Boosting the activity of a Au/CeO2/Al2O3 catalyst for the WGS reaction

1 Boosting the activity of a Au/CeO2/Al2O3 catalyst for the WGS reaction. T.R. Reina*, S. Ivanova, M.A. Centeno and J.A. Odriozola Departamento de Química Inorgánica, Universidad de Sevilla e Instituto de Ciencias de Materiales de Sevilla Centro mixto US-CSIC Avda. Américo Vespucio 49, 41092 Seville, Spain (*) corresponding author: [email protected] Abstract Herein a strategy to design highly efficient Au/CeO2/Al2O3 based WGS catalysts is proposed. The inclusion of transition metals, namely Fe, Cu and Zn as CeO2 dopant is considered. All the promoters successfully increased the WGS performance of the undoped sample. The activity improvement can be correlated to structural and/or redox features induced by the dopants. The comparative characterization of the doped samples by means of XRD, Raman spectroscopy and OSC evaluation permits an accurate understanding of the boosted WGS activity arising from the Ce-promoter interaction. This study establishes distinction among both, structural and redox sources of promotion and provides a useful strategy to develop highly active Au/CeO2 based catalysts for the WGS reaction. Keywords: WGS, Gold catalysts, cerium oxide, dopants, 2 1. Introduction The expected depletion of oil-based energy sources and their associated problems as for example serious pollution and global warming, among others, makes mandatory the investigation on alternative energies more environmentally respectful and with longer lifetime [1]. The H2-fuel cell technology is recognized as real approach to replace internal combustion engines in vehicles. These devices, use the chemical energy of H2 to efficiently produce electricity generating only water as a by-product [2,3]. The production of hydrogen for fuel cell applications requires several processing steps, including CO removal by means of the water gas shift (WGS) and the selective CO oxidation and/or methanation since the presence of carbon monoxide inactivates the fuel cell anode [4]. The WGS reaction is the most relevant CO elimination process in terms of CO amount to face (4-10 % vol.) in comparison to the finals steps of the hydrogen purification (few ppm - 1 % vol.). Moreover, the WGS reaction not only eliminates CO but produces an extra amount of H2 in the stream improving the efficiency of the process. Therefore the development of efficient catalysts for the WGS reaction is of crucial importance to guarantee the success of the hydrogen technology. Traditionally, the industrial catalysts for the low temperature WGS are based on Cu-ZnO-Al2O3 mixtures [5-7]. Nevertheless these systems present several drawbacks including its pyrophoricity and obligatorily activation steps [8]. In order to overcome these problems, noble metal based catalysts are under intensive investigation nowadays [6, 9-11]. Pt and Pd based catalysts have been reported as very promising systems for this process, nevertheless they still required activation either by previous reduction or under the WGS stream itself [12-14]. On the other hand, supported gold based catalysts do not present such pre-activation limitation and can be used as prepared. However there is a critical point regarding gold catalysts and the shift reaction: 3 contrary to platinum, gold cannot carry out the water activation and requires the indispensable assistance of the support to complete the catalytic cycle [15]. Therefore the choice of the support gains relevancy in the WGS. We have recently reviewed the literature involving gold based materials for the water gas shift in the last 20 years [16]. Despite the diversity of supports employed, cerium oxide outstands as the most convenient option. The suitability of CeO2 for this process is related to its high oxygen mobility and storage capacity [17,18]. In addition, at medium temperatures and reductive atmospheres, such as the WGS conditions, oxygen vacancies can be created on ceria lattice. Very frequently these punctual defects have been correlated with the catalytic activity in the WGS since they act as water activation sites, being vital to initiate the reaction [19, 20]. Afterwards, the activated water and CO interacts at the Au-CeO2 interface that plays also a major role in the process [21] Classically gold nanoparticles are dispersed on bulk ceria resulting in very efficient Au/CeO2 materials for the WGS [22-25]. As an alternative Au/CeO2/Al2O3 catalysts can be proposed. This approach provides several advantages compared to the traditional Au/CeO2 configuration: (i) reduced amount of ceria in the catalysts formulation (ii) greater Au-CeO2 interface (iii) enhanced oxygen mobility, and iv) possibility to potentiate the later by the addition of ceria promoters [26]. These dopants may affect the structural and electronic properties of ceria resulting in a beneficial effect on the WGS activity [27, 28]. With this premises the use of 3 late transition metals Fe, Cu and Zn as ceria promoters of the Au/CeO2/Al2O3 catalyst is evaluated in this paper. An accurate comparative study of the Au/CeO2-MOx/Al2O3 catalyst’s structural and redox modifications produced by the added metals is then correlated to the observed catalytic behavior. 4 2. Experimental 2.1. Catalyst preparation Support preparation The supports were synthesized by a conventional co-precipitation method. The necessary amounts of metal nitrate precursor (cerium, copper, iron or zinc nitrate - Sigma-Aldrich) were impregnated on γ-alumina powder (Sasol). The impregnation was carried out in 50 mL of ethanol, evaporated in rotary vapour till dry solid obtaining at reduced pressure and 50 ºC. The solid was then treated with NH3 solution (10 mol L-1) during 30 min in order to assure the full conversion of the nitrates to hydroxides, except for the Cu where an important loss of the metals as amino complexes should be avoided. The support was then filtered, dried and calcined at 500 ºC during 4 hours. The intended composition of the samples was 15 wt. % of Ce-M mixed oxide (M= Zn, Cu, Fe,) with 2 wt. % of doping metal oxide. According to previous results of our group, 2 wt.% of metal oxide is the optimum amount of dopant for this kind of systems [26, 28] Gold deposition The gold was deposited by the direct anionic exchange method (DAE), assisted by NH3 as described elsewhere [29]. A 10-4 mol L-1 aqueous solution of the gold precursor HAuCl4 (Alfa Aesar) was used in order to obtain a final Au loading of 2 wt. %. The support was sieved and the 100-200 μm fractions retained. After Au deposition, the solid was dried at 100 °C overnight and calcined in air at 350°C for 4 h. To simplify the nomenclature, oxygen is omitted from the labels. For example, the Au/CeCu/Al solid theoretically contains 2 wt.% of Au and 15 wt.% CeO2-CuO mixed oxide dispersed on alumina. 5 2.2. Catalyst characterization X-ray microfluorescence spectrometry (XRMF) was used to determine the chemical compositions and the analysis was performed in EDAX Eagle III spectrophotometer with Rh source of radiation. The specific surface area was determined by N2 adsorption-desorption measurements at liquid nitrogen temperature. The experiments were carried out on a Micrometrics Tristar II instrument. Before the analysis, the samples were degassed for 2h at 250ºC in vacuum. BET equation was applied to obtain the specific surface area of the studied samples. X-ray diffraction (XRD) analysis was carried out on X`Pert Pro PANalytical instrument. Diffraction patterns were recorded using Cu Kα radiation (40 mA, 45kV) over a 2θ-range of 10 to 80º and using a step size of 0.05º and a step time of 240 s. The Raman spectra were recorded in a dispersive Horiva Jobin Yvon LabRam HR800 spectrometer, with a 20 mW He–Ne green laser (532.14 nm) operating 5 mW. The microscope used a 50x objective and a confocal pinhole of 1000 µm. For the oxygen storage capacity (OSC) measurements 100 mg of catalyst were loaded and activated into a U-shaped quartz reactor at 350 ºC during 1 hour in 5% O2/ He flow (50 mL/min). Then, the system is cooled and set to the desired temperature (150, 250 and 350 ºC). For each temperature, ten O2 pulses of 1 mL were injected every 2 min. The sample is then subjected after 10 min He degasification to four alternating series of pulses (CO–O2–CO–O2–CO–O2-CO–O2). The OSC is determined by the average amount of CO2 per pulse formed after the first CO pulse of the alternated ones. This method is based on the one presented by Duprez et al. [30,31]. The 6 gas composition at the exit of the reactor was analyzed by a mass spectrometer PFEIFFER Vacuum PrismaPlus controlled by Quadera® software. Several assumptions were contemplated for the OSC calculations. Concretely, it was considered that i) only oxygen atoms bonded to the cerium participate in the oxygen storage process; ii) the surface is assumed homogeneous iii) only one of the four oxygen atoms is involved in the storage (CeO2  Ce2O3 + “O”); and iv) null gold metal contribution to the reduction, e.g. the gold metal could not be reoxidized. 2.3. Catalytic Activity Water-gas shift reaction was performed in a stainless steel tubular flow reactor (0.75 cm ID) at atmospheric pressure in the 140-350 ºC temperature range .The catalysts were pelletized and sieved with the 600-800 μm fraction employed for the test. The following conditions were applied: catalyst bed volume 1.5 cm3, space velocity 4000 h-1 and gas composition: 30% vol. H2O + 4.5% vol. CO balance in N2. Water was injected into flowing gas stream by HPLC pump, vaporized and homogenized with the gas feed before entering the reactor. The CO and CO2 content was determined by on line ABB gas analyzer and the activity expressed in terms of CO conversion. The most active sample was submitted to long-term stability test, using gas composition imitating the outlet of an ethanol reformer - 30% H2O, 9% CO, 11% CO2 and 50% H2. 7 3. Results and discussion The actual composition of the prepared catalysts and their parent supports is presented in Table 1. All the samples are composed mainly by alumina (more than 80 wt. %) and about 15 wt. % of cerium oxide. The dopants loadings are close to the targeted value (2 wt. %) except for the Zn doped sample for which only 50 % of ZnO is incorporated to the catalyst. Most likely the ammonia treatment caused partial dissolution and loss of Zn during the filtration process. Regarding gold loadings, all the catalysts presented gold amounts very close to the nominal value pointing that the gold deposition was successfully carried out. Curiously, the doped supports admit slightly higher quantities of gold. It is well accepted that gold uptake depend on the point of zero charge (PZC) of the support [32], hence this result indicates that dopants could have smoothly modified this parameter. In addition, the presence of dopants may involve the creation of oxygen vacancies on ceria. These vacancies are viewed as electron rich sites where gold particles tend to nucleate [33, 34]. Therefore, the enhanced gold uptake observed for the doped ceria samples could suggests also higher concentration of oxygen vacancies on these systems compared to the unmodified one. Concerning the textural properties, all the samples are mesoporous materials with specific surface areas neighboring that of the primary support. Compared with a conventional Au/CeO2 material, these catalysts benefit of larger SBET due to the presence of the Al2O3 carrier which can be considered as a textural promoter in the prepared systems. XRD diffraction patterns of the studied supports and catalysts are presented in Figure 1A and 1B respectively. For all samples the diffraction lines corresponding to cubic CeO2 fluorite type structure (JCPDS# 00-004-0593) and to γ-Al2O3 phase (JCPDS# 00-048-0367) are detected. No 8 signals due to the crystalline Cu, Zn or Fe oxide phases are found. The low quantity of doping oxide, its high dispersion, or amorphous character may account for this fact. Similarly, no signals of metallic gold are observed indicating that gold particles are rather well dispersed and under the detection limit of the technique (< 5 nm). Further relevant information about the structural changes on ceria induced by the dopants can be extracted from the XRD. For example the crystallite size calculations using CeO2 (111) diffractions points some dissimilarities among the catalysts (Table 1). While Zn and Fe doped samples preserve the crystallite size of the undoped solid (about 5.5 nm) the Cu doped solid is composed by slightly larger ceria particles (7.3 nm). Consequently, smaller Au-CeO2 interface is expected for the Cu promoted sample. In addition, ceria lattice parameter (a) was estimated using the expression that relates the lattice parameter of the fluorite structure with the diffraction angle: a = √ℎ2+ 𝑘2+ 𝑙2 ( 𝜆 2sin𝜃) equation 1 The (111) CeO2 diffraction line (where no contribution of the γ-alumina phase is observed) was selected for this calculation being the λ factor that of Cu filament (1.5405 Å). The obtained values are listed in Table 1. From these data it can be concluded that only Fe introduces a notable change in the ceria lattice dimensions in both the support and the catalyst. A contraction of the CeO2 cell was obtained when iron is used as a dopant indicating the formation of a Ce-Fe solid solution. Actually considering the size of both cations, the decrease of ceria lattice parameter is expected. Indeed, iron oxide fulfills Hume-Rothery´s miscibility rules and the solid solution formation is favored. This phenomenon of Ce-Fe solid solution was broader addressed 9 elsewhere [26, 34] and basically its main consequence is the formation of oxygen vacancies to balance the lattice charges upon substitution. Complementary structural information is obtained through Raman spectroscopy experiments. Figure 2 shows the Raman spectra of the gold catalyst’s supports. For all the samples a band centered at about 458 cm-1 appeared. This band is ascribed to the F2g Raman mode of ceria fluorite cell corresponding to oxygen symmetric breathing vibration around Ce4+ [35]. Shifting and broadening of this vibration mode is observed for all the dopants indicating alterations of ceria structure and/or ceria particle size in good agreement with the XRD data discussed above. Moreover this type of modifications on the shape and position of the F2g Raman mode suggest a change on ceria electronic properties due to the presence of the dopants. In particular the observed blue shift in all doped samples indicates a strong Ce-M interaction [15]. Very importantly, for the Fe doped solid a second band centered at 530 cm-1 was detected. This band has been widely reported in literature and is associated with the presence of the oxygen vacancies preferentially placed on the surface of the cubic CeO2 [36, 37]. As commented in the XRD section, the inclusion of Fe3+ cation on the ceria lattice produces the formation of oxygen vacancies in the structure as a result of the Ce-Fe solid solution, confirmed by the Raman experiments providing a spectroscopic evidence of the existence of such punctual defects. This higher oxygen vacancies concentration may also explain the greater gold uptake observed for the Fe doped solid in comparison to the other two dopants. In summary from both XRD and Raman data an intimate Ce-M interaction can be assumed. This effect influences the structural and electronic properties of the resulting materials. In particular Fe modifies the oxygen vacancies population on ceria structure which may lead to redox altered properties. For an accurate understanding of the redox properties of the presented catalysts the 16 [5] M.V. Twigg, Catalysis handbook second ed. Wolf Scientific Texts, London 1989. [6] C. Ratnasamy, J. Wagner, Catal. Rev. 51 (2009) 325-440. [7] D.L. Trimm, Appl. Catal. A 296 (2005) 1-11. [8] T. Giroux , S.Hwang, Y. Liu, W. Ruettinger, L. Shore Appl. Catal. B 55 (2005) 185-200. [9] R.J. Farrauto, Y. Liu, W. Ruettinger, O. Ilinich, L. Shore, T. Giroux, Catal. Rev. 49 (2007) 141–196. [10] K.C. Petallidou, A.M. Efstathiou, Appl. Catal. B 140–141 (2013) 333–347. [11] Q. Fu, H. Saltsburg, M. Flytzani-Stephanopolous, Science, 310 (2003) 935-938. [12] P. Panagiotopoulou, J. Papavasiliou, G. Avgouropoulos, T. Ioannides, D.I. Kondarides, Chem. Eng. J. 134 (2007) 16–22. [13] M. Boaro, M. Vicario, J. Llorca, C. de Leitenburg, G. Dolcetti, A. Trovarelli, Appl. Catal. B 88 (2009) 272–282. [14] S. Colussi, L. Katta, F. Amoroso, R. J. Farrauto, A.Trovarelli Catal Comm. 5 (2014) 63-66 [15] M. Gonzalez Castaño, T.R. Reina, S. Ivanova, M.A. Centeno, J.A. Odriozola J. Catal. 314 (2014) 1–9. [16] T.R.Reina, M.Gonzalez-Castaño, S.Palma, S. Ivanova, J.A. Odriozola in “Heterogeneous gold catalysts and catalysis” eds. Z.Ma and S.Dai, RSC catalysis series 18. (2014) 111-139. [17] Q. Fu, W. Deng, H. Saltsburg, M. Flytzani-Stephanopoulos Appl. Catal., B, 56 (2005) 57- 68. [18] G. Jacobs, E. Chenu, P.M. Patterson, L. Williams, D. Sparks, G. Thomas, B.H. Davis, Appl. Catal., A, 258 (2004) 203-214. [19] F. Vindigni, M. Manzoli, A. Damin, T. Tabakova, A. Zecchina, Chem. Eur. J., 2011, 17, 4356-4361. 17 [20] D. Andreeva, I. Ivanov, L. Ilieva, J. W. Sobczak, G. Avdeev, T. Tabakova, Appl. Catal., A., 333 (2007) 153-160. [21] J.A. Rodriguez, Catal. Today, 160 (2011) 3-10. [22] T. Tabakova, V. Idakiev, D. Andreeva, I. Mitov, Appl. Catal. A 202 (2000) 91–97. [23] R. Burch, Phys. Chem. Chem. Phys. 47 (2006) 5483–5500. [24] H. Sakurai, T. Akita, S. Tsubota, M. Kiuchi, M. Haruta Appl. Catal., A 291 (2005) 179-187. [25] C. H. Kim, L. T. Thompson, J. Catal. 230 (2005) 66-74. [26] T.R. Reina, S. Ivanova, J.J. Delgado, I. Ivanov, T. Tabakova, V. Idakiev, M.A. Centeno, J.A. Odriozola, ChemCatChem 6 (2014) 1401-1409. [27] G. Avgouropoulos, M. Manzoli, F. Boccuzzi, T. Tabakova, J.Papavasiliou, T. Ioannides, V. Idakiev, J. Catal. 256 (2008) 237-247. [28] T.R. Reina, S. Ivanova, V. Idakiev, J.J. Delgado, I. Ivanov, T. Tabakova, M.A. Centeno, J.A. Odriozola, Catal. Sci. Technol. 3 (2013) 779–787. [29] S. Ivanova, C. Petit, V. Pitchon, Appl. Catal. A 267 (2004) 191–201. [30] S. Kacimi, J. Barbier , R. Taha, D. Duprez. Catal. Lett. 22 (1993) 343-50. [31] S. Bedrane, C. Descorme, D. Duprez. Catal.Today 75 (2002) 401-405. [32] F. Moreau, G.C. Bond, Appl. Catal. A 302 (2006) 110–117. [33] W.Y. Hernández, M.A. Centeno, F. Romero-Sarria, J.A. Odriozola, J. Phys. Chem. C 113 (2009) 5629–5635. [34] O.H. Laguna, F. Romero-Sarria, M.A. Centeno, J.A. Odriozola, J. Catal. 276 (2010) 360–370. [35] T.R.Reina, A.Alvarez, S.Ivanova, J.A. Odriozola, M.A. Centeno ChemCatChem 4 (2012) 512-520. 18 [36] H. Bao, X. Chen, J. Fang, Z. Jiang, W. Huang Catal. Lett., 125 (2008) 160–167. [37] Z.Y. Pu, J.Q. Lu, M.F. Luo, Y.L. Xie J. Phys Chem.C, 111 (2007) 18695–18702. [38] O. H. Laguna, M. A. Centeno, F. Romero-Sarria and J. A. Odriozola, Catal. Today 172 (2011) 118-123. [39] S. Bedrane, C. Descorme and D. Duprez, Catal.Today 75 (2002) 401-405. [40] J.A. Rodriguez, J.C. Hanson, D. Stacchiola, S.D. Senanayake, Phys. Chem. Chem. Phys. 15 (2013) 12004–12025. [41] C.T. Campbell, Nature Chemistry 4 (2012) 597-598. [42] T.R Reina, W.Xu, S.Ivanova, M.A. Centeno J.Hanson, J.A. Rodriguez, J.A. Odriozola Catal. Today 205 (2013) 41-48.