Contents lists available at ScienceDirect

Fuel Processing Technology

journal homepage: www.elsevier.com/locate/fuproc

Research article

Optimum operating conditions in ethanol steam reforming over a Ni/La2O3-αAl2O3 catalyst in a fluidized bed reactor

Carolina Monteroa,⁎, Aingeru Remirob, Pedro Luis Benitoc, Javier Bilbaob, Ana G. Gayubob

a Chemical Engineering Faculty, Central University of Ecuador, Ciudad Universitaria-Ritter s/n y Bolivia, Quito, Ecuadorb Chemical Engineering Department, University of the Basque Country, P.O. Box 644, 48080, Bilbao, Spainc Mining and Metallurgical Engineering and Materials Science Department, University of the Basque Country, C/Nieves Cano 12, 01006 Vitoria-Gasteiz, Spain

A R T I C L E I N F O

Keywords:EthanolHydrogen productionSteam reformingFluidized reactor

A B S T R A C T

This manuscript analyzes the steam reforming of ethanol (SRE) over a Ni/La2O3-αAl2O3 catalyst in a fluidizedbed reactor under a wide range of operating conditions (500–650 °C, space time up to 0.35 gcatalysth/gEtOH, andsteam/ethanol (S/E) molar ratio in the feed between 3 and 9) in order to select optimum conditions for max-imizing H2 production. The significance the individual reactions in the reaction mechanism have on productsdistribution and the role of the catalyst in the extent of these reactions has also been analyzed. Blank runs(without catalyst) have been performed to test the contribution of thermal routes to this mechanism. Ethyleneand acetaldehyde are intermediate products in the kinetic scheme, whose presence is only observed whenethanol conversion is not full. The increase in temperature enhances the reforming and decomposition of ethanoland acetaldehyde and, when the catalyst is used, CH4 reforming and reverse WGS reactions are also promoted, sothat the yield of H2 and CO increases, that of CH4 decreases and the one of CO2 remains almost constant withtemperature. The increase in S/E molar ratio increases H2 yield, but attenuates the rate of some reactions in-volved in the process. 600 °C, a space time of 0.35 gcatalysth/gEtOH and S/E = 6 are suitable conditions formaximizing ethanol conversion (100%) and H2 yield (82%) with high catalyst stability.

1. Introduction

The foreseen 30% growth in worldwide energy demand for 2040,together with the increasing social awareness concerning the negativeconsequences of the use of fossil fuels, have boosted the development oftechnologies for maximizing energy production from renewablesources, so that 37% of power generation will be from renewable re-sources in 2040, compared to 23% today [1]. Among these, biorefinerytechnologies aimed at converting different biomass types into chemicalsand fuels have a relevant role [2], and the reforming of biomass derivedoxygenates has gained an important strategic interest because of theincreasing demand of H2 for use as a fuel, and as raw material in pet-rochemical industry and agrochemistry [3].

Among biomass derived oxygenates, bio-ethanol has great interestas raw material for producing H2 by reforming [4,5], mainly due to thegood perspectives for its production from lignocellulosic biomass, witha forecasted increase from the current 270 L/t biomass to 400 L/t bio-mass in 2030, as a result of the advance in the technology of enzymatichydrolysis-fermentation [6]. Moreover, the steam reforming (SR) ofbio-ethanol (~86% H2O) avoids the high cost required for its

dehydration (estimated at 50% of the total product cost [7]) in order tobe used as a fuel (dehydrated ethanol).

The steam reforming of ethanol (SRE) is an endothermic processthat proceeds at relatively low temperatures (between 300 and 800 °C),with the following stoichiometry:

C2H5OH+3H2O→6H2+2CO2 ΔH2980=173.3 kJ/mol (1)

The use of steam/ethanol (S/E) molar ratio in the feed above thestoichiometric value (S/E = 3) improves H2 selectivity and attenuatesdeactivation by coke deposition [8]. Nevertheless, the reaction me-chanism is complex due to secondary reactions that take place in par-allel to the steam reforming reaction and generate intermediate pro-ducts and by-products, thus reducing H2 yield. Among the secondaryreactions, the following are considered [9–11]:

↔ +Ethanol dehydrogenation: C H OH C H O H2 5 2 4 2 (2)

→ +Ethanol dehydration: C H OH C H H O2 5 2 4 2 (3)

→ + +Ethanol decomposition: C H OH H CO CH2 5 2 4 (4)

+ → +Acetic acid formation: C H OH H O CH COOH 2H2 5 2 3 2 (5)

http://dx.doi.org/10.1016/j.fuproc.2017.10.003Received 5 April 2017; Received in revised form 6 September 2017; Accepted 5 October 2017

⁎ Corresponding author.E-mail address: cdmontero@uce.edu.ec (C. Montero).

Fuel Processing Technology 169 (2018) 207–216

0378-3820/ © 2017 Elsevier B.V. All rights reserved.

MARK

→ + +Acetone formation: 2 C H OH CH COCH CO H2 5 3 3 2 (6)

+ → +Acetone steam reforming: CH COCH 2H O 3CO 5H3 3 2 2 (7)

+ → +Acetic acid steam reforming: CH COOH 2H O 2CO 4H3 2 2 2 (8)

+ → + +Incomplete ethanol reforming: C H OH H O CH 2H CO2 5 2 4 2 2

(9)

→ +Acetaldehyde decomposition: C H O CH CO2 4 4 (10)

+ → +Acetaldehyde reforming: C H O H O 2CO 3H2 4 2 2 (11)

+ → +C H O 3H O 2CO 5H2 4 2 2 2 (12)

+ ↔ +Water Gas Shift reaction: CO H O H CO2 2 2 (13)

+ ↔ +

Methane steam reforming reverse to methanation of CO and CO( )

: CH H O CO 3H2

4 2 2 (14)

+ ↔ +CH 2H O CO 4H4 2 2 2 (15)

+ → +Ethylene steam reforming: C H 2H O 4H 2CO2 4 2 2 (16)

Other reactions take also place, but they are not relevant for pro-ducts distribution, although they contribute to catalyst deactivation byformation or elimination (by gasification) of coke (C):

→ →Ethylene polymerization: C H polymers C2 4 (17)

↔ +Boudouard reaction: 2CO C CO2 (18)

→ +Methane decomposition: CH 2H C4 2 (19)

+ → +Coke gasification: C H O CO H2 2 (20)

Due to the complexity of the reaction scheme, the yield and se-lectivity of H2 is highly affected by reaction conditions (temperature, S/E molar ratio and space time), as well as by catalyst composition.Consequently, the industrial viability of the SRE process requires thedevelopment of highly active and selective catalysts for H2 formation(minimize secondary reactions), which are stable and hardly affectedby coke formation. With this objective in mind, several reviews haveanalyzed the use of catalysts with both noble and non-noble metalssupported on different oxides [9,12–15]. Noble metal catalysts, espe-cially Rh based catalysts, are highly active and selective for SRE[16,17], but their practical applications are limited by their high cost.Among the non-noble catalysts, those based on Ni and Co are the moststudied due to their high CeCe bond breakage activity [18–23].

Furthermore, it is well established that an increase in metal contentin Ni based catalysts improves ethanol conversion, but does not guar-antee a higher H2 selectivity. Thus, Han et al. [24] determined an op-timum content of 15 wt% Ni for a catalyst prepared by sol-gel tech-nique, which showed a high Ni dispersion and resistance to cokedeposition. Gayubo et al. [25] reported an optimum content of 10 wt%Ni in Ni/SiO2 and Ni/α-Al2O3 catalysts prepared by incipient wetnessimpregnation, which is due to a higher Ni content leading to a sig-nificant agglomeration of metal crystals.

γ-Al2O3 has been widely used as support because of its high thermaland mechanical stability, together with a high specific surface area,which also improves the dispersion of the active phase. Nevertheless, itsacidity promotes ethanol dehydration reaction and, as a result, cokedeposition via ethylene, which causes a rapid deactivation of the cat-alyst. Consequently, several methods have been studied for the neu-tralization of its acidity. The addition of basic additives, such as CaO,lowers the support acidity and weakens the interaction between Ni andAl2O3, which facilitates the reduction of Ni+2 species to Ni0 [26,27].However, Ca contents above 5 wt% increase Ni active particle size,which causes a lower H2 yield [27] and promotes the formation ofencapsulating coke responsible for the rapid deactivation of the catalyst[26]. The addition of MgO to γ-Al2O3 caused similar results to thoseobtained by doping with CaO [28–30]. Nevertheless, the doping of the

support ZrO2 with CaO did not affect Ni reducibility and hindered cokedeposition [31]. Furthermore, the addition of La2O3 provides stabilityto Ni/Al2O3 catalyst by lowering coke formation rate [32–34]. In aprevious work, a Ni/La2O3-αAl2O3 catalyst used in the SRE reactionachieved an equilibration state subsequent to a reaction-regenerationcycle (consisting in the steam reforming at 700 °C followed by cokecombustion with air at 550 °C), which allowed attaining a reproducibleperformance in successive reaction-regeneration cycles [35].

In view of this background, this work analyzes the effect operatingconditions (temperature, S/E molar ratio and space time) have on thebehavior of an equilibrated Ni/La2O3-αAl2O3 catalyst in the SRE pro-cess, in order to determine the conditions maximizing ethanol conver-sion and H2 yield. The study covers a wide range of operating condi-tions, including those for kinetic regime control (a limited extent of theindividual reactions) in order to get information on the nature of thereaction intermediates. Moreover, the significance individual reactionsof the complex reaction mechanism (Eqs. 1–16) have on product dis-tribution has also been analyzed, which has allowed establishing anoriginal kinetic scheme accounting for the significance of individualreactions under the conditions studied. With the aim of assessing thecontribution of the thermal reaction steps to the overall scheme, blankruns (without catalyst) have also been performed. Moreover, catalyststability has been approached by means of long duration runs (200 h),as it is an essential aspect for larger scale applications. A Ni/La2O3-αAl2O3 catalyst subjected to thermal equilibration treatment [35] hasthe additional interest that the results are reproducible when successivereaction-regeneration cycles are carried out. The fluidized bed reactoris also interesting for ensuring bed temperature homogeneity because,in addition to controlling this variable, it also has good perspectives forscaling-up.

2. Experimental section

2.1. Catalyst preparation

The catalyst was prepared by incipient wetness impregnationmethod [36], and with a composition (nominal contents of 10 wt% Niand 9 wt% La2O3) determined in a previous work [25]. It has beenproven that once calcined at 550 °C fSor 2 h in air, and subsequent to anequilibration treatment, the catalyst achieves a reproducible kineticbehavior in reaction-regeneration cycles [35]. Prior to the kinetic runsthe catalyst was reduced in situ at 700 °C for 2 h by using a H2–He flow(10 vol% H2). The properties (Table 1) have been determined as fol-lows: composition, by inductively coupled plasma and atomic emissionspectroscopy (ICP-AES) in a Thermo X7-II spectrometer; surface area(SBET) and porous structure, by N2 adsorption − desorption in aQuantacrome Autosorb IQ2 in physisorption mode; metal surface areaand dispersion, by H2 chemisorption in a Quantacrome Autosorb IQ2.The metal content in the catalyst is very close to the nominal value,which confirms that the preparation method by incipient wet impreg-nation is suitable. A comparison of the physical and metallic propertiesof Ni catalysts with different supports and their effect on the kineticperformance in the SRE reaction has already been studied by the au-thors [25]. It was proven that the addition of La2O3 to α-Al2O3 supporthad a minor unfavorable effect on the physical and metal properties of

Table 1Physical and chemical properties of Ni/La2O3–αAl2O3 catalyst.

Technique Property Results

N2 physisorption SBET (m2/g) 35ICP Ni (%) 8.8

La (%) 6.8H2 chemisorption Metal dispersion (%) 4.7

Metal active surface (m2/g) 3.1XRD Crystal size (Å), 2θ = 52° 106

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

208

the Ni supported catalyst, but it significantly contributed to the at-tenuation of catalyst deactivation.

Temperature programmed reduction (TPR) measurements wereconducted on an AutoChem II 2920 Micromeritics. The TPR profile ofthis catalyst showed three reduction peaks [35]: i) a peak below 380 °C,corresponding to the reduction of bulk NiO clusters with low interac-tion with the support; ii) a main reduction band, in the 400–700 °Crange (which can be decomposed into two peaks), usually ascribed inthe literature to the reduction of dispersed NiOx species, which areprobably amorphous and interact strongly with the support [37], or tothe reduction of LaNiO3 [38]; iii) a small peak above 700 °C, corre-sponding to the NiAl2O4 spinel, which has probably been formed bymigration of Ni atoms on Al2O3 [39].

The X-ray diffraction (XRD) pattern measured on a Bruker D8Advance diffractometer with a CuKα1 radiation showed diffractionlines corresponding to the reflection of Al2O3 phase and Ni0 phase (at2θ angle of 44.5°, 51.9° and 76.4°). LaAlO3 phase (La2O3 combined withα-Al2O3) is also detected [40].

2.2. Reaction equipment and experimental conditions

The runs were carried out in automated reaction equipment(Microactivity reference-PID Eng & Tech) provided with an isothermalfluidized bed reactor (22 mm internal diameter and total length of460 mm), connected on-line to a gas chromatograph (Agilent 3000)provided with four modules for the analysis of products: 1) permanentgases (O2, N2, H2, CO, and CH4) with 5A molecular sieve capillarycolumn; 2) light oxygenates (C2−), CO2 and water, with Plot Q capil-lary column; 3) C2-C4 hydrocarbons, with alumina capillary column; 4)oxygenated compounds (C2+) with Stabilwax type column. Calibrationstandards of known concentration were used to quantify and identifythe compounds. The balance of atoms (C, H, O) is closed in all runsabove 98%. As an example, Figs. S1 and S2 of supplementary in-formation show the C/H, C/O and H/O ratios at the inlet (lines) and atoutlet of the reactor (points) along time on stream, for two sets of ex-perimental conditions.

The use of a fluidized bed is interesting in order to guarantee theisothermicity of the bed and also to attenuate deactivation by cokedeposition [33,41]. In order to ensure a correct fluidization of the bedand avoid internal and external diffusional restrictions in the catalystparticles, the following conditions have been established [41]: a bedmade up of catalyst (particle diameter between 0.15 and 0.25 mm) andinert solid (CSi, 37 μm) in a mass ratio inert:catalyst > 8:1; bed height/bed diameter of 2; upwards gas flow rate at the reactor inlet~7.6 cm3 s−1, corresponding to a gas linear velocity of 2.4 cm s−1,which accounts for 6 times the minimum fluidization velocity, corre-sponding to a vigorous fluidization regime.

The conditions of the catalytic runs are as follows: temperaturebetween 500 and 650 °C; steam/ethanol molar ratio (S/E) between 3and 9; space time between 0.02 and 0.35 gcatalysth/gEtOH (with a catalystmass between 0.03 and 0.525 g); total pressure, 1.4 bar; partial pres-sure of ethanol, 0.08 bar.

Blank runs (without catalyst) have been performed in order to de-limit the temperature from which there is a noticeable contribution ofthermal routes to the reaction mechanism. These runs have been carriedout with inert solid (CSi) in the fluidized bed, in the 500–700 °C range,with an S/E molar ratio between 3 and 9, and maintaining the samehydrodynamic conditions as in the runs with catalyst.

2.3. Reaction indices

The kinetic behavior has been quantified by considering the fol-lowing reaction indices:

Conversion of ethanol (X), which is calculated from its molar flowrate (F) at the inlet and outlet (unreacted ethanol) of the catalytic re-actor:

=−X F F

Finlet outlet

inlet (21)

The yield of each product (Yi) is calculated as the ratio between themolar flow rate of product i (Fi) and the maximum molar flow rate thatmay be obtained according to stoichiometry when ethanol is fed intothe reactor:

=⋅

Y Fυ Fi

i

i inlet (22)

where υi = 6 for H2, υi = 2 for CO2, CO and CH4, and υi = 1 foracetaldehyde and ethylene.

The selectivity of each gaseous product (Si) is calculated as the ratiobetween the molar flow rate of product i (Fi) and the molar flow rate ofall reaction products (ethanol and water excluded):

=∑ − −

S FF F Fi

i

ii E W

(23)

3. Results and discussion

3.1. Contribution of thermal routes to SRE mechanism

Figs. 1 and 2 show the composition (molar fraction on a water freebasis) of the products stream at the reactor outlet obtained in the runswithout catalyst at different temperatures and S/E molar ratios. Fig. 1corresponds to ethanol (Fig. 1a) and H2 (Fig. 1b), whereas each graph inFig. 2 corresponds to a carbon product (acetaldehyde, ethylene, CO andCH4). The composition of CO2 (below 2% under all the studied condi-tions) is not shown. Moreover, neither acetone nor acetic acid has beenobserved in the products stream in the temperature range studied.

It is observed that at 500 °C ethanol conversion is low and theproducts stream is mainly composed of ethanol (~95%), with lowconcentrations of H2 (~2.5%), acetaldehyde (~2.5%), and ethylene(~0.1%). Moreover, at this temperature, the results are not affected byS/E molar ratio.

The increase in temperature up to 600 °C involves a slight increasein ethanol conversion. However, above this temperature ethanol con-version rapidly increases, so that its molar fraction decreases noticeably(Fig. 1a) and H2 molar fraction increases almost linearly, achieving avalue of 0.36 at 700 °C and for S/E = 9 (Fig. 1b). The molar fractions ofcarbon products have different trends with temperature (Fig. 2). Acet-aldehyde is the major product in the 600–650 °C range (Fig. 2a),whereas the formation of CO and CH4 increases exponentially withtemperature above 650 °C (Fig. 2c, d), and the formation of CO2 (notshown) is almost insignificant. These results are evidence that an in-crease in temperature promotes firstly ethanol dehydrogenation to formacetaldehyde (Eq. 2) (whose presence is significant at 500 °C), andsubsequently ethanol dehydration to form ethylene (Eq. 3) (whoseformation is noticeable above 550 °C), whereas the decomposition ofethanol (Eq. 4) or acetaldehyde (Eq. 10) to produce CO and CH4 areonly significant above 650 °C, and they are enhanced exponentiallywith temperature. The similar concentration of CO and CH4 obtained athigh temperature, as well as the almost insignificant formation of CO2 isevidence that, in the absence of catalyst, WGS reaction (Eq. 13), me-thane steam reforming (Eqs. 14 and 15) and direct steam reforming ofethanol (Eq. 1) do not have a significant contribution to the reactionmechanism in the 500–700 °C temperature range.

Furthermore, an increase in S/E molar ratio has lower effect thantemperature on these results. In general, all the thermal routes arepromoted by increasing S/E molar ratio from 3 to 6, although a furtherincrease in this variable has no significant effect.

There are scarce papers in the literature reporting results withoutcatalyst in order to analyze the thermal routes for ethanol conversionunder steam reforming conditions. The aforementioned results areconsistent with those by Fatsikostas and Verykios [42] obtained by

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

209

experimentation at temperature programmed under conditions of SREreaction (S/E = 3 in the 500–800 °C range). These authors concludedthat ethanol conversion is activated from 600 °C on, and is significantabove 700 °C. At low temperature, ethanol dehydrogenation was theprevailing reaction, whereas dehydration and cracking/dissociationwere the main reactions at high temperature. Melchor-Hernández et al.also obtained a significant ethanol conversion above 600 °C for S/E = 3[43]. Barattini et al. [44] operated in a fixed bed reactor with quartz asinert solid with S/E = 3 and found that ethanol conversion was sig-nificant above 430 °C, and complete at 790 °C. At low temperature(400 °C), they mainly obtained ethylene and acetic acid, whereas athigher temperature (730 °C) small amounts of CO, CH4 and H2 wereobtained and ethylene selectivity remained stable in the studied tem-perature range, and acetaldehyde was not detected, with the latter re-sult differing from that in this paper and in the one by Fatsikostas andVerykios [42].

Based on these results, and although the higher reaction rate of thecatalyzed steps minimize the contribution of the thermal routes, the500–650 °C range has been selected in order to analyze the effect ofoperating variables on the catalyst performance (next section). Thecontribution of thermal routes is minimal within this range, and

therefore conclusions concerning activity, selectivity and stability of thecatalyst may be inferred.

3.2. Effect of operating conditions for 10Ni/La2O3-αAl2O3 catalyst

This section analyzes the effect operating conditions (temperature,space time and S/E molar ratio) have on the reaction indices at zerotime on stream, which have been obtained by extrapolating the resultsobtained in runs of 20 h duration to zero time.

3.2.1. Effect of temperatureThe effect that temperature has on the reaction indices is shown in

Figs. 3 and 4, in which the values are shown for ethanol conversion andH2 yield (Fig. 3), and carbon products yields (Fig. 4). The results cor-respond to S/E molar ratio of 3 (which is the stoichiometric ratio forSRE reaction) and for two values of space time, low (0.04 gcatalysth/gEtOH) and high (0.35 gcatalysth/gEtOH), with the aim of establishing theeffect of temperature under conditions far from equilibrium (kineticregime) and close to equilibrium (thermodynamic regime), respec-tively. Results for two values of space time are obtained in order to getinformation of interest for both optimizing reaction conditions (whichshould correspond to thermodynamic regime for a high throughput)and gaining knowledge on the intermediates and reaction mechanism,with the operating conditions corresponding in this case to the kineticregime, under which the extent of the individual reaction is limited. Itshould be pointed out, that neither acetone nor acetic acid is present inthe products stream in the experimental conditions studied. The ex-planation may lie in the fact that the reactions corresponding to theirformation/disappearing (Eqs. 5–8) are very rapid or do not occur. Infact, taking into account that these products are not observed in theblank runs either, it is evidenced that these reactions do not occur.

Under kinetic regime (space time of 0.04 gcatalysth/gEtOH, continuouslines), high ethanol conversion (0.6) is obtained at 500 °C, which pro-gressively increases with temperature and almost full conversion isachieved at 650 °C (Fig. 3). For a high space time (0.35 gcatalysth/gEtOH(dashed lines), which is close to thermodynamic regime) ethanol con-version is full for the whole temperature range studied. This effect oftemperature is consistent with that observed in the literature for dif-ferent catalysts and reaction conditions [43,45–49]. Thus, according toFasikostas et al. [45], who used 30%Ni/La2O3/Al2O3 catalyst, S/E = 3and a space time of 0.29 gcatalysth/gEtOH, conversion increases from 0.2at 600 °C to 0.92 at 800 °C, and is full at 850 °C. Melchor-Hernándezet al. [43] used 10%Ni/8%La2O3-γAl2O catalyst prepared by sol-gelmethod, S/E = 3 and space time ~0.07 gcatalysth/gEtOH and obtainedhigh ethanol conversion at low temperature (0.88 at 450 °C and 0.94 at500 °C) and total conversion at 600 °C, although H2 yield was ratherlow (0.04 at 400 °C and 0.50 at 600 °C). They also proved that this yieldwas slightly higher for a higher La2O3 content in the support. The in-crease in ethanol conversion with temperature reported by Llera et al.[46] is more pronounced (from 0.10 to 0.55 in the 600–650 °C tem-perature range) when they used Ni/Al2O3/LDH catalyst, S/E = 5.5 anda low space time, whereas Patel et al. [47] obtained 0.95 conversion at600 °C and full conversion at 700 °C with Ni/Ce2O3/Zr2O3 catalyst, fora high space time and S/E = 9. In the steam reforming of bioethanol(14 wt% ethanol) on 10%Ni/6%La2O3-Al2O3 catalyst, full ethanolconversion was achieved at 350 °C, which is partially attributed to thehigh S/E ratio (~15) [49].

Concerning the H2 yield obtained in this work (empty symbols inFig. 3), it increases with temperature (from 0.19 at 500 °C to 0.33 at650 °C) under kinetic regime conditions (low space time), although to alesser extent than ethanol conversion, especially above 600 °C. Underconditions with full ethanol conversion (high space time), H2 yield in-creases with temperature in the 500–650 °C range (from 0.40 to 0.73),but in a less pronounced way above 600 °C. It should be mentioned thatthese results correspond to S/E = 3, and that a higher H2 yield (0.88)was obtained for a higher S/E ratio (S/E = 9) at 600 °C.

0.00

0.25

0.50

0.75

1.00

500 550 600 650 700

x*EtOH

Temperature, ºC

a

0.00

0.25

0.50

0.75

1.00

500 550 600 650 700

x*H2

Temperature, ºC

S/E

3

6

9

b

Fig. 1. Molar fractions (dry basis) of ethanol (a) and H2 (b) obtained in runs withoutcatalyst at different temperatures and S/E molar ratios.

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

210

This evolution of H2 yield, which is different to the evolution ofethanol conversion, is explained by the different effect of temperatureon the reactions of the overall mechanism (Eqs. 1–16), which also in-volves a noticeable change in the yields of carbon products with tem-perature, as shown in Fig. 4. Under kinetic regime conditions (low

space time, 0.04 gcatalysth/gEtOH) (Fig. 4a), the order of the yields at500 °C is as follows: acetaldehyde (0.24) > CO2 (0.08) > CO(0.06) > CH4 (0.05) > C2H4 (0.01). It should be emphasized thatunder these conditions, H2/CO2 molar ratio (6.91) is noticeably higherthan the stoichiometric value (3) corresponding to ethanol reformingreaction (Eq. 1). These results are evidence that, at 500 °C, the fastestreaction in the kinetic scheme is the dehydrogenation to acetaldehyde(Eq. 2), which is noticeably faster than ethanol reforming (Eq. 1).However, the latter is faster than the decomposition reactions (of bothethanol and acetaldehyde, Eqs. 4 and 10), with the WGS reaction (Eq.13), methane reforming (Eqs. 14–15) and dehydration to ethylene (Eq.3) being apparently slow reactions, specially the last one. When tem-perature increases above 600 °C, acetaldehyde yield decreases notice-ably, (slightly lower than 0.1 at 650 °C), whereas the yield of the re-maining carbon products increases, especially that of ethylene, with thelower increase corresponding to methane. Consequently, H2/CO2 molarratio decreases asymptotically with temperature (to 5.3 at 650 °C).These results give evidence that an increase in temperature promotesmainly ethanol dehydration (Eq. 3), as well as ethanol and acet-aldehyde decomposition (Eqs. 4 and 10) and the reforming of acet-aldehyde (Eqs. 11 and 12) and CH4 (Eqs. 14–15).

For a high space time value (0.35 gcatalysth/gEtOH, Fig. 4b), that is,under conditions with total ethanol conversion and close to thermo-dynamic equilibrium, there is almost total absence of ethylene in thewhole temperature range studied, and acetaldehyde formation is notobserved above 550 °C. This result is consistent with the role bothcompounds play as intermediate products in the reaction scheme, andtherefore their transformation reactions (reforming of ethylene anddecomposition/reforming of acetaldehyde) are full under the condi-tions studied. Furthermore, methane yield decreases almost linearlywith temperature (from 0.45 at 500 °C to 0.13 at 650 °C) and a

0.00

0.05

0.10

0.15

0.20

500 550 600 650 700

x*C2H4O

Temperature, ºC

aS/E

3

6

9

0.00

0.05

0.10

0.15

0.20

500 550 600 650 700

x*C2=

Temperature, ºC

b

0.00

0.05

0.10

0.15

0.20

500 550 600 650 700

x*CO

Temperature, ºC

c S/E

3

6

9

0.00

0.05

0.10

0.15

0.20

500 550 600 650 700

x*CH4

Temperature, ºC

d

Fig. 2. Molar fractions (dry basis) of acetaldehyde (a), ethylene (b), CO (c) and CH4 (d) obtained in runs without catalyst at different temperatures and S/E molar ratios.

0

0.2

0.4

0.6

0.8

1

450 500 550 600 650 700

XEtOH

,

YH2

Temperature, ºC

XEtOH H2

Fig. 3. Effect of temperature on ethanol conversion and H2 yield in the SRE process overNi/La2O3–αAl2O3 catalyst with a space time of 0.04 gcatalysth/gEtOH (blue rhombus) and0.35 gcatalysth/gEtOH (black circles). S/E molar ratio = 3. (For interpretation of the re-ferences to colour in this figure legend, the reader is referred to the web version of thisarticle.)

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

211

significant increase in CO yield takes place at the same time (similarlyto that obtained for H2 yield under these conditions, Fig. 3), whichevidences an important contribution of methane steam reforming re-action (Eq. 14) to the overall reaction mechanism at high temperatures(because this reaction is highly endothermic), with a noticeable in-crease in H2 yield. The CO2 yield is high and almost constant (~0.46) inthe 500–600 °C range, and decreases slightly above 600 °C due to theshift in the thermodynamic equilibrium of the exothermic WGS reaction(the reverse WGS reaction is enhanced). The fact that CO2 yield remainsalmost constant in the 500–600 °C range, even though there is an in-crease in the reaction rate of steam reforming of oxygenates and hy-drocarbons, should be attributed to CO2 consumption in both the me-thanation reaction (reverse Eq. 15) and the subsequent CH4 reforming,as well as in the dry reforming of oxygenates and hydrocarbons, aspointed out by Fatsikostas et al. [45].

As previously mentioned, the properties of the support have animportant role in products distribution. Thus, Melchor-Hernández et al.(2013) obtained a maximum ethylene yield of ~0.35 in the 500–550 °Crange, with a subsequent decrease to ~0.20 at 600 °C, with the yieldbeing lower for a higher La content in the support [43]. This highethylene yield was a consequence of the slightly acidic support (γ-Al2O3), which is highly active for ethanol dehydration reaction. On thecontrary, with a Ni/SiO2 catalyst and under similar conditions to thoseused in this work, Vicente et al. [10] reported total absence of ethylenein the product stream due to the very low acidity of SiO2 support

hindering ethylene dehydration reaction. Moreover, with this supportwithout activity for ethanol dehydration, the competence of catalyticreactions with thermal reactions avoided the formation of ethyleneeven at high temperatures.

3.2.2. Effect of space timeFig. 5 shows the effect of space time on ethanol conversion and H2

yield for different temperatures, and Fig. 6 shows the effect of spacetime on the carbon products yields for two temperatures (500 °C(Fig. 6a) and 650 °C (Fig. 6b), which are the limits of the studied range).These results correspond to a S/E molar ratio of 3, and similar trendshave been observed for the ratios of 6 and 9.

Ethanol conversion increases in a pronounced way with space time(Fig. 5), although more moderately at high temperatures due to thesignificant contribution of thermal routes to the reaction mechanism.For all the studied temperatures, H2 yield (empty symbols in Fig. 5)increases in a pronounced way with space time in the0.02–0.09 gcatalysth/gEtOH range. A maximum is achieved for0.09 gcatalysth/gEtOH, when ethanol conversion is full, and below 550 °CH2 yield decreases with an increase in space time due to the relevanceof methanation reactions at these temperatures (reverse Eqs. 14–15),and tends towards thermodynamic equilibrium values [11]. Conse-quently, the yield of CH4 increases (Fig. 6a), although less noticeably astemperature is increased.

Similarly to those observed in Fig. 5 for ethanol conversion and H2

yield, two zones are also identified in the evolution with space time ofcarbon products yields (Fig. 6), although the range in which the effectof space time is noticeable depends on temperature, as follows: i) at500 °C (Fig. 6a) there is a noticeable effect of space time on the yields ofall carbon products in the 0.02–0.18 gcatalysth/gEtOH range; ii) at 650 °C(Fig. 6b) a significant variation in the yields of products takes placebelow 0.09 gcatalyst/hgEtOH (that is, before attaining full ethanol con-version). For both temperatures, acetaldehyde and ethylene are themain products at low space time values, specially the former, whichconfirms that ethanol dehydrogenation and dehydration reactions,especially the former, are the fastest reactions in the kinetic scheme.Moreover, at both temperatures, CO and CO2 yields increase sharply forlow space time values, and subsequently CO2 yield continues increasingasymptotically towards the thermodynamic equilibrium value [11],whereas CO yield, which peaks at 0.09 gcatalysth/gEtOH, decreases slowlytowards the corresponding thermodynamic equilibrium value [11]. At500 °C, the increase in CH4 yield is almost parallel to that of CO2, andprogressively attenuates with the increase in space time (Fig. 6a).Furthermore, CH4 formation is faster at 650 °C (Fig. 6b), because a

0

0.2

0.4

0.6

450 500 550 600 650 700

Yi

Temperature, ºC

CH4

CO

CO2

C2=

C2H

4O

a

0

0.002

0.004

0.006

0.008

0.01

0

0.2

0.4

0.6

450 500 550 600 650 700

Yi

Temperature, ºC

b

Fig. 4. Effect of temperature on carbon products yields in the SRE process over Ni/La2O3–αAl2O3 catalyst with a space time of 0.04 gcatalysth/gEtOH (a) and 0.35 gcatalysth/gEtOH (b). S/E molar ratio = 3.

0

0.2

0.4

0.6

0.8

1

0 0.1 0.2 0.3 0.4

XEtOH

,

YH2

space time, gcatalyst

h/gEtOH

500

550

600

650

XEtOH

H2

T,ºC

Fig. 5. Effect of space time on ethanol conversion and H2 yield in the SRE process overNi/La2O3–αAl2O3 catalyst at different temperatures. S/E molar ratio = 3.

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

212

considerable yield of 0.1 (twice that at 500 °C) is attained at the lowestspace time studied (0.04 gcatalysth/gEtOH), which confirms the im-portance of the thermal routes in ethanol and acetaldehyde decom-position reactions (Eqs. 4 and 10, respectively) at this temperature. Forhigher values of space time, CH4 yield keeps almost constant, which isexplained because its higher formation rate from ethanol decomposi-tion is balanced by its higher disappearance rate by steam reforming. Athigh temperatures, CH4 reforming thermodynamics is enhanced overmethanation reactions [10].

Similar trends in the effect of space time in ethanol conversion andproducts yields have been observed in the literature for other Ni basedcatalysts, although the range of space time depends on catalyst com-position and on the remaining reaction conditions. Thus, Llera et al.[46], obtained a conversion of 0.50 for 0.035 gcatalysth/gEtOH, and totalconversion for 0.14 gcatalysth/gEtOH, at 650 °C and S/E = 5.5, on a Ni/Al2O3/LDH catalyst. According to these authors, H2 yield increasedwith space time to 0.80 for 0.14 gcatalysth/gEtOH, and decreased slightlyfor higher space time. With a skeletal (non-supported) Ni-based cata-lyst, Zhang et al. [48] obtained total ethanol conversion for a space timeabove 0.15 gcatalysth/gEtOH, whereas H2 yield remained almost constant(0.60) in the 0.07–0.35 gcatalysth/gEtOH range, at 450 °C with S/E = 8.

An increase in ethanol conversion with space time was also observed byPatel et al. [47] on a Ni/Ce2O3/Zr2O3 catalyst, with values of up to 0.95for 0.60 gcatalysth/gEtOH, at 650 °C and with S/E = 4. According to theseauthors, H2 yield varied in the 0.70–0.80 range, which is slightly higherthan that shown in Fig. 5, probably due to the higher S/E ratio used. Fora commercial 15%Ni/Al2O3 catalyst used in the 200–600 °C range withS/E = 10, two zones were also identified by Wu et al. [50], below andabove a certain value of space time (corresponding to a contact time of1 s). Thus, the increase in CO2 yield with space time is less pronouncedin the second zone (> 1 s). Furthermore, CO yield peaks for a givenspace time at 600 °C, whereas CH4 yield (higher at low temperature)increased with space time in the first zone (< 1 s), and acetaldehydeyield decreased continuously with space time, especially at 600 °C.

3.2.3. Effect of steam/ethanol molar ratioThe effect of this variable has been studied in less detail than that of

temperature in the literature. Most of the research has been carried outwith S/E molar ratio = 3, which corresponds to the stoichiometricvalue [11–13,51]. The results with S/E values above 6 correspond tostudies on bio-ethanol steam reforming, whose characteristic ethanolcontent is around 14 wt%, and therefore S/E ratio is of around 15 [52].

Fig. 7 shows the evolution with S/E molar ratio (in the 3–9 range) ofethanol conversion and H2 yield at 600 °C for a space time of0.18 gcatalysth/gEtOH. For this high value of space time, ethanol con-version is full for all the S/E molar ratios studied, but a significant effectof this variable on H2 yield is observed, as predicted by thermo-dynamics [11]. Thus, H2 yield increases almost linearly by increasing S/E molar ratio, because the excess of water favors the reactions produ-cing H2, such as ethanol and methane reforming, and WGS reaction.This effect explains the evolution of products selectivity with S/E molarratio shown in Fig. 8, corresponding to the same temperature and spacetime as Fig. 7. As S/E ratio is increased there is an almost linear increasein the selectivities of H2 and CO2, although the latter to a lesser extent,whereas the selectivities to CO and CH4 progressively decrease, as aconsequence of the increase in the rates of WGS and CH4 reformingreactions.

Similar trends in carbon products selectivity by changing S/E ratiohave been also observed by Carrera-Cerritos et al. [18] and Li et al.[53]. According to Li et al. [53], an increase in CO2 yield of the sameorder as the decrease in CH4 yield is explained because the WGS re-action is enhanced to the same extent as the CO methanation reaction ishindered.

0

0.1

0.2

0.3

0.4

0.5

0 0.1 0.2 0.3 0.4

Yi

space time, gcatalyst

h/gEtOH

CH4

CO

CO2

C2=

C2H

4O

500 ºC

a

0

0.1

0.2

0.3

0.4

0.5

0 0.1 0.2 0.3 0.4

Yi

space time, gcatalyst

h/gEtOH

650 ºC

b

Fig. 6. Effect of space time on carbon products yield in the SRE process over Ni/La2O3–αAl2O3 catalyst at 500 °C (a) and 650 (b) °C. S/E molar ratio = 3.

0.2

0.4

0.6

0.8

1

0 3 6 9 12

XEtOH

,

YH2

S/E

XEtOH

H2

Fig. 7. Effect of S/E molar ratio on ethanol conversion and H2 yield in the SRE processover Ni/La2O3–αAl2O3 at 600 °C and space time of 0.18 gcatalysth/gEtOH.

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

213

The selectivities of acetaldehyde and ethylene have not been plottedin Fig. 8 because their presence has not been detected for this highvalue of space time. Under conditions they appear in the reactionmedium (space time below 0.09 gcatalysth/gEtOH), their selectivities de-crease with an increase in the S/E molar ratio.

In order to complement the previous results, corresponding to totalethanol conversion, the effect of S/E molar ratio on ethanol conversionand H2 yield has been studied with runs at low values of space time(0.04 gcatalysth/gEtOH). As observed in Fig. 9, ethanol conversion in the500–550 °C range (kinetic regime) is lower as the S/E molar ratio isincreased. This result is evidence that an excess of water over thestoichiometric value attenuates the global reaction mechanism forethanol steam reforming (Eqs. 1–16) due to the lower rate of some ofthe reactions in the kinetic scheme, such as ethanol dehydrogenationand dehydration, and ethanol and acetaldehyde decomposition. Thisattenuation should be taken into account in future studies for the de-velopment of a kinetic model for this process. In the 500–550 °C tem-perature range, H2 yield is slightly affected by an increase in the S/Emolar ratio because the drop in ethanol conversion is balanced by the

increase in H2 selectivity. A similar effect of S/E ratio was observed byCarrera-Cerritos et al. [18] on a 10Ni/La2O3-Al2O3 catalyst used in afixed-bed reactor, at 600 °C and for a space time of 0.10 gcatalysth/gEtOH.These authors obtained full conversion with S/E = 3, whereas theconversion decreased to 0.89 for S/E = 6, and H2 yield did not increaseas S/E ratio was increased.

Nevertheless, Fig. 9 shows that, in the 550–650 °C range, H2 yieldnoticeably increases when S/E ratio is increased because an increase inS/E molar ratio under these conditions (of high conversion) hardlyattenuates ethanol conversion.

Taking into account the above mentioned results concerning theeffect of operating conditions on products distribution obtained on aNi/La2O3-αAl2O3 catalyst, and their explanation based on the relevanceof the possible steps in the reaction mechanism (Eqs. 1–16), the kineticscheme plotted in Fig. 10 has been established. This scheme only con-siders the reactions that affect products distribution in the range ofreaction conditions studied, and is useful for quantifying this distribu-tion by means of a kinetic model. In this scheme, continuous arrowsdenote irreversible reactions, whereas dashed arrows denote reversiblereactions. Blue arrows indicate the reactions that are favored by anincrease in S/E molar ratio (enhancing H2 production), whereas blackarrows denote reactions attenuated by increasing S/E molar ratio (de-creasing ethanol conversion).

3.3. Catalyst stability

In order to test the stability of the catalyst, long duration runs(200 h time on stream) were performed under conditions suitable formaximizing H2 yield, which in view of the results in previous sectionsare as follows: 600–650 °C; space time, 0.35 gcatalysth/gEtOH, and; S/Emolar ratio, 6. A higher S/E ratio was not considered because only aslight increase in H2 yield would be achieved at the expense of a highercost involving steam generation and product separation (product molarfractions decrease due to dilution).

The evolutions of ethanol conversion and products yield under theseconditions are plotted in Fig. 11a (600 °C) and Fig. 11b (650 °C). Asobserved, intermediate products in the reaction scheme of the process(acetaldehyde and ethylene) are not formed, and the catalyst is highlystable. Thus, ethanol conversion is almost complete (> 0.97) after200 h time on stream, and there is also a very small variation in pro-ducts yields. The yield of the carbon products is different at bothtemperatures due to the opposite effect of temperature on the equili-brium of WGS and methane steam reforming reactions, as previouslycommented. Thus, the equilibrium of WGS reaction shifts to the left astemperature is higher, thereby increasing CO yield and decreasing CO2

yield, whereas methane steam reforming equilibrium is favored (higherCH4 yield). As a consequence of the opposite effect of temperature onthe previously mentioned reactions, H2 yield remains almost the sameat both temperatures. These results are consistent with those de-termined in a previous paper for the relationship between coke de-position and the concentration of ethanol and those of acetaldehydeand ethylene in the reaction medium, which are the precursors ofamorphous coke causing catalyst deactivation by blocking Ni sites [41].

4. Conclusions

The effect operating conditions have on products distribution in theSRE over Ni/La2O3-αAl2O3 catalysts is complex, as it is a consequenceof numerous parallel reactions that are activated by the catalyst, anddue to the contribution of thermal (non-catalytic) routes. The kineticscheme proposed explains the effect operating conditions (temperature,space time and steam/ethanol molar ratio) have on product distribution(H2, CO, CO2, CH4, ethylene, acetaldehyde), based on the extent of theindividual reaction steps.

The thermal routes contribute significantly to the extent of somereactions, and therefore to products distribution, i.e., below 600 °C by

0

0.2

0.4

0.6

0.8

0 3 6 9 12

SI

S/E

H2

CO2

CH4

CO

Fig. 8. Effect of S/E molar ratio on carbon products selectivity in the SRE process overNi/La2O3–αAl2O3 at 600 °C and space time of 0.18 gcatalysth/gEtOH.

0

0.2

0.4

0.6

0.8

1

450 500 550 600 650 700

XEtOH

,

YH2

Temperature, ºC

XEtOH

H2

S/E

3

6

Fig. 9. Effect of temperature on ethanol conversion and H2 yield in the SRE process overNi/La2O3–αAl2O3 for two different S/E molar ratios. Space time of 0.04 gcatalysth/gEtOH.

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

214

means of ethanol dehydrogenation and dehydration and above 600 °Cby ethanol and acetaldehyde decomposition.

In the catalytic steam reforming over Ni/La2O3-αAl2O3, bothethanol conversion and H2 yield increase considerably with tempera-ture in the 500–650 °C range, although the increase in H2 yield at-tenuates above 600 °C because the WGS reaction equilibrium shiftstowards the left at high temperature. Furthermore, a space time above

0.09 gcatalysth/gEtOH leads to almost full ethanol conversion and pro-ducts yields do not vary significantly, especially above 600 °C, as theyapproach values corresponding to thermodynamic equilibrium. Theeffect of space time is less noticeable at high temperature (650 °C) dueto the significant contribution of the thermal routes to the reactionmechanism. The effect steam/ethanol (S/E) molar ratio has on con-version and product selectivity is complex. Under conditions of highconversion (high values of temperature and space time), an increase inS/E ratio above the stoichiometric value (3) enhances H2 selectivity dueto the promotion of ethanol and methane steam reforming and WGSreaction, whereas the selectivity to by-products (CO, CH4, ethylene andacetaldehyde) decreases. Nevertheless, under conditions far fromthermodynamic equilibrium, ethanol conversion decreases by in-creasing the S/E ratio above the stoichiometric value due to the at-tenuation in the rate of some reactions in the kinetic scheme.

In order to maximize H2 yield, while minimizing by-product for-mation and enhancing catalyst stability, the following operating con-ditions are recommended: 600 °C, space time above 0.35 gcatalysth/gEtOHand S/E molar ratio = 6. Under these conditions, a H2 yield of 82% isachieved, which remains constant along 200 h time on stream.

Acknowledgments

This work has been carried out with financial support from theMinistry of Science and Technology of the Spanish Government (ProjectCTQ2012-35263) and the ERDF Funds (Project CTQ2015-68883-R), theUniversity of the Basque Country (Project UFI 11/39) and the BasqueGovernment (Project IT748-13). Carolina Montero is grateful with forher Ph.D grant from National Secretariat of Higher Education, Science,Technology and Innovation of Ecuador SENESCYT (Contract20110560).

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.fuproc.2017.10.003.

References

[1] International Energy Agency, World energy outlook 2016, https:www.aei.org/newsroom/news/2016/november/world-energy-outlook-2016.html.

[2] P. Lanzafame, G. Centi, S. Perathoner, Evolving scenarios for biorefineries and theimpact on catalysis, Catal. Today 234 (2014) 2–12.

[3] P. Mohanty, K.K. Pant, R. Mittal, R., Hydrogen generation from biomass materials:challenges and opportunities, in: P.D. Lund, J. Byrne, G. Berndes, I.A. Vasalos

HYDROGEN

ETHANOL

EthyleneAcetaldehyde

MethaneCO

CO2

Dehydration (Eq. 3)

Decomposition (Eq. 10)

SR (Eq.16)SR (Eqs. 12-13)

Dehydrogenation (Eq. 2)

Decomposition (Eq. 4)

SR/Methanation (Eq. 14)

SR (Eq. 1)

SR/Methanation(Eqs. 14-15)

WGS/r-WGS (Eq. 13)

Fig. 10. Kinetic scheme proposed for the SRE process overNi/La2O3–αAl2O3 catalyst.

0.0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200

XE

tO

H, Y

i

time on stream, h

a

S/E=6

600 °C

XEtOH

H2

CO2

CH4

CO

0.0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200

XE

tO

H, Y

i

time on stream, h

S/E=6

650 °C

b

Fig. 11. Evolution with time on stream of ethanol conversion and products yield in theSRE process over Ni/La2O3–αAl2O3 at 600 °C (a) and 650 °C (b) for 0.35 gcatalysth/gEtOHand S/E molar ratio = 6.

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

215

(Eds.), Advances in Bioenergy: The Sustainability Challenge, John Wiley & Sons,Ltd, Oxford, 2016, pp. 93–108.

[4] A. Hedayati, O. Le Corre, B. Lacarrière, J. Llorca, Dynamic simulation of pure hy-drogen production via ethanol steam reforming in a catalytic membrane reactor,Energy 117 (Part 2) (2016) 316–324.

[5] D. Li, X. Li, J. Gong, Catalytic reforming of oxygenates: state of the art and futureprospects, Chem. Rev. 116 (2016) 11529–11653.

[6] H.B. Aditiya, T.M.I. Mahlia, W.T. Chong, H. Nur, A.H. Sebayang, Second generationbioethanol production: a critical review, Renew. Sust. Energ. Rev. 66 (2016)631–653.

[7] Y. Sharma, A. Kumar, R. Prasad, S. Upadhyay, Ethanol steam reforming for hy-drogen production: latest and effective catalyst modification strategies to minimizecarbonaceous deactivation, Renew. Sust. Energ. Rev. 74 (2017) 89–103.

[8] S. Sun, W. Yan, P. Sun, J. Chen, Thermodynamic analysis of ethanol reforming forhydrogen production, Energy 44 (2012) 911–924.

[9] M. Ni, D. Leung, M. Leung, A review on reforming bio-ethanol for hydrogen pro-duction, Int. J. Hydrog. Energy 32 (2007) 3238–3247.

[10] J. Vicente, J. Ereña, C. Montero, M.J. Azkoiti, J. Bilbao, A.G. Gayubo, Reactionpathway for ethanol steam reforming on a Ni/SiO2 catalyst including coke forma-tion, Int. J. Hydrog. Energy 39 (2014) 18820–18834.

[11] C. Montero, L. Oar-Arteta, A. Remiro, A. Arandia, J. Bilbao, A.G. Gayubo,Thermodynamic comparison between bio-oil and ethanol steam reforming, Int. J.Hydrog. Energy 40 (2015) 15963–15971.

[12] G. Nahar, V. Dupont, Hydrogen via steam reforming of liquid biofeedstock, Biofuels3 (2012) 167–191.

[13] J.L. Contreras, J. Salmones, J.A. Colín-Luna, L. Nuño, B. Quintana, I. Córdova, et al.,Catalysts for H2 production using the ethanol steam reforming (a review), Int. J.Hydrog. Energy 39 (2014) 18835–18853.

[14] A. Kumar, R. Prasad, Y. Sharma, Steam reforming of ethanol: production of re-newable hydrogen, Int. J. Environ. Res. Dev. 3 (2014) 203–212.

[15] T. Hou, S. Zhang, Y. Chen, D. Wang, W. Cai, Hydrogen production from ethanolreforming: catalysts and reaction mechanism, Renew. Sust. Energ. Rev. 44 (2015)132–148.

[16] P. Osorio-Vargas, C.H. Campos, R.M. Navarro, J.L.G. Fierro, P. Reyes, Improvedethanol steam reforming on Rh/Al2O3 catalysts doped with CeO2 or/and La2O3:influence in reaction pathways including coke formation, Appl. Catal. A Gen. 505(2015) 159–172.

[17] P.K. Sharma, N. Saxena, P.K. Roy, A. Bhatt, Hydrogen generation by ethanol steamreforming over Rh/Al2O3 and Rh/CeZrO2 catalysts: a comparative study, Int. J.Hydrog. Energy 41 (2016) 6123–6133.

[18] R. Carrera Cerritos, R. Fuentes Ramírez, A. Aguilera Alvarado, J. Martínez Rosales,T. Viveros García, I. Galindo Esquivel, Steam reforming of ethanol over Ni/Al2O3-La2O3 catalysts synthesized by sol-gel, Ind. Eng. Chem. Res. 50 (2011) 2576–2584.

[19] R. Trane-Restrup, S. Dahl, A.D. Jensen, Steam reforming of ethanol: effects ofsupport and additives on Ni-based catalysts, Int. J. Hydrog. Energy 38 (2013)15105–15118.

[20] A.F. Lucredio, J.D.A. Bellido, A. Zawadzki, E.M. Assaf, Co catalysts supported onSiO2 and γ-Al2O3 applied to ethanol steam reforming: effect of the solvent used inthe catalyst preparation method, Fuel 90 (2011) 1424–1430.

[21] X. Zhao, G. Lu, Modulating and controlling active species dispersion over Ni-Cobimetallic catalysts for enhancement of hydrogen production of ethanol steam re-forming, Int. J. Hydrog. Energy 41 (2016) 3349–3362.

[22] A.J. Vizcaino, A. Carrero, J.A. Calles, Comparison of ethanol steam reforming usingCo and Ni catalysts supported on SBA-15 modified by Ca and Mg, Fuel Process.Technol. 146 (2016) 99–109.

[23] F. Carvalho, Y. Asencios, J. Bellido, E. Assaf, Bio-ethanol steam reforming for hy-drogen production over Co3O4/CeO2 catalysts synthesized by one-step poly-merization method, Fuel Process. Technol. 142 (2016) 182–191.

[24] S.J. Han, Y. Bang, J. Yoo, J.G. SeoI, K. Song, Hydrogen production by steam re-forming of ethanol over mesoporous Ni–Al2O3–ZrO2 xerogel catalysts: effect ofnickel content, Int. J. Hydrog. Energy 38 (2013) 8285–8292.

[25] A.G. Gayubo, J. Vicente, J. Ereña, C. Montero, M. Olazar, J. Bilbao, Comparison ofNi and Co catalysts for ethanol steam reformingin a fluidized bed reactor, Catal.Lett. 144 (2014) 1134–1143.

[26] C.K.S. Choong, Z. Zhong, L. Huang, Z. Wang, T.P. Ang, A. Borgna, et al., Effect ofcalcium addition on catalytic ethanol steam reforming of Ni/Al2O3: I. Catalyticstability, electronic properties and coking mechanism, Appl. Catal. A Gen. 407(2011) 145–154.

[27] K. Elias, A. Lucrédio, E. Assaf, Effect of CaO addition on acid properties of Ni–Ca/Al2O3 catalysts applied to ethanol steam reforming, Int. J. Hydrog. Energy 38(2013) 4407–4417.

[28] M.N. Barroso, A.E. Galetti, M.F. Gomez, L.A. Arrúa, M.C. Abello, Ni-catalysts sup-ported on ZnxMg1−xAl2O4 for ethanol steam reforming: influence of the substitu-tion for Mg on catalytic activity and stability, Chem. Eng. J. 222 (2013) 142–149.

[29] G.P. Szijjártó, Z. Pászti, I. Sajó, A. Erdőhelyi, G. Radnóczi, A. Tompos, Nature of theactive sites in Ni/MgAl2O4-based catalysts designed for steam reforming of ethanol,J. Catal. 305 (2013) 290–306.

[30] G. Zeng, R. Gu, Y. Li, The preparation and catalytic behavior of a shell–core Ni/Mg–Al catalyst for ethanol steam reforming, Int. J. Hydrog. Energy 38 (2013)11256–11267.

[31] V. Nichele, M. Signoretto, F. Pinna, F. Menegazzo, I. Rossetti, G. Cruciani, et al., Ni/ZrO2 catalysts in ethanol steam reforming: inhibition of coke formation by CaO-doping, Appl. Catal. B Environ. 150–151 (2014) 12–20.

[32] M.C. Sánchez-Sánchez, R.M. Navarro, J.L.G. Fierro, Ethanol steam reforming overNi/La–Al2O3 catalysts: influence of lanthanum loading, Catal. Today 129 (2007)336–345.

[33] J. Vicente, C. Montero, J. Ereña, M.J. Azkoiti, J. Bilbao, A.G. Gayubo, Coke deac-tivation of Ni and Co catalysts in ethanol steam reforming at mild temperatures in afluidized bed reactor, Int. J. Hydrog. Energy 39 (2014) 12586–12596.

[34] H. Ma, L. Zeng, H. Tian, D. Li, X. Wang, X. Li, et al., Efficient hydrogen productionfrom ethanol steam reforming over La-modified ordered mesoporous Ni-basedcatalysts, Appl. Catal. B Environ. 181 (2016) 321–331.

[35] C. Montero, A. Remiro, A. Arandia, P.L. Benito, J. Bilbao, A.G. Gayubo,Reproducible performance of a Ni/La2O3-αAl2O3 catalyst in ethanol steam re-forming under reaction-regeneration cycles, Fuel Process. Technol. 152 (2016)215–222.

[36] A. Remiro, B. Valle, A.T. Aguayo, J. Bilbao, A.G. Gayubo, Operating conditions forattenuating Ni/La2O3–αAl2O3 catalyst deactivation in the steam reforming of bio-oil aqueous fraction, Fuel Process. Technol. 115 (2013) 222–232.

[37] J. Mazumder, H.I. de Lasa, Fluidizable La2O3 promoted Ni/γ-Al2O3 catalyst forsteam gasification of biomass: effect of catalyst preparation conditions, Appl. Catal.B Environ. 168–169 (2015) 250–265.

[38] M. Sugisawa, K. Takanabe, M. Harada, J. Kubota, K. Domen, Effects of La additionto Ni/Al2O3 catalysts on rates and carbon deposition during steam reforming of n-dodecane, Fuel Process. Technol. 92 (2011) 21–25.

[39] C. Li, Y. Chen, Temperature-programmed-reduction studies of nickel oxide/aluminacatalysts: effects of the preparation method, Thermochim. Acta 256 (1995)457–465.

[40] B. Valle, B. Aramburu, A. Remiro, J. Bilbao, A.G. Gayubo, Effect of calcination/reduction conditions of Ni/La2O3–αAl2O3 catalyst on its activity and stability forhydrogen production by steam reforming of raw bio-oil/ethanol, Appl. Catal. BEnviron. 147 (2014) 402–410.

[41] C. Montero, A. Ochoa, P. Castaño, J. Bilbao, A.G. Gayubo, Monitoring Ni0 and cokeevolution during the deactivation of a Ni/La2O3–αAl2O3 catalyst in ethanol steamreforming in a fluidized bed, J. Catal. 331 (2015) 181–192.

[42] A.N. Fatsikostas, X.E. Verykios, Reaction network of steam reforming of ethanolover Ni-based catalysts, J. Catal. 225 (2004) 439–452.

[43] C. Melchor-Hernández, A. Gómez-Cortés, G. Díaz, Hydrogen production by steamreforming of ethanol over nickel supported on La-modified alumina catalysts pre-pared by sol-gel, Fuel 107 (2013) 828–835.

[44] L. Barattini, G. Ramis, C. Resini, G. Busca, M. Sisani, U. Costantino, Reaction path ofethanol and acetic acid steam reforming over Ni–Zn–Al catalysts. Flow reactorstudies, Chem. Eng. J. 153 (2009) 43–49.

[45] A.N. Fatsikostas, D.I. Kondarides, X.E. Verykios, Production of hydrogen for fuelcells by reformation of biomass-derived ethanol, Catal. Today 75 (2002) 145–155.

[46] I. Llera, V. Mas, M.L. Bergamini, M. Laborde, N. Amadeo, Bio-ethanol steam re-forming on Ni based catalyst. Kinetic study, Chem. Eng. Sci. 71 (2012) 356–366.

[47] M. Patel, T. Jindal, K. Pant, Kinetic study of steam reforming of ethanol on Ni-basedceria-zirconia catalyst, Ind. Eng. Chem. Res. 52 (2013) 15763–15771.

[48] C. Zhang, S. Li, G. Wu, Z. Huang, Z. Han, T. Wang, et al., Steam reforming of ethanolover skeletal Ni-based catalysts: a temperature programmed desorption and kineticstudy, AICHE J. 60 (2014) 635–644.

[49] M. Dan, L. Senila, M. Roman, M. Mihet, M.D. Lazar, From wood wastes to hydrogen– preparation and catalytic steam reforming of crude bio-ethanol obtained from firewood, Renew. Energy 74 (2015) 27–36.

[50] Y. Wu, J.C. Santos, P. Li, J.-. Yu, A.F. Cunha, A.E. Rodrigues, Simplified kineticmodel for steam reforming of ethanol on a Ni/Al2O3 catalyst, Can. J. Chem. Eng. 92(2014) 116–130.

[51] P.D. Vaidya, A.E. Rodrigues, Insight into steam reforming of ethanol to producehydrogen for fuel cells, Chem. Eng. J. 117 (2006) 39–49.

[52] B. Banach, A. Machocki, P. Rybak, A. Denis, W. Grzegorczyk, W. Gac, Selectiveproduction of hydrogen by steam reforming of bio-ethanol, Catal. Today 176 (2011)28–35.

[53] S. Li, M. Li, C. Zhang, S. Wang, X. Ma, J. Gong, Steam reforming of ethanol over Ni/ZrO2 catalysts: effect of support on product distribution, Int. J. Hydrog. Energy 37(2012) 2940–2949.

C. Montero et al. Fuel Processing Technology 169 (2018) 207–216

216