Hydrogen production from ethanol for PEM fuel cells. An integrated fuel processor comprising ethanol steam reforming and preferential oxidation of CO

doi:10.1016/j.cattod.2009.02.006

Catalysis Today

Volume 146, Issues 1–2, 15 August 2009, Pages 110-123

https://doi.org/10.1016/j.cattod.2009.02.006 Get rights and content

Abstract

The aim of the work was to study the performance of ceria catalysts to convert ethanol to hydrogen in a combined system including ethanol steam reforming and PROX. The roles of the active oxide component, partially reduced ceria, and the metal component, Pt, in the ethanol steam reforming mechanism were investigated by diffuse reflectance infrared spectroscopy (DRIFTS) carried out under steady state reaction conditions. The main mechanism was found to proceed by (1) dissociative adsorption of ethanol to ethoxy species; (2) dehydrogenation of ethoxy species to adsorbed acetaldehyde; (3) oxidation of acetaldehyde species by ceria OH groups to acetate; (4) acetate demethanation to CH₄ and carbonate species; (5) carbonate decomposition to CO₂; and presumably (6) CH₄ decomposition steps. Though Pt improved the initial ethanol conversion rate by facilitating hydrogen transfer and demethanation steps, the Pt–ceria interface was quickly lost to the buildup of carbon-containing species, thus hindering the Pt from effectively demethanating the acetate intermediate. Unpromoted ceria, though less active, was a significantly more stable catalyst.

The steps for the PROX reaction in the presence of acetaldehyde were found to include: (1) decomposition of acetaldehyde leading to CO and methane; (2) hydrogenation of acetaldehyde producing ethanol; and (3) oxidation of the CO.

Introduction

Nowadays, the absence of a suitable hydrogen distribution infrastructure is one of the main barriers to the use of hydrogen as an energy carrier or as a fuel for energy generation [1], [2]. In the early stages of the transition to a hydrogen economy, hydrogen should be produced locally at refueling stations using small-scale fuel processors [3]. These fuel processors pose new challenges for catalyst design as a consequence of their operating conditions and the need for high performance at low cost [4]. Depending on the applications, hydrogen rich streams with very low CO contents (below 10 ppm) are necessary and then further processing steps are required. In this case, the reformer is followed by a clean up step such as the water–gas shift reaction (WGS), CO preferential oxidation (PROX), pressure swing adsorption (PSA) or CO methanation. The selection of the different stages of the fuel processor affects significantly the efficiency and the cost of the fuel processor [5]. Different approaches have been proposed to reduce the complexity of the fuel processor such as the use of membrane reactors, which integrates the reaction and the separation steps within the same reactor, or the development of new catalysts designed specifically for the small-scale processors [4]. They must be active, selective to hydrogen and stable under the reaction conditions.

For example, a fuel processor to generate hydrogen from ethanol is a rather attractive technology since ethanol is a renewable raw material obtained from fermentation of various agricultural products (e.g., sugar cane). This light alcohol is a high energy density carrier for hydrogen and can be more easily transported by vehicle or pumped to a point of use, where it can be reformed to release the hydrogen with a fuel processor. However, the development of suitable catalysts to effect this reaction is proving to be a challenge. In the literature, the majority of authors [6], [7], [8], [9], [10], [11], [12], [13], [14], [15] reported using supported metals as catalysts for the ethanol reforming reactions. In general, these catalysts exhibited optimal performance at high temperatures (between 873 and 1023 K) and consequently, large amounts of CO were formed, which increases the fuel processor cost, due to the added complexity of the downstream CO conversion process. In addition, catalyst deactivation due to carbon deposition is an important issue [16], [17], [18]. A different catalyst design approach is to use metal oxides as catalysts for steam reforming of ethanol [19], [20], [21], [22]. In spite of their lower activity than supported metal catalysts, metal oxides are capable of producing hydrogen free of CO as well as carbon deposits, depending on the reaction conditions used. This may eliminate some purification steps and as a result reduce the costs. However, a wide range of undesirable by-products (e.g., ethene, acetaldehyde and acetone) is formed during steam reforming of ethanol over metal oxides in comparison with supported metal catalysts. For instance, acetaldehyde produced upstream may impact the performance of the downstream PROX catalyst. Therefore, the design of an appropriate catalyst that can handle both the reforming as well as the purification steps would go a long way to improve the overall efficiency of the fuel processor.

Ceria and ceria-containing mixed oxides have been proposed as catalytically active components for ethanol conversion reactions due to their high oxygen storage capacity, which improves catalyst stability [23], [24], [25], [26], [27]. In addition, the strong metal–support interaction prevents metal particle sintering, which also contributes to catalyst deactivation. Recently, we have investigated the performance of supported Pt catalysts for steam reforming of ethanol [26], [27]. At low reaction temperature, the Pt/CeZrO₂ catalyst underwent significant deactivation during ethanol decomposition and steam reforming reactions. Co-feeding oxygen decreased the deactivation rate of the catalyst but adversely impacted the selectivity to hydrogen. Increasing the reaction temperature greatly improved the stability of the catalyst [26]. A reaction mechanism was proposed based upon results obtained from in situ diffuse reflectance infrared spectroscopy (DRIFTS) analyses carried out under reaction conditions. The effect of the support nature and metal dispersion on the performance of supported Pt catalysts during steam reforming of ethanol was also studied [27]. H₂ and CO production were facilitated over Pt/CeO₂ and Pt/CeZrO₂, whereas the acetaldehyde and ethene formation were favored on Pt/ZrO₂. However, regardless of the support used, all the catalysts significantly deactivated during the reaction at low temperature. Unpromoted cerium oxide has also been used as a catalyst for the steam reforming of ethanol [19], [20], [22]. CeO₂ exhibited activity for steam reforming of ethanol while CO was not detected or it was formed in very low concentrations.

Many catalysts have been considered for the PROX reaction. A recent work published by Park et al. [28] describes an overview of the performance of many catalytic systems, including Au, Pt, Cu, Rh and Ru supported on different oxides. However, as far as we know in the literature, there is no data related to the impact of oxygenates on the catalytic performance of the PROX reaction.

The aim of this work is to study the performance of catalysts used to convert ethanol to hydrogen in a combined system including steam reforming and PROX reactions. CeO₂ and Pt/CeO₂ catalyst were tested for the steam reforming of ethanol and the role of the support on the reaction mechanism was investigated using diffuse reflectance infrared spectroscopy carried out under steady state reaction conditions. Attention was also focused on the activity of Pt/CeO₂ catalysts for CO and acetaldehyde removal.

Section snippets

Catalyst preparation

CeO₂ and Al₂O₃ were used as supports. Al₂O₃ was supplied by NORPRO and CeO₂ support was obtained by calcination of (NH₄)₂Ce(NO₃)₆ in a muffle furnace at 1073 K for 1 h.

Platinum was added to the supports by the incipient wetness impregnation technique using an aqueous solution of H₂PtCl₆·6H₂O. After impregnation, the samples were dried at 393 K and calcined under air flow (50 mL/min) at 673 K for 2 h. The following catalysts were obtained: 0.5%Pt/CeO₂, 1.0%Pt/CeO₂, 1.5%Pt/CeO₂, 2.0%Pt/CeO₂ and 1.5%

Catalyst characterization

The BET surface area of CeO₂ support was very low (14 m²/g) and it was not measurably affected by the addition of platinum. Alumina exhibited a BET surface area of 200 m²/g. The Pt dispersion values obtained were 58 and 80% for 1.5%Pt/CeO₂ and 1.5%Pt/Al₂O₃ catalyst, respectively.

The TPR profiles of CeO₂ and 1.5%Pt/CeO₂ are presented in Fig. 1. For the CeO₂, a small H₂ uptake at 817 K and a strong H₂ consumption at 1230 K were observed. The first peak is attributed to the surface reduction of CeO₂

Conclusions

While 1.5%Pt/CeO₂ was found to be a significantly more active catalyst than ceria alone for ethanol steam reforming, the catalyst was found to deactivate very rapidly. The mechanism of ethanol decomposition was probed by TPD and DRIFTS, while the mechanism of ethanol steam reforming was monitored by steady state DRIFTS. The mechanism of ethanol steam reforming was found to involve six main steps, including (1) dissociation of ethanol on partially reduced ceria to ethoxy species and a bridging

Acknowledgements

This work received financial support of CTENERG/FINEP-01.04.0525.00. CAER acknowledges the Commonwealth of Kentucky for financial support.

References (50)

J.M. Ogden
Int. J. Hydrogen Energy
(1999)
P. Forsberg et al.
Int. J. Hydrogen Energy
(2007)
H. Song et al.
Catal. Today
(2007)
G. Jacobs et al.
J. Catal.
(2007)
L.P.R. Profeti et al.
J. Power Sources
(2008)
M. Veronica et al.
Chem. Eng. J.
(2008)
A.N. Fatsikostas et al.
J. Catal.
(2004)
A. Erdohelyi et al.
Catal. Today
(2006)
V. Fierro et al.
Catal. Today
(2002)
J.C. Vargas et al.
Catal. Today
(2005)

T. Nishiguchi et al.

Appl. Catal. A

(2005)

J. Llorca et al.

J. Catal.

(2004)

W. Cai et al.

J. Catal.

(2008)

J. Kugai et al.

J. Catal.

(2006)

S.M. de Lima et al.

J. Catal.

(2008)

S.M. de Lima et al.

Catal. Today

(2008)

E.D. Park et al.

Catal. Today

(2009)

M. Boudart

Adv. Catal.

(1969)

P. Pantu et al.

Appl. Catal. A

(2002)

F.B. Passos et al.

Catal. Today

(2005)

M. Dömök et al.

Appl. Catal. B

(2007)

L.V. Mattos et al.

J. Catal.

(2005)

A. Yee et al.

J. Catal.

(1999)

M.A.S. Baldanza et al.

J. Catal.

(2000)

E.M. Cordi et al.

J. Catal.

(1996)

Cited by (58)

Insights into the state of ceria during ethanol steam reforming over Ir/CeO<inf>2</inf>
2022, Catalysis Science and Technology
Highly active Ir/CeO₂ is investigated following 240 h membrane-assisted ethanol steam reforming at 873 K to elucidate the effect of H₂ removal on catalyst activity. Sub-nm Ir species aggregated into particles with 1.4 nm mean diameter retaining ca. 78% Ir dispersion. Surface area declined by 83% due to ceria particle growth which were extensively reduced and difficult to re-oxidize at 673 K. This distinct CeO_2−x state includes 60–65% Ce^III near its surface. Correlations with ethanol and acetaldehyde steam reforming kinetic analyses of Ir/CeO₂, Ir/CeO_2−x, and a partially re-oxidized catalyst suggest that the initiating ethanol dehydrogenation involves only Ir sites and possibly CeO₂ at their perimeter. In contrast, ceria controls the subsequent acetaldehyde conversion via activation of steam and probably also directly as oxidant. Product distributions reveal that acetaldehyde reforming via acetic acid is favored on Ir/CeO₂ at lower temperatures while its decomposition into CH₄ and CO prevails on Ir/CeO_2−x at elevated temperatures.
Molecular structure models of amorphous bismuth and cerium carboxylate catalyst precursors
2022, Catalysis Today
As our societal need for materials and energy has grown, so has our need for catalyst processes in hydrogen production. A major function in these applications, for both homogenous and heterogeneous catalysis processes, is the synthesis of an active metal catalyst. It must first be soluble to control the physical properties of the metal being used. Recent work in metal precursors has begun to turn toward these metal carboxylate types of material. Structural models are proposed for bismuth 2-ethylhexanoate and 2,2-dimethyloctanoate and cerium 2-ethylhexanoate. The bismuth compounds have been characterized at different ratios of bismuth to carboxylate as solutions of the free acids. Their structures are most consistent with a Bi₄(RCO₂)₁₂ motif where the Bi ions are arranged in a flattened tetrahedron with Bi – Bi distances of about 4.3 Å. There is evidence for Bi – O – Bi linkages at low free acid concentrations. The cerium compound is most consistent with a linear tetracerium molecule where the Ce – Ce distances repeat at about 4.3 Å out to 16.4 Å. The models were generated by analogy with known crystal structures and compared to high-energy x-ray scattering data. To further evaluate the models, DFT calculations were made, and the equilibrium geometries were compared. The vibrational spectra calculated from those geometries are presented and compared to the experimental results. Magnetization vs. temperature data was collected on the cerium compound, and its behavior was consistent with the proposed model. A geometrical approach to determining the dimensionality and relative positions of the metal ions in these structures is presented.
The effects of support morphology on the performance of Pt/CeO<inf>2</inf> catalysts for the low temperature steam reforming of ethanol
2021, Applied Catalysis B: Environmental
Citation Excerpt :
In addition, the physicochemical properties of the support have been found to strongly affect the activity of dispersed platinum towards the WGS. In particular, the activity has been found to increase by about 1–2 orders of magnitude when Pt is supported on “reducible” compared to “irreducible” metal oxides [6,24]. CeO2 has been widely studied for the SR of ethanol due to its unique characteristics as a redox catalyst (such as easy Ce4+ reducibility and high mobility of the O2− species within its lattice) [25].
A series of Pt/CeO₂ catalysts with different support morphologies was studied for the low temperature ethanol steam reforming. Different ceria nanostructures (nanocube_NC, nanorod_NR and flower-like_FL) were prepared by the hydrothermal method and are compared with CeO₂ obtained by precipitation. Support morphology was found to influence certain reaction routes, even though the reaction scheme was the same regardless of the support. Concerning catalyst stability at 573 K, Pt/CeO₂- PPT exhibited rapid deactivation, whereas Pt/CeO₂- NR remained rather stable. XPS surface analysis before and after reaction at 573 K revealed significant degree of Pt reduction on CeO₂ nanorods, directly related to the reduction of Ce⁴⁺ into Ce³⁺. DRIFTS experiments under reaction conditions for 5 h, revealed an accumulation of acetates on the Pt/CeO₂- PPT catalyst, likely due to the imbalance between the rate of acetate species decomposition to CH_x (not facilitated over Pt^δ+ of the catalyst with the lower Pt dispersion), and their desorption as CH₄.
Insights into the essential roles of tin and chloride species within Cu–CeO<inf>2</inf> based catalysts for CO preferential oxidation in H<inf>2</inf>-rich stream
2021, Journal of Power Sources
Citation Excerpt :
Hydrogen for PEMFCs is primarily obtained by reforming the alcohols and hydrocarbons, and followed by water gas shift (WGS) reaction. Nonetheless, H2 obtained still contains 0.3%–1% of carbon monoxide (CO) owing to the thermodynamic equilibrium limitation, easily poisoning the Pt-based electrode of PEMFCs [3,4]. Consequently, it is essential to develop effective methods to decrease the amount of CO below 10 ppm in the H2-rich streams [5], and CO preferential oxidation (CO-PROX) has been considered as a simple and efficient process [6,7].
Sn modified Cu–Ce catalysts have been synthesized by the solvent-free combustion method for preferential oxidation of CO in H₂-rich stream (CO-PROX) and characterized by N₂ adsorption-desorption (BET), H₂ temperature-programmed reduction (H₂-TPR), X-ray diffraction (XRD), Raman, in-situ DRIFTS, X-ray photoelectron spectroscopy (XPS), and Oxygen storage capacity (OSC). The results show that the incorporation of Sn species accelerates the redox cycle among Sn, Cu and Ce, thus strengthening the Cu–Ce interaction and promoting the formation of surface and bulk lattice Cu-[O_x]-Ce structures Meanwhile, to balance the charge, more oxygen vacancies are formed and thus facilitate the migration of oxygen species. Combined with the more strong absorbed ability of SnO₂ for O₂, active Cu-[O_x]-Ce structures promotes the easier conversion of more CO into CO₂. However, the existence of Cl ions decreases catalytic performance because the formation of CuCl makes the Cu–Ce interaction weakened, resulting in the formation of fewer Cu-[O_x]-Ce structures and oxygen vacancies. Although CuCl is prone to be in favor of absorbing more CO, there are not enough O species available, hence a decrease in performance is observed.
A study on catalytic hydrogen production: Thermodynamic and experimental analysis of serial OSR-PROX system
2018, Fuel Processing Technology
Steady state performance of OSR and serial OSR-PROX reactors were investigated. OSR of methane/propane and PROX were conducted over novel Pt-Ni/δ-Al₂O₃ and Pt-Sn/AC catalysts, respectively. OSR (623–773 K) and PROX (383–388 K) reactors' temperature combination, S/C ratio (2.25–6) in the OSR feed, and W/F_O₂ ratio (1.5–12.2 mg-cat·min/μmol) in the PROX feed, which was adjusted through oxygen addition to OSR outlet at rates on the basis of CO molar flow in the OSR outlet, were used as the experimental parameters. Simultaneous use of high temperature and high S/C ratio led to increased H₂ and H₂/CO ratios, and decreased CH₄ and CO₂ concentrations in OSR product. In serial tests, the highest achieved H₂ concentration was 66% (dry based inert free) at S/C ratio of 6 at 723 K, while the lowest CO concentration was 700 ppm at OSR-PROX temperature combination of 723 & 383 K. Maximum achieved methane and propane conversions at 723 K were 70% and 100%, respectively. Oxygen fed to OSR was totally consumed in all tests. OSR temperature of 723 K, S/C ratio of 5, PROX temperature of 383 K and W/F_O₂ ratio of 1.53 in the PROX feed were the optimal parameters targeting high H₂/CO product ratio. The experimental product distribution trends were consistent with the results of the thermodynamic equilibrium calculations.
Evolution of the structure of unpromoted and potassium-promoted ceria-supported nickel catalysts in the steam reforming of ethanol
2018, Applied Catalysis B: Environmental
Unpromoted and potassium-promoted Ni/CeO₂ catalysts, fresh, reduced and used in the SRE were characterized by TEM (as a leading method), SEM, XRD, hydrogen chemisorption, H₂-TPR and Raman spectroscopy. The evolution of the nature and the size of the nickel active phase (different in both catalysts) as well as distribution of potassium is disclosed. The promoter does not protect the catalyst against formation of carbonaceous deposit and does not improve its stability in the SRE. More important is high dispersion of nickel. Smaller nickel crystallites in Ni/CeO₂ ensure its better stability. In both catalysts, fragments of nickel (devoid of potassium in the case of KNi/CeO₂) are pushed from ceria by carbonaceous fibres formed. Their nature is different in both catalysts. On the Ni/CeO₂ amorphous (disordered) long and thick as well as short and thin graphitic fibres are formed. The KNi/CeO₂ produces very large amount of only graphitic fibres, longer and thicker than those in unpromoted catalyst. Much intensive production of graphitic fibres, which with time-of-stream can encapsulate greater number of nickel crystallites, is considered as responsible for faster deactivation of KNi/CeO₂ catalyst in the SRE.

View all citing articles on Scopus

¹: Present address: Universidade Estadual do Oeste do Paraná - Unioeste, Campus de Toledo, Rua da Faculdade, 645, Jd. La Salle, CEP 85903-000, Toledo, Brazil.

View full text

Hydrogen production from ethanol for PEM fuel cells. An integrated fuel processor comprising ethanol steam reforming and preferential oxidation of CO

Abstract

Introduction

Section snippets

Catalyst preparation

Catalyst characterization

Conclusions

Acknowledgements

Int. J. Hydrogen Energy

Int. J. Hydrogen Energy

Catal. Today

J. Catal.

J. Power Sources

Chem. Eng. J.

J. Catal.

Catal. Today

Catal. Today

Catal. Today

Appl. Catal. A

J. Catal.

J. Catal.

J. Catal.

J. Catal.

Catal. Today

Catal. Today

Adv. Catal.

Appl. Catal. A

Catal. Today

Appl. Catal. B

J. Catal.

J. Catal.

J. Catal.

J. Catal.