous, which particle size was smaller than the former. The particle diameter of EMC was merely 50 nm, it dispersed well and the superficial activity sites distributed evenly, which benefited to improve the catalytic performance of MSR.

3.4. XRD characterization of catalysts

XRD patterns of different catalysts can be seen in Figure 5. The average grain size calculated by Scherrer formula at 2θ ≈ 35.5°, 38.7° and the texture parameters of different catalysts are listed in Table 2. There were no peaks assigned to Al₂O₃ in the four patterns, indicating that Al₂O₃ existed as amorphous or the content of Al₂O₃ was low. Diffraction peaks of CuO appeared in all the four catalysts at 2θ of 38.7°, CuO peaks of WWC were very sharp, the grain size of CuO was comparatively large, up to 18.8 nm as shown in Table 2. The CuO peaks of EWC became smoother when using ethanol as solvent, in which the ZnO peaks at 2θ＝34.4°, and 36.3° overlapped with CuO peak at 2θ＝35.5°. CuO grain size of EWC reduced to 12 nm, which indicated that viscous ethanol restricted the growth of precursor in mother liquid, resulting in a decrease in CuO grain size after calcination.

From patterns of WMC and EMC in Figure 5, it can be observed that the CuO peak becomes smoother, after introducing microwaves. Weak peaks of ZnO could still be seen in WMC, however, no peaks assigned to ZnO were detected in EMC. It indicated that isomorphous substitution took place in mother liquid when microwave irradiation was introduced, partial Cu²⁺ of CuC₂O₄ was incorporated into ZnC₂O₄ and almost all Zn²⁺ of ZnC₂O₄ was incorporated into CuC₂O₄, thereby, the original crystal structure were destroyed and the crystal degree declined after substitution. Cu-O-Zn solid solutions were obtained after calcining WMP and EMP, in which there existed strong synergy between CuO and ZnO. CuO crystal was surrounded averagely by ZnO crystal, which restricted the growth of CuO crystallites, so the grain size of CuO in WMC and EMC were merely 8.9 nm and 8.4 nm respectively [24].

As shown in Table 2, catalysts prepared under microwave irradiation possessed bigger surface area and pore volume than that prepared in water bath, surface area and pore volume of EMC were 77.2 m²/g and 0.36 cm³/g respectively. This is because there was much more (Cu,Zn)C₂O₄ in precursor prepared under microwave irradiation, the precursor were converted into (Cu,Zn)O (or Cu-O-Zn) solid solution after calcination, in which ZnO and CuO arrayed homogeneously and closely, restricting the agglomeration of CuO, therefore the catalysts were very fine and possessed large surface area. The mother liquid was sol-like after aging when ethanol was selected as solvent, the catalyst possessed large surface area after calcining corresponding precursor [26], so the pore volume of EWC and EMC was comparatively large as shown in Table 2.

3.5. H₂-TPR characterization of catalysts

Figure 6 shows H₂-TPR curves of different catalysts. Ethanol restricted the growth of precurs phase and favored forming fine catalysts which made the reduction process easy. It manifested that reduction temperature of catalysts prepared using ethanol as solvent was lower than that using water. The catalysts prepared by water bath heating were difficult to reduce, the reduction temperature of WWC and EWC was 240 °C approximately, while the reduction temperature of catalysts prepared by microwave irradiation was lower, it was 220 °C or so. It proved again that microwave irradiation selected the phase in the generation process of precursor, the catalyst arranged orderly in micro structure, which decreased the difficulty of reducing CuO by H₂.

As shown in Figure 6, one low temperature reduction peak and one high emerged in curves of WWC, EWC, and WMC, however, three peaks emerged in curve of EMC, the first two were divided to low temperature reduction peaks, the third was divided to high temperature reduction peak. Low temperature reduction peaks were assigned to reduction of scattered phase CuO (Peak I), while the other was assigned to reduction of bulk phase CuO (Peak II). The amount of isomorphous substitution was little in mother liquids aging heated by water bath, thus, its synergy between CuO and ZnO was weak and there mainly existed bulk phase CuO in corresponding catalyst, so their Peak II was larger. The amount of substitution between Cu²⁺ and Zn²⁺ was great when aging under microwave irradiation, they mainly generated (Cu,Zn)C₂O₄ in precursors which were calcined into Cu-O-Zn solid solutions. The catalysts were homogeneous and dispersive, the synergy was strong, so the Peak I was larger. Especially, the amount of substitution for EMC was great, the reduction peak assigned to high temperature was much smaller than others [27,28].

3.6. XPS and AES characterization of catalysts

XPS spectra of catalysts are shown in Figure 7. As shown in Figure 7(a), all catalysts showed a distinctive BE (binding energy) of Cu 2p_3/2 around 932~933 eV, companying a characteristic satellite peak between 940 ~ 945 eV due to the electron shakeup process, which indicated that Cu species were present as CuO. The BE of Cu 2p_3/2 around 932 ~ 933 eV in spectra of WWC, EWC, WMC and EMC were 932.0 eV, 932.05 eV, 932.10 eV and 932.95 eV respectively, increasing gradually. The BE of Zn2p_3/2 (Figure 7(b)) in corresponding spectra were 1021.45 eV, 1020.6 eV, 1020.35 eV and 1020.15 eV, decreasing gradually, which were lower than the BE (1022.2 eV) of pure ZnO [29].

It illustrated that isomorphous substitution took place when introducing microwave in aging process. The chemical environment and energy status of copper and zinc changed in the gotten Cu-O-Zn solid solution. Because the electronegativity of zinc was higher than copper, the outermost electron of copper shifted towards zinc, the electron density of copper decreased and its binding energy increased, while the electron density of zinc increased and its binding energy decreased [23].

Table 3 lists the data analysis based on XPS and AES spectra and the superficial elements concentration of different catalysts. EMC possessed higher superficial copper content and lower superficial zinc content, its ratio of X_Cu/X_Zn was up to 8.98, which exceeded all others, it can be verified by the intensity of spectra in Figure 7 and Figure 8. In Table 3, it also could be observed that ethanol solvent help to improve the content of superficial copper atoms. Besides, microwave irradiation raised the yield of (Cu,Zn)C₂O₄ in precursor, leading to lower KE (kinetic energy) of Cu LMM and higher KE of Zn LMM in final catalyst as shown in Figure 8. This was coincident with the research result of literature [15].

3.7. Catalytic performance test for MSR reaction

The catalytic performance of different catalysts for MSR is listed in Table 4. When comparing WWC and EWC, it can be concluded that ethanol solvent is conducive to reducing the CO selectivity of the catalyst. As can be seen from the XRD data, ethanol optimized precursor crystal phase composition, thereby reducing the CO selectivity. When comparing WWC and WMC, it can be concluded that microwave heating is conducive to improving the methanol conversion rate of the catalyst. As can be seen from the BET data and SEM images, microwave radiation increases the specific surface area of the catalyst, improves the dispersity of copper, increases more active sites, and thus improves the methanol conversion rate. Besides, as can be seen from the XRD data, microwave irradiation prompted the isomorphous substitution between cooper and zinc oxalate precursors, Cu-O-Zn solid solution formed after calcination, which possessed strong synergy and benefited the MSR reaction.

Thus, it illustrated that both ethanol and microwave irradiation benefited to improve the catalytic performance of MSR, the catalyst EMC prepared by ethanol solvent and microwave heating exhibited optimal catalytic performance, the conversion of methanol was 91.2%, the space time yield (STY) of H₂ reached 516.7 mL·g^-1·h^-1, its selectivity of CO was only 0.29%; In stark contrast, the catalyst WWC prepared by water solvent and water bath heating showed the worst catalytic performance, its conversion of methanol was only 53.6%, and the STY of H₂ was 300.0 mL·g^-1·h^-1, but its selectivity of CO was as high as 1.53%.

4. Conclusion

In the preparing process of CuO/ZnO/Al₂O₃ catalyst precursors, ethanol solvent selected the crystal phase composition in precursor and restricted its growth. Microwave irradiation promoted the isomorphous substitution between oxalate copper and oxalate zinc, the precursor mainly contain (Cu,Zn)C₂O₄ and exhibit nano-fibriform morphology. Either using ethanol as solvent or adopting microwave irradiation as aging heating manner benefited MSR reaction, the catalyst EMC prepared by ethanol solvent and microwave heating showed the optimal catalytic performance, the maximum STY_H2 reached 516.7 mL·g^-1·h^-1, the conversion of CH₃OH reached 91.2%, and the selectivity of CO was only 0.29%.