ZrO2 catalyst for enhancing water–gas-shift performance

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

The significant role of oxygen vacancy in Cu/ZrO2 catalyst for enhancing wateregas-shift performance Chongqi Chen, Chunxiao Ruan, Yingying Zhan, Xingyi Lin, Qi Zheng*, Kemei Wei National Engineering Research Center of Chemical Fertilizer Catalysts, Fuzhou University, Gongye Road 523, Fuzhou, Fujian 350002, PR China

article info

abstract

Article history:

Three Cu/ZrO2 catalysts were synthesized utilizing co-precipitation (CP), deposition

Received 8 August 2013

eprecipitation (DP) and depositionehydrothermal (DH) methods, respectively. The micro-

Received in revised form

structure and texture of those catalysts are characterized by means of XRD, SEM, N2-

30 September 2013

physisorption, Raman and EPR characterizations. It is demonstrated that different mor-

Accepted 14 October 2013

phologies and textures of ZrO2 are formed, and the micro- and crystal structure of Cu

Available online 16 November 2013

nanoparticles as well as the concentration of oxygen vacancies of ZrO2 are distinguish from

Keywords:

properties of the as-synthesized Cu/ZrO2 catalysts. It is found that the synergy interaction

each other. In addition, H2-TPR technique is employed to investigate the reducibility Cu/ZrO2

between CueZrO2 obtained by the DH method is the strongest, owning to the possession of

Water gas shift

the largest amount of oxygen vacancies. Furthermore, their catalytic activities with respect

Oxygen vacancy

to the water gas shift reaction are also performed, and the Cu/ZrO2-DH shows high cata-

Preparation method

lytic activity, the reasons are the well dispersion and small crystallite size of Cu, the largest amount of oxygen vacancies, as well as the strongest interaction between CueZrO2. Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

The three-pronged challenges for the 21st century are risking environment pollution, irreversible global climate change and uncertain oil supply [1]. Hydrogen is one of the promising energetic carriers for mobile and/or residential fuel cell, and possesses the merits of: (1) sustainable fuel resource, i.e. steam reforming of biomass and electrolysis of water; (2) much easier to be stored and transported compared with wind or solar energy; (3) no air pollutant and greenhouse gas emissions [2,3] during the buring of hydrogen. Therefore, widespread attentions have been focused on it, aiming to

reduce our dependence on fossil fuels and environment pollution [4,5]. As concerning with the hydrogen generation, a number of methods have been explored that can roughly be cataloged as: (1) hydrolysis of borohydride [6]; (2) reaction of Al with water [7]; (3) photocatalytic water-splitting [8]; (4) steam reforming of natural gas and/or biomass [9]. Among all these methods, steam reforming of natural gas, to date, is still the predominant way for hydrogen generation [10]. For the steam reforming of natural gas, water gas shift (WGS) reaction (CO þ H2O 4 CO2 þ H2 DH ¼ 41.1 kJ/mol) is implemented to enrich H2 concentration and eliminate CO. In the recent two decades, due to the keen

* Corresponding author. Tel.: þ86 591 8373 1234 8112; fax: þ86 591 8373 8808. E-mail address: [email protected] (Q. Zheng). 0360-3199/$ e see front matter Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijhydene.2013.10.074

318

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

demand of hydrogen for fuel cells, renewing interests have been focused on exploring promising catalysts that are suitable for the WGS reaction in fuel cell applications [11,12]. Both precious metal (e.g. Pt, Au or Ru) and non-precious metal (e.g. Cu or Ni) based catalysts have been widely examined. Various oxides are chosen as support, such as CeO2 [13], Al2O3 [14], TiO2 [15], ZrO2 [16], and their complex, like CeO2eZrO2 [17], CeO2eLa2O3 [18] and so on. Considering the high cost of the precious metals, more and more attentions have been focused on the non-precious ones. The Cu-based catalysts have found to be shown high WGS activity and stability in low and medium temperature (200  Ce300  C) [13,19], and have also been widely exployed. Besides, ZrO2 has also been widely used as a support for Cu-based catalysts in the field of methanol synthesis [20], steam reforming of methanol [21], etc. because it can activate and stabilize copper particles; meanwhile, the used of ZrO2 as a support for precious metal, like Pt [22] and Au [23], in the WGS reaction has also been well documented. Therefore, it is reasonable to consider that the ZrO2 supported Cu catalyst would show good catalytic performance for the WGS reaction. Cu/ZrO2 catalyst for the WGS reaction was first reported by Ko et al [24], it was found that the catalytic performance of Cu/ ZrO2 catalyst was depending on the loading amount of Cu. The influence of the crystalline structure of ZrO2 on the activity of Cu/ZrO2 catalyst in the WGS reaction was investigated by Agulia and co-workers [25], and it was concluded that the activity of copper supported on tetragonal ZrO2 is much higher than the one supported on monoclinic ZrO2. DFT calculation of Cu/ZrO2 catalyst was performed by Tang et al. [26] for the WGS reaction, and the catalytic active site was supposed to be in the vicinity of Cu/oxides interfaces and the ZrO2 support acts as a charge buffer to accept/release electrons from/to the Cu particle. It was reported that oxygen coordinatively unsaturated Zr sites, i.e. oxygen vacancy, is easy to be generated by calcination of ZrO2, especially in the reduction atmosphere [27], however, the function of oxygen vacancy in the Cu/ZrO2 catalyst for the WGS reaction has not been well studied. Herein, in an attempt to develop efficient binary Cu/ZrO2 catalysts with different amount of oxygen vacancies for the WGS reaction, a series of Cu/ZrO2 catalysts were fabricated by means of three different preparation methods including Coprecipitation (CP), Depositioneprecipitation (DP) and Depositionehydrothermal (DH) method, and their catalytic performances for WGS reaction were also investigated. The chemical and physical properties of those catalysts were characterized by XRD, SEM, Raman, N2-physisorption, EPR and TPR techniques, special attention is paid to the effect of preparation methods on the generation of oxygen vacancy for Cu/ZrO2 catalysts and the relationship between the physicochemical properties and catalytic performances.

2.

Experimental method

2.1.

Catalysts preparation

2.1.1.

Materials

Zirconium oxychloride (ZrOCl2$8H2O), Copper (II) nitrate (Cu(NO3)2$3H2O) and Potassium hydroxide (KOH) are all analytical grades and used without further purification.

2.1.2.

Synthesis

The Cu/ZrO2 catalysts with copper content of 25 wt% (calculated by CuO) were prepared by three different methods. For the co-precipitation method (CP), a certain amount of ZrOCl2$8H2O and Cu(NO3)2$3H2O were dissolved in 300 mL deionized water in a four-neck flask, and then KOH solution (0.5 mol/L) was added dropwise at 65  C under vigorous stirring until the pH ¼ 10, after that, the suspension was maintained at 65  C for another 1.5 h, washed with deionized water until pH value of supernatant is neutral. For the Depositioneprecipitation method (DP), the support was firstly synthesized by co-precipitation method, typically, 150 mL ZrOCl2$8H2O solution was first put in a four-neck flask, following by adding KOH aqueous solution (0.5 mol/L) dropwise at 65  C under vigorous stirring until the pH ¼ 10, subsequently by maintaining the suspension at 65  C for another 1.5 h, washing with deionized water until pH value of supernatant is neutral, dried at 120  C for 8 h, and then calcined at 350  C for 4 h; secondly, the CuO was deposited: 4.5 g of the above-synthesized ZrO2 was dispersed in 300 mL of deionized water, and a certain amount of Cu(NO3)2,3H2O was added, and then KOH aqueous solution (0.5 mol/L) was added dropwise at 65  C under vigorous stirring until the pH ¼ 10, after that, the suspension was maintained at 65  C for another 1.5 h, washed with deionized water until pH value of supernatant is neutral. For the Depositionehydrothermal method (DH), the support was prepared by hydrothermal method according to the literature protocols [16], in detail, 60 mL ZrOCl2$8H2O aqueous solution (0.4 mol/L) was transferred into a 100 mL Teflon-lined stainless steel autoclave and maintained at 150  C for 6 h, the resultant precipitate was then washed thoroughly with deionized water, dried at 120  C for 8 h and calcined at 350  C in static air for 4 h; afterwards, CuO was deposited same as the DP method. All the as-prepared precursors were further dried at 120  C for 12 h, and then calcined at 300  C for 2 h (heating rate was 5  C/min) purging with hydrogen. The as-obtained products are denoted as Cu/ZrO2-CP, Cu/ZrO2-DP and Cu/ ZrO2-DH, respectively.

2.2.

Characterizations

The powder X-ray diffraction (XRD) patterns of the samples were recorded by a PANalytical X0 Pert Pro diffractometer using Co Ka radiation (l ¼ 0.179 nm) at 40 kV and 40 mA. Their morphologies were investigated by Field Emission Scanning Electron Microscope (FE-SEM, Hitachi-S4800). Raman spectra were collected at room temperature on a Renishaw Invia Plus instrument using a semiconductor laser as an illumination source (532 nm). To determine the textural properties of the as-prepared samples, nitrogen adsorptionedesorption measurements were carried out at 77 K using a Micrometrics ASAP 2020 system after the sample was degassed at 200  C in a vacuum for 4 h. Temperature-programmed reduction (TPR) measurement was carried out on an AutoChem 2910 instrument. The H2TPR was performed by passing 10 vol.% H2/Ar (flowing rate ¼ 30 mL/min) on 50 mg catalyst at a heating rate of 10  C/min. Prior to the measurement, the samples were pretreated under Argon atmosphere at 200  C for 30 min, then the system was cooled to ambient temperature purging

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

with pure Argon gas. The hydrogen consumption was monitored using a thermal Conductivity Detector (TCD). Cudispersion was determined by selective N2O chemisorption method: in a typical procedure, a H2-TPR experiment was firstly carried out, after that the system was cooled down to room temperature purging with Ar, an oxidation procedure with 3.35 vol.% N2O/He gas mixture was executed at 50  C for 1 h (N2O þ 2 Cu ¼ Cu2O þ N2); finally, another H2-TPR experiment was performed. The dispersion of Cu sites was calculation by the Hydrogen Consumption (Chydrogen) of the twice H2-TPR experiment, X(Cu) % ¼ 2  Chydrogen’/ Chydrogen  100%, where the Chydrogen and Chydrogen’ are the Hydrogen Consumption of the first and second cycle of H2TPR experiment, respectively. Electron Paramagnetic Resonance (EPR) spectroscopy was recorded on a Bruker EMX A300 spectrometer at room temperature. The settings are as follows: center field is 3512.48 G, microwave frequency is 9.84 GHz and the power is 6.34 mW.

2.3.

Evaluation of catalytic performance

The catalytic activities of the catalysts for WGS reaction were tested in a fixed bed reactor at atmospheric pressure from 150  C to 350  C at an interval of 50  C. Typically, 1.5 mL of catalysts (20e40 mesh) was used and the space velocity was calculated to be 4000 h1; feed gas was a model reformates containing 10 vol. % CO, 60 vol. % H2, 12 vol. % CO2 and balance with N2; the ratio of vapor to feed gas was maintained at 1:1. The residual water of the outlet was removed by a condenser before entering a gas chromatograph equipped with a thermal conductivity detector (TCD). The activity was expressed by the conversion of CO, defined as: XCO (%) ¼ (1  V0 CO/VCO)/ (1 þ V0 CO)  100%, where VCO and V0 CO are the inlet and outlet content of CO, respectively.

3.

319

Results and discussion

3.1. Crystal structures and textures of the Cu/ZrO2 catalysts The morphologies of Cu/ZrO2 catalysts prepared by three different methods and reduced with H2 were examined by SEM technique, and the results are shown in Fig. 1aec. As clearly seen in Fig. 1, the morphologies of the three samples are so distinguish from each other. It is notable that the Cu/ ZrO2-CP sample obtained by co-precipitation method afforded the massive agglomerates of small particles; while in Fig. 1b, leaf-like crystals are observed, the size of each leaf-like structure is ranging from 50 to 150 nm in width and 100e500 nm in length; when the deposition-hydrothermal method was employed, the pre-synthesized ZrO2 is found to be disk-like and organized regularly. The morphology of the hydrothermally prepared ZrO2 is magnified and inserted in Fig. 1c, it can be seen that the disk-like ZrO2 is composed of a great deal of small globular nanoparticles with a large number of wormhole-like channels, as is supposed to be good for Cu distribution. Besides, the crystal structure of the three Cu/ZrO2 catalysts was investigated by XRD characterization, and the resulting patterns are shown in Fig. 1d. Two sets of diffraction patterns are found for all the three Cu/ZrO2 samples: those marked with “*” are indexed to ZrO2, while those labeled with “#” is ascribed to cubic metallic Cu (JCPDS file no. 01-089-2838). The diffraction peaks of ZrO2 for the Cu/ZrO2-DH are stronger and can be attributed to well crystalline monoclinic ZrO2 (JCPDS file no. 01-089-9066), while the other two are amorphous or compose of very fine crystals. Furthermore, the crystallite sizes of Cu are estimated from the half width of (111) reflection

Fig. 1 e Morphologies and crystal structures of Cu/ZrO2 catalysts prepared by different methods: (a) Cu/ZrO2-CP; (b) Cu/ZrO2DP; (c) Cu/ZrO2-DH; (d) XRD patterns.

320

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

Table 1 e Texture parameters of Cu/ZrO2 catalysts prepared by different methods. Samples a

m-ZrO2 Cu/ZrO2-CP Cu/ZrO2-DP Cu/ZrO2-DH a

BET surface area (m2/g)

Pore volume (cm3/g)

Cu crystallite size (nm)

Cu sites dispersion (%)

152 177 162 99

/ 0.125 0.139 0.122

/ 38.5 63.4 34.0

/ 39.0 17.7 28.0

The m-ZrO2 is fabricated by hydrothermal method, which is described in the section 2.1.2.

by DebyeeSherrer equation and summarized in Table 1. The calculation results showed that the Cu crystallite size of Cu/ ZrO2-DH is 34.0 nm, which is similar to the Cu/ZrO2-CP (38.5 nm), but the Cu crystallite size of Cu/ZrO2-DP is much larger. In addition, no remarkable shift of the diffraction peaks comparing with the pure Cu and ZrO2 reveals that lattice expansion and/or shrinkage of Cu/ZrO2 crystals should be neglected, i.e. no solid solution is formed during the synthesized process. Furthermore, the structural information of the as-prepared Cu/ZrO2 catalysts was studied by the Raman spectra, as is shown in Fig. 2. It can be seen that there are three main peaks located at 174, 465 and 622 cm1 in Fig. 2a and c, which are ascribed to the Raman-active modes for the monoclinic phase of ZrO2 [28], indicating that the ZrO2 in Cu/ZrO2-CP is fine crystals with monoclinic phase. Whereas, in Fig. 2b, a broad band centered at 523 cm1 can be observed, indicating the appearance of amorphous ZrO2 for the Cu/ZrO2-DP sample. Besides, an intense band at 273 cm1 with a shoulder at 323 cm1 and a broad band at 1086 cm1 are corresponding to the copper oxides species. In comparison with the Raman vibrational spectra of a CuO single crystal [29], the Raman peaks in those as-prepared Cu–ZrO2 catalysts are broadened and downshifted; the peak at 273 cm1 can be assigned to the Ag mode, and the peaks at 323 cm1 to the Bg modes [30]. According to the previous report [31], the broadenings and downshifts of the Raman peaks are mainly attributed to the

quantum confinement effect of nano-sized CuO, i.e. the red shift and broadening of the Raman spectra increase with decrease in grain size. As is consistent with the Cu crystal size of reduced Cu/ZrO2 catalyst, i.e. the Cu/ZrO2-DP shows the largest Cu crystal size, while the Cu/ZrO2-CP and Cu/ZrO2-DH have the smaller Cu particle sizes. It is well known that the better the dispersion of Cu is, the higher catalytic performance of supported Cu catalyst for WGS will be, therefore, the dispersion of Cu sites on the surface of ZrO2 is evaluated, and the results are presented in Table 1. The Cu dispersion values are 39.0, 17.7 and 28.0% for Cu/ZrO2-CP, Cu/ZrO2-DP and Cu/ZrO2-DH, respectively. The reason must be originated from distinguish physical and chemical properties of the ZrO2-supports, thus resulting in different strength of interaction between CuO and ZrO2. As indicates by the above SEM, XRD results, it can be concluded that: (1) the fine ZrO2 crystals derived from co-precipitation method is favor for the CuO dispersion; (2) when concerning the depositioneprecipitation and depositionehydrothermal method, the ZrO2 support prepared by hydrothermal method, which possesses the uniform and fine ZrO2 crystals, leads to better CuO dispersion. The texture of the as-synthesized sample was characterized by N2-physisorption experiment and the corresponding texture parameters are listed in Table 1. One can see that the BET specific surface areas for the three samples follow the sequence: Cu/ZrO2-CP > Cu/ZrO2-DP > Cu/ZrO2-DH. The reason must be that the Cu/ZrO2-CP is composed of fine ZrO2 crystals and amorphous ZrO2 for Cu/ZrO2-DP after calcination, while the Cu/ZrO2-DH is well crystallized, as is in accordance to the above SEM, XRD and Raman characterizations. The largest pore volume is found for the Cu/ZrO2-DP sample, as is ascribed to larger grain size of those leaf-like structures.

3.2.

Fig. 2 e Raman spectra of the as-prepared catalysts: (a) Cu/ ZrO2-CP; (b) Cu/ZrO2-DP; (c) Cu/ZrO2-DH.

Microstructures of the Cu/ZrO2 catalysts

Fig. 3 shows the EPR spectra of the freshly prepared Cu/ZrO2 samples without calcination. Spectral analysis of the different peaks allows us to discern the contribution of different signals: all the spectra can be interpreted in terms of at least three overlapping components. The most prominent signals present a line shape characteristic of species with an axial (or near axial) symmetry with g//(2.38) > gt (2.04) > ge (2.02) and fourline hyperfine splitting due to the interaction of the unpaired electrons with nuclei of spin I ¼ 3/2 (signal type A) [32]. Type A signals are attributed without ambiguity to the isolated Cu2þ ions in a distorted octahedral [33,34]. Another one (signal A0 ) with g ¼ 2.08 is existed in all the three spectra, it was reported that the signals A0 and A were closely correlated with the dehydration, which can be attributed to Cu2þ ions before and

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

321

Fig. 3 e EPR spectra of Cu/ZrO2 catalysts prepared by different methods: (a) Cu/ZrO2-CP; (b) Cu/ZrO2-DP; (c) Cu/ZrO2-DH. Right panel is the magnification of the region rectangle by dash line from the left panel.

after dehydration [35], it was confirmed by that after pretreatment of all the three freshly prepared catalyst at 300  C for 2 h with H2, the signals A0 disappeared (not shown). The other one (signal B) with g ¼ 2.23 consists of a single line shape can be assigned to the magnetic interaction of two or more adjacent Cu2þ ions. When carefully examination the spectrum Fig. 3a, it can be found that there is a widely broaden peak (signal O), which should be attributed to Cu2þ ions in the presence of strong dipolar interactions [35], and its larger line width (and consequently its unresolved hyperfine splitting) might be attributed to dipolar broadening effects produced by mutual interactions between paramagnetic Cu2þ ions. It is to say, the formation of small well dispersed particles of CuO in ZrO2 solid could be responsible for the O signal; when the CuO species transform into large aggregates, the O signal may disappears, as is in accordance with the dispersion of the Cu sites results by utilizing N2O titration method. Owing to the intense signals of Cu2þ are collected for the as-prepared Cu/ZrO2 catalysts, no signals for Zr3þ can be observed in the EPR spectra. The distinction of various signals in EPR spectra is thought to be closely related to the microstructure of catalyst. In Fig. 3, the different g factor values of signal A reflect differences of the coordination environment for the Cu2þ ions, i.e. the

location of Cu2þ ions. It was reported that shifts of g// to lower values and A// to higher values result in increasing of the crystal lattice distortion, from a pure tetragonally distorted octahedral coordination to square pyramidal and, further, to square planar coordinations [32,36]. It can be carefully seen in the right panel of Fig. 3, from sample a to c, the g//values of signal A shift to a little bit lower gradually, suggesting that the crystal lattice distortion increases likewise. The isolated Cu2þ ions in a distorted octahedral environment can be assigned to hexa-coordination, thus the crystal lattice distortion of the asprepared catalysts leads to the coordination status of Cu2þ ions from some degree of coordinative unsaturation (sample c) to saturation (sample a) [32]. Herein, in order to keep electric neutrality of the as-synthesized samples, coordinative unsaturation must result in oxygen deficit, i.e. oxygen vacancy, which is thought to be benefit for the Cu-based catalyst in the WGS reaction.

3.3. Investigation on the reducibility of the Cu/ZrO2 catalyst In order to investigate the reducibility of as-prepared Cu/ ZrO2 catalysts, TPR technique is employed and the typical

Fig. 4 e H2-TPR profiles of the first cycle (a) and second cycle after oxidation (b) for those Cu/ZrO2 catalysts prepared by different methods.

322

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

H2-TPR profiles are shown in Fig. 4. Fig. 4a shows the H2TPR profiles of those freshly prepared catalysts, while Fig. 4b is the H2-TPR results from those catalysts firstly reduced with 10 vol.% H2/Ar at 300  C for 2 h following by re-oxidation with 3 vol.% O2/He at 300  C for 2 h. In literature, various Cu/ZrO2 catalysts have been investigated by TPR characterization, and diverse results were acquired. Five peaks at 431 K (peak a1), 452 K (peak a2), 470 K (peak b1), 507 K (peak b2) and 579 K (peak g) were obtained by Ma et al. [35] during the reduction of Cu/m-ZrO2 catalyst. The peaks a1 and a2 were related to the surface oxygen vacancies of monoclinic zirconia with special geometry structure; the two b peaks were ascribed to the reduction of highly dispersion copper species with different environment or interaction with monoclinic zirconia support; and the g peak could be contributed to the reduction of bulk copper oxide. A double peak around 200  C and an additional peak at high temperature were observed for the copper catalysts supported on monoclinic ZrO2 by Aguila et al. [25], who attributed the double peak to the sequential reduction of highly dispersed Cu2þ species to Cu1þ and Cu0 and the other one to the reduction of CuO bulk. In Fig. 4b, three reduction peaks can be observed: the lower one located at about 150  C can be decompounded into two peaks (a1 and a2) with similar peak area, as can be assigned to the reduction of CuO with small crystallite size and well dispersed on the surface of ZrO2 supports, from Cu2þ species to Cu1þ and Cu0, respectively [25], and the peak areas of peak a for all the three samples are in good consistent with the results of N2O titration; The high temperature reduction peak (peak b) is incontestable attributed to the reduction of moderate copper oxide (crystalline) interacted with surface oxygen vacancies of ZrO2; as for peak g, it could be attributed to the reduction of copper oxide incorporated into the ZrO2 lattice and the large isolated copper oxide phase (crystalline). It is well known that the variation of reduction temperature of bulk oxide should be related to its particle size and/or promotion effect of another species that interacts with it. The reduction of pure CuO is featured by a single reduction peak at a considerably higher temperature of 320  C. It is thus concluded that there is a CueZrO2 interaction which facilitates the reduction of the supported copper species in our cases. The lower the reduction temperature is, the stronger the interaction between Cu and ZrO2 will be. As seen in Fig. 4b, the Cu/ZrO2-DH catalyst shows the strongest interaction between Cu and ZrO2. Most important of all, in Fig. 4b, the fraction of moderate copper oxide (peak b), which is consider to be has a greatly positive effect on the WGS catalytic activity over supported CuO catalyst, is observed to be intense and the peaks position are ranked as: Cu/ZrO2-DH < Cu/ZrO2-DP < Cu/ZrO2-CP, indicating that the interaction between Cu and ZrO2 also follows the sequence. When correlating to the EPR results, as may be resulting from different amount of oxygen vacancies of Cu/ZrO2 catalysts prepared by different methods, i.e. the existence of oxygen vacancy results in the strong interaction between Cu and ZrO2, thus lowing the reduction temperature of peak b.

3.4.

Catalytic performances of the Cu/ZrO2 catalysts

Fig. 5 shows the CO conversion vs. temperature curves of the Cu/ZrO2 catalysts prepared by different methods. One can be seen that the catalytic performances of the as-synthesized catalysts behave differently on the basis of the prepared methods, and the reactivity follows the order: Cu/ZrO2DP < Cu/ZrO2-CP < Cu/ZrO2-DH. The one prepared by DH method shows the high activity towards WGS reaction, while the other two catalysts display similar activity during the whole testing temperature range. For a particular state of the catalysts, the CO conversion of Cu/ZrO2-DH sample is higher than that of Cu/CeO2, which is supposed to be an active catalyst among the most active WGS systems [37], suggesting that the monoclinic ZrO2 supported Cu catalyst is also a promising candidate for the WGS reaction other than Cu/CeO2 catalyst. Cu/ZrO2 was also prepared by Aguila et al. [25] with respect to the WGS reaction, it was reported that Cu dispersed on the surface of tetragonal ZrO2, showing superior catalytic activity for the WGS reaction, while the one supported on monoclinic ZrO2 showed almost no CO conversion. It should be noticed that the monoclinic ZrO2 in the literature is obtained by calcination of hydrous zirconium at 700  C for 3 h with a BET surface area of 36 m2/g, comparing with 107 m2/g for tetragonal ZrO2 obtained by refluxing of hydrous zirconium with NH4OH; the dispersions of Cu on the surface of the two ZrO2 are also distinct evidently from each other, as can be concluded from the TPR results. In the present work, a monoclinic ZrO2 (m-ZrO2) with a BET surface area of 152 m2/g was fabricated by hydrothermal method. Subsequently, after deposition of Cu nanoparticles, the Cu/m-ZrO2 catalyst shows excellent catalytic behavior towards the WGS reaction. Therefore, it is hard to draw the conclusion that which crystal structure of ZrO2 is better for supported Cu with respect to the WGS reaction, the dispersion of copper, chemical and physical structure of ZrO2 support as well as interaction between

Fig. 5 e Catalytic actives of Cu/ZrO2 catalysts prepared by different methods over the WGS reaction: (a) Cu/ZrO2-CP; (b) Cu/ZrO2-DP; (c) Cu/ZrO2-DH; (d) Cu/CeO2 [37].

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

copper and ZrO2 are supposed to be integrally influencing the catalytic performance of copper based catalysts during the WGS reaction. Cu has been found to possess both the abilities of CO adsorption and H2O dissociation, and it is known that pure Cu alone is active for the WGS reaction, although the catalytic activity measured is not high. Several DFT calculations confirmed that the WGS reaction can indeed occur on clean Cu (100) and Cu (111) surfaces [38,39]. In line with this, most of the researchers proposed that metallic Cu is active center for the Cu/oxide composite systems and characterization results also claimed that metallic Cu is the only stable species in those systems under reaction conditions [40,41]. On the basis of the above characterization results, it can be found that after the reduction with hydrogen, all the three as-prepared catalysts show only metallic Cu, no Cu2O or CuO are detected, in the XRD patterns, indicating the metallic Cu is the active site in our cases. It is clear that the Cu sites dispersion follows the sequence: Cu/ZrO2CP > Cu/ZrO2-DH > Cu/ZrO2-DP, and the crystal sizes of Cu crystals are Cu/ZrO2-CP z Cu/ZrO2-DH < Cu/ZrO2-DP. It is well known that the better the Cu sites dispersion and the smaller crystal sizes of Cu crystals are, the higher catalytic activity of Cu catalyst will be. Moreover, the Cu/ZrO2-CP one possesses the largest BET surface area, which is regarded as an ability of adsorbing the most reactant. Thus, based on N2O titration, XRD and N2-physisorption characterization results, the Cu/ZrO2-CP may show the best catalytic performance towards the WGS reaction. However, in actual, the activities of the as-prepared catalysts follow the order: Cu/ZrO2-DP < Cu/ZrO2-CP < Cu/ ZrO2-DH, although the Cu/ZrO2-DH does not show the best Cu sites dispersion, and the largest BET surface area. The reasons will be clarified whereafter. To date, the reaction mechanism of WGS still in dispute as either simply redox mechanism or involving formate intermediates mechanism [25]. Recently, another mechanism so called carboxyl-mediated mechanism [26] is supposed, for which the formate is thought to be just a spectator during the WGS reaction process [42]. In both the redox and carboxyl mechanisms, the activation of H2O, leading to the formation of species responsible for oxidizing CO, is a kinetically important step [43]. Hereon, the oxygen vacancy can play a direct role in the dissociation of H2O [44,45]. Most important of all, the dissociation of H2O process is thought to be the ratedetermining step during the WGS reaction. It was reported that the oxygen vacancy sites on the support may activate water and release H2 (gas) by direct water decomposition or following the water dissociation into OHad and Had [45]. In our cases, the route of water dissociation through the oxygen vacancy sites is still confused, so more detail about the effect of oxygen vacancy sites during the WGS reaction is being carried out. Based on the above characterizations, it can be concluded from the EPR and H2-TPR results that the Cu/ZrO2DH possesses more oxygen vacancies than the other two. Thus, Cu/ZrO2-DH sample is supposed to be the most efficient in the activation of H2O, resulting in showing the best catalytic performance of WGS reaction. Furthermore, the largest amount of moderate copper oxide also greatly promotes its catalytic performance. Therefore, under the combined action of crystal size of Cu sites, dispersion, the amount of moderate copper oxide and the present of oxygen vacancies, the Cu/

323

ZrO2-DH catalyst shows the best catalytic performance with respect to the WGS reaction.

4.

Conclusions

Three Cu/ZrO2 catalysts were successfully synthesized utilizing co-precipitation, depositioneprecipitation and depositionehydrothermal methods. It is demonstrated that different morphologies and textures of ZrO2 supports were formed; the interaction between Cu and ZrO2, the crystallite sizes of and dispersion of Cu nanoparticles, as well as the concentration of oxygen vacancies in ZrO2 are distinguish from each other. It is concluded that the DH method is ideal to fabricate Cu/ZrO2 catalyst with monoclinic ZrO2, small particle sizes of Cu crystals, large amount of oxygen vacancies, thus showing stronger synergy interaction between CueZrO2 and higher catalytic activity for the WGS reaction compared to the other two.

Acknowledgments The authors acknowledge the financial support from the Department of Science of the People’s Republic of China (20771025) and the Technology Development Program of Fuzhou University (2011-XY-5).

references

[1] Andrews J, Shabani B. Re-envisioning the role of hydrogen in a sustainable energy economy. Int J Hydrogen Energy 2012;37:1184e203. [2] Chen WH, Hsieh TC, Jiang TL. An experimental study on carbon monoxide conversion and hydrogen generation from water gas shift reaction. Energy Convers Manage 2008;49:2801e8. [3] Urbaniec K, Friedl A, Huisingh D, Claassen P. Hydrogen for a sustainable global economy. J Clean Prod 2010;18(Suppl. 1):S1e3. [4] Aki H, Taniguchi Y, Tamura I, Kegasa A, Hayakawa H, Ishikawa Y, et al. Fuel cells and energy networks of electricity, heat, and hydrogen: a demonstration in hydrogen-fueled apartments. Int J Hydrogen Energy 2012;37:1204e13. [5] Hwang JJ. Review on development and demonstration of hydrogen fuel cell scooters. Renew Sustain Energy Rev 2012;16:3803e15. [6] Hu L, Ceccato R, Raj R. Ultrahigh figure-of-merit for hydrogen generation from sodium borohydride using ternary metal catalysts. J Power Sources 2011;196:69e75. [7] Gai WZ, Liu WH, Deng ZY, Zhou JG. Reaction of Al powder with water for hydrogen generation under ambient condition. Int J Hydrogen Energy 2012;37:13132e40. [8] Chou L, Liu R, He W, Geh N, Lin Y, Hou EYF, et al. Direct oxygen and hydrogen production by photo water splitting using a robust bioinspired manganese-oxo oligomer complex/tungsten oxide catalytic system. Int J Hydrogen Energy 2012;37:8889e96. [9] Guo Y, Liu X, Azmat MU, Xu W, Ren J, Wang Y, et al. Hydrogen production by aqueous-phase reforming of

324

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 3 1 7 e3 2 4

glycerol over Ni-B catalysts. Int J Hydrogen Energy 2012;37:227e34. Izquierdo U, Barrio VL, Cambra JF, Requies J, Gu¨emez MB, Arias PL, et al. Hydrogen production from methane and natural gas steam reforming in conventional and microreactor reaction systems. Int J Hydrogen Energy 2012;37:7026e33. Tabakova T, Idakiev V, Avgouropoulos G, Papavasiliou J, Manzoli M, Boccuzzi F, et al. Highly active copper catalyst for low-temperature wateregas shift reaction prepared via a CueMn spinel oxide precursor. Appl Catal A Gen 2013;451:184e91. Zhai Y, Pierre D, Si R, Deng W, Ferrin P, Nilekar AU, et al. Alkali-stabilized PteOHx species catalyze low-temperature wateregas shift reactions. Science 2010;329:1633e6. Si R, Raitano J, Yi N, Zhang L, Chan S-W, FlytzaniStephanopoulos M. Structure sensitivity of the lowtemperature wateregas shift reaction on CueCeO2 catalysts. Catal Today 2012;180:68e80. Shekhar M, Wang J, Lee WS, Williams WD, Kim SM, Stach EA, et al. Size and support effects for the wateregas shift catalysis over gold nanoparticles supported on model Al2O3 and TiO2. J Am Chem Soc 2012;134:4700e8. Kim SS, Lee HH, Hong SC. The effect of the morphological characteristics of TiO2 supports on the reverse wateregas shift reaction over Pt/TiO2 catalysts. Appl Catal B: Environ 2012;119e120:100e8. Zhang Y, Zhan Y, Chen C, Cao Y, Lin X, Zheng Q. Highly efficient Au/ZrO2 catalysts for low-temperature wateregas shift reaction: effect of pre-calcination temperature of ZrO2. Int J Hydrogen Energy 2012;37:12292e300. Pilasombat R, Daly H, Goguet A, Breen JP, Burch R, Hardacre C, et al. Investigation of the effect of the preparation method on the activity and stability of Au/ CeZrO4 catalysts for the low temperature water gas shift reaction. Catal Today 2012;180:131e8. Hla SS, Morpeth LD, Sun Y, Duffy GJ, Ilyushechkin AY, Roberts DG, et al. A CeO2eLa2O3-based Cu catalyst for the processing of coal-derived syngases via high-temperature wateregas shift reaction. Fuel 2013;114:178e86. Ruan CX, Chen CQ, Zhang YJ, Lin XY, Zhan YY, Zheng Q. Cu/ ZrO2 catalyst for low-temperature wateregas shift reaction: preparation, characterization and performance. Chinese J Catal 2012;33:842e9. Liu J, Shi J, He D, Zhang Q, Wu X, Liang Y, et al. Surface active structure of ultra-fine Cu/ZrO2 catalysts used for the CO2 þ H2 to methanol reaction. Appl Catal A Gen 2001;218:113e9. Esposito S, Turco M, Bagnasco G, Cammarano C, Pernice P, Aronne A. Highly dispersed solegel synthesized CueZrO2 materials as catalysts for oxidative steam reforming of methanol. Appl Catal A Gen 2010;372:48e57. Graf PO, de Vlieger DJM, Mojet BL, Lefferts L. New insights in reactivity of hydroxyl groups in water gas shift reaction on Pt/ZrO2. J Catal 2009;262:181e7. Zhang Y, Zhan Y, Cao Y, Chen C, Lin X, Zheng Q. Lowtemperature wateregas shift reaction over Au/ZrO2 catalysts using hydrothermally synthesized zirconia as supports. Chinese J Catal 2012;33:230e6. Ko JB, Bae CM, Jung YS, Kim DH. CueZrO2 catalysts for wateregas-shift reaction at low temperatures. Catal Lett 2005;105:157e61. Aguila G, Guerrero S, Araya P. Influence of the crystalline structure of ZrO2 on the activity of Cu/ZrO2 catalysts on the water gas shift reaction. Catal Commun 2008;9:2550e4.

[26] Tang QL, Liu ZP. Identification of the active Cu phase in the wateregas shift reaction overCu/ZrO2 from first principles. J Phys Chem C 2010;114:8423e30. [27] Zhao Q, Wang X, Cai T. The study of surface properties of ZrO2. Appl Surf Sci 2004;225:7e13. [28] Li M, Feng Z, Xiong G, Ying P, Xin Q, Li C. Phase transformation in the surface region of zirconia detected by UV Raman spectroscopy. J Phys Chem B 2001;105:8107e11. [29] Irwin JC, Chrzanowski J, Wei T, Lockwood DJ, Wold A. Raman scattering from single crystals of cupric oxide. Phys C Supercond 1990;166:456e64. [30] Wang W, Liu Z, Liu Y, Xu C, Zheng C, Wang G. A simple wetchemical synthesis and characterization of CuO nanorods. Appl Phys A 2003;76:417e20. [31] Xu JF, Ji W, Shen ZX, Li WS, Tang SH, Ye XR, et al. Raman spectra of CuO nanocrystals. J Raman Spectrosc 1999;30:413e5. [32] Martinez-Arias A, Fernandez-Garcia M, Soria J, Conesa JC. Spectroscopic study of a Cu/CeO2 catalyst subjected to redox treatments in carbon monoxide and oxygen. J Catal 1999;182:367e77. [33] Liu Z, Ji W, Dong L, Chen Y. Electron spin resonance studies of CuO supported on tetragonal ZrO2. J Catal 1997;172:243e6. [34] Liu Z, Amiridis MD, Chen Y. Characterization of CuO supported on tetragonal ZrO2 catalysts for N2O decomposition to N2. J Phys Chem B 2005;109:1251e5. [35] Aboukaı¨s A, Bennani A, Lamonier-Dulongpont C, Abi-Aad E, Wrobel G. Redox behaviour of copper(II) species on CuCe oxide catalysts: electron paramagnetic resonance (EPR) study. Colloid Surf A 1996;115:171e7. [36] Kucherov AV, Slinkin AA, Kondrat’ev DA, Bondarenko TN, Rubinstein AM, Minachev KM. Cu2þ-cation location and reactivity in mordenite and ZSM-5: e.s.r.-study. Zeolites 1985;5:320e4. [37] She Y, Zheng Q, Li L, Zhan Y, Chen C, Zheng Y, et al. Rare earth oxide modified CuO/CeO2 catalysts for the wateregas shift reaction. Int J Hydrogen Energy 2009;34:8929e36. [38] Gokhale AA, Dumesic JA, Mavrikakis M. On the mechanism of low-temperature water gas shift reaction on copper. J Am Chem Soc 2008;130:1402e14. [39] Tang QL, Chen ZX, He X. A theoretical study of the water gas shift reaction mechanism on Cu (1 1 1) model system. Surf Sci 2009;603:2138e44. [40] Li L, Zhan Y, Zheng Q, Zheng Y, Lin X, Li D, et al. Wateregas shift reaction over aluminum promoted Cu/CeO2: nanocatalysts characterized by XRD, BET, TPR and Cyclic Voltammetry (CV). Catal Lett 2007;118:91e7. [41] Rodriguez JA, Hanson JC, Wen W, Wang X, Brito JL, Martı´nezArias A, et al. In-situ characterization of wateregas shift catalysts using time-resolved X-ray diffraction. Catal Today 2009;145:188e94. [42] Meunier FC, Reid D, Goguet A, Shekhtman S, Hardacre C, Burch R, et al. Quantitative analysis of the reactivity of formate species seen by DRIFTS over a Au/Ce(La)O2 wateregas shift catalyst: first unambiguous evidence of the minority role of formates as reaction intermediates. J Catal 2007;247:277e87. [43 Campbell CT. Finding the rate-determining step in a mechanism: comparing dedonder relations with the “degree of rate control”. J Catal 2001;204:520e4. [44] Henderson MA. The interaction of water with solid surfaces: fundamental aspects revisited. Surf Sci Rep 2002;46:1e308. ´ S, Sanz JF. [45] Oviedo J, Sa´nchez-de-Armas R, Miguel MA Methanol and water dissociation on TiO2 (110): the role of surface oxygen. J Phys Chem C 2008;112:17737e40.