Bioenergy systems for the future 9 h2 production from bioalcohols and biomethane steam reforming in membrane reactors

Bioenergy systems for the future 9 h2 production from bioalcohols and biomethane steam reforming in membrane reactors Bioenergy systems for the future 9 h2 production from bioalcohols and biomethane steam reforming in membrane reactors Bioenergy systems for the future 9 h2 production from bioalcohols and biomethane steam reforming in membrane reactors Bioenergy systems for the future 9 h2 production from bioalcohols and biomethane steam reforming in membrane reactors Bioenergy systems for the future 9 h2 production from bioalcohols and biomethane steam reforming in membrane reactors

and biomethane steam reforming

in membrane reactors

A Iulianelli*, F Dalena*, A Basile†

*University of Calabria, Rende, Italy,†Institute on Membrane Technology (ITM-CNR),Rende, Italy

MW molecular weight of diffusing gas

n dependence factor of the hydrogen flux on the hydrogen partial pressure

p pressure

PEMFC proton exchange membrane fuel cell

pH2 hydrogen partial pressure

pH2
perm hydrogen partial pressures in the permeate side

pH2
ret hydrogen partial pressures in the retentate side

PSS porous stainless steel

R universal gas constant

Bioenergy Systems for the Future http://dx.doi.org/10.1016/B978-0-08-101031-0.00009-0

© 2017 Elsevier Ltd All rights reserved.

In the meanwhile, conventional techniques used for producing hydrogen need to bestrongly revised and redesigned to match the requirements of a sustainable develop-ment Therefore, the introduction of alternative systems coupled to the exploitation ofrenewable sources such as the membrane reactors (MRs) has fostered the production

of high-grade hydrogen Furthermore, the International Energy Agency has recentlypublished the technology road map toward hydrogen and fuel cells, pointing out thatthe world economy is still based on the derived of fossil-fuel utilization and thatrenewable sources exploitation is strictly required to move on a world hydrogen-basedenergy system (Van der Hoeven, 2015) Proton exchange membrane fuel cells(PEMFCs) convert efficiently the chemical energy into electricity via indirect reaction

of hydrogen and oxygen and, furthermore, are zero-pollutant emission devices (Bose

et al., 2011) Low-temperature PEMFCs normally work at temperature lower than

100°C and tolerate up to 10 ppm of CO in the high-grade hydrogen stream supplied,owing to the CO poisoning effect on the anodic Pt-based catalyst Nowadays,

Fig 9.1 Prevision about hydrogen utilization in a large-scale sustainable energy system.Reproduced with permission from Dunn, S., 2002 Hydrogen futures: toward a sustainableenergy system Int J Hydrogen Energy 27, 235-264

hydrogen is industrially produced by steam reforming of natural gas in conventionalreactors (CRs) Hence, the reformed stream coming out from the CR contains besideshydrogen also such by-products as CO, CH4, and CO2 As a consequence, thishydrogen-rich stream needs to be purified before supplying a PEMFC, and conven-tionally, this takes place via a multistage process (Iulianelli et al., 2012) As an alter-native solution to the aforementioned multistage process, the application of the MRtechnology constitutes an interesting option for intensifying the whole process, bycombining the reforming reaction for generating hydrogen and its purification in asingle stage (Basile et al., 2016; Lu et al., 2007) Nevertheless, only a fewindustrial-scale MR applications can be noticed, and one of them is represented bythe Tokyo Gas Company Ltd (E-net [1]) More in details, the most significant aspect

of a MR is represented by the presence of the membrane in the reaction system, whichmakes possible to display the thermodynamic equilibrium conversion of the equiva-lent CR (in an equilibrium-restricted reaction) due to the selective removal of a prod-uct from the reaction system (i.e., hydrogen), promoting the so-called shift effect onthe reaction itself, which proceeds with a higher product formation and consequentconversion enhancement However, MR technology can be applied in various fields,each of them depending on which kind of process itself has to be performed and, con-sequently, on which kind of membrane has to be considered as best solution There-fore, the MRs can be classified by taking into account the role of the membranestoward the removal/addition of various chemical species and, referring to their mate-rials or structures, giving particular relevance to the selectivity of the permeation ofsuch products with respect to other ones Hence, it is possible to summarize the dif-ferent classes of MRs in

(1) dense and porous inorganic MRs (Iulianelli et al., 2012; Lu et al., 2007; Lin, 2001),(2) polymeric MRs (Bose et al., 2011; Gu et al., 2016),

(3) zeolite MRs (Wang and Yan, 2015),

(4) photocatalytic MRs (Mozia, 2010),

(5) enzyme MRs (Uragami, 2011),

(6) membrane bioreactors (Lebrero et al., 2013; Calabro`, 2013),

(7) electrochemical MRs (fuel cells, electrolytic cells, etc.) (Datta et al., 2015; Chatenet

et al., 2010)

In particular, the purposes of this chapter are related to the MR technology application

to biofuel reforming reaction for producing hydrogen, with particular relevance to theapplication of Pd-based membranes

A real interest toward the application of MR technology was noticed when innovativeinorganic membrane materials and development of high-temperature membraneprocesses took place (Basile et al., 2013; Menendez, 2011) Industrially, variousheterogeneous gas-solid catalytic processes, conventionally carried out in fixed-,fluidized-, or trickle-bed reactors, are mostly performed at high temperatures and

H 2 production from bioalcohols and biomethane steam reforming in membrane reactors 323

in chemically harsh ambient Consequently, inorganic membranes are favored rial with respect to the polymeric ones for their utilization in MRs It is well recog-nized that the separation behavior of an inorganic membrane housed in an MRgives several benefits to enhance the performance of a catalytic system Here, we pre-sent the advantages due to the utilization of inorganic MRs to integrate a chemicalreaction involving a biosource as a reactant (e.g., ethanol, methanol, glycerol, biogas,and acetic acid) with a membrane process (i.e., H2separation).

mate-Thus, by taking into account the crucial role of the inorganic membrane, it is useful

to subdivide them as generically resumed below:

l Macroporous membranes, with a pore size>50 nm

l Mesoporous membranes, with a pore size between 2 and 50 nm

l Microporous membranes, with smaller pore size of 2 nm

l Dense membranes, with pore size<0.5 nm

Regarding the mechanism regulating the gas transport through the inorganic branes, it is worth noting that for dense membranes, the transport mechanism is rep-resented by solution-diffusion, while for porous membranes, different mechanismsoften compete with each other to control the process Among the most common ofthem used for describing a gas permeation process through porous membranes, it ispossible to consider mainly those reported below:

mem-l Poiseuille mechanism, which occurs in case of the average pore diameter results to be muchlarger than the mean free path of the molecules; consequently, the collisions within the var-ious molecules are more frequent than those within molecules and porous walls:

Ji¼
s  d

2 pore

where ε¼membrane void fraction, d2

pore¼pore diameter, R¼universal constant,

T ¼ temperature, p ¼ pressure, τ ¼ tortuosity rp ¼ pressure gradient, and η ¼ viscosity

l Knudsen diffusion mechanism, which occurs in case of the pore diameter results to becomparable or less than the mean free path; the quantum momentum is transferred bythe collisions between the molecules and the wall of the pores Applying the kinetic theory

of gases to a single straight and cylindrical pore, the Knudsen diffusion coefficient can bedefined as

where ε¼membrane void fraction, dpore¼pore diameter, R¼universal constant,

T ¼ temperature, τ ¼ tortuosity, η ¼ viscosity, and MW ¼ molecular weight

On the contrary, metallic membranes have been deeply studied due to the teristics of high hydrogen permselectivity of dense metal walls, particularly appeal-ing in the field of hydrogen separation and purification Among them, Pd and itsalloys have been extensively studied because of its potentiality as a useful materialfor membrane fabrication (Al-Mufachi et al., 2015; Yun and Oyama, 2011)

Nevertheless, as high the hydrogen permselectivity over all of the other gases as lowthe permeability (and vice versa), while the cost of the membranes strictly depends

on the thickness of membrane material (Pd and Pd-alloy) As a solution to this ical issue, composite Pd-based membranes (consisting of thin metal films coatedover porous supports) have been particularly studied because they exhibit highhydrogen permeability and selectivity values depending on the Pd-alloy layer cov-ering the porous support (Li et al., 2016)

crit-Then, taking into account that MR technology has been and is particularly used inhydrogen production from the reforming of hydrocarbons and/or alcohols, the utiliza-tion of dense and thin-wall Pd-based membranes showing full hydrogen per-mselectivity should allow both high-grade hydrogen stream and hydrogen recovery

as well Therefore, in the last ten years, an extensive literature has been addressed

to hydrogen production using both self-supported and supported Pd-based MRs,highlighting their benefits when applied in reforming processes over the CRs asshortly reported here below (Itoh, 2012; Basile et al., 2011b; Mallada andMen
endez, 2008; Basile, 2008; Dolan et al., 2006; Me
enendez, 2011; Adhikari andFernand, 2006; Yan et al., 2011; Zornoza et al., 2015):

1 Intensified process in which the chemical reaction for producing hydrogen and its tion/purification are performed in only one process device

separa-2 Higher conversions than CRs (exercised at the same MR operating conditions) or, at milderconditions, the same conversion of CRs

3 Removal of high-grade hydrogen permeated stream

4 Improvement of both hydrogen yield and selectivity

The transport mechanism of the hydrogen permeation through a dense palladium (orits alloy) wall (Fig 9.2) is represented by the solution-diffusion one, which occursspecifically in six steps as reported below:

1 H2molecule adsorption from the membrane side at higher H2partial pressure

2 Dissociation of H2molecules on the surface

3 Reversible dissociative chemisorption of atomic H2.

4 Reversible dissolution of atomic H2in the metal lattice of the membrane

5 Diffusion into the metal of atomic H2proceeds from the side of the membrane at a higher H2pressure to the side at lower pressure

6 Desorption of recombined atomic H2into molecular form

From a theoretical point of view, the solution-diffusion mechanism evolves in threetypes of fluxes:

J3¼ D Cð 2
C1Þ (9.5)which represent the H2adsorption on the membrane side at higher partial pressure, thedissociation into atomic H2and reversible dissociative chemisorption of atomic H2,and the final desorption of recombined H2molecules, respectively At steady-stateconditions, the aforementioned fluxes are equal ðJH 2¼ J1¼ J2¼ J3Þ, and by adding

whereαdiffis a diffusion coefficient describing the relationship between the resistance

of H2transport on the membrane surface and the H2dissociation into the metal lattice:

Consequently, the H2permeating flux can be expressed by the following generalequation:

to 0.5, the transport resistance is represented by the H2dissociation into the Pd layer,and then, Eq.(9.7)becomes the Sieverts-Fick law (see Eq.9.10):

Al2O3, and B2O3, or porous stainless steel (PSS) PSS supports have a thermal sion coefficient close to that of palladium (i.e., PSS supports) and, then, allow highmechanical durability and simplify the gas sealing, but at relatively high temperatures,they alloy with palladium, leading to the decrease of hydrogen permeability(Li et al., 2016)

expan-Nevertheless, composite membranes allow the reduction of the palladium contentand the cost as well, favoring meanwhile a more consistent mechanical resistance Inthis scientific area, Yun and Ted Oyama stated that a composite Pd-based membranesupported on Al2O3results to be more effective than the utilization of other poroussupports in terms of membrane cost and hydrogen permeation characteristics (Yunand Oyama, 2011) This is more emphasized in the case of dense palladium orpalladium-alloy layer ranging between 2.5 and 5μm of thickness, for which these

H 2 production from bioalcohols and biomethane steam reforming in membrane reactors 327

membranes can represent a concrete solution at larger scale for achieving a effective and high hydrogen permeable and permselective membrane operation.Another important aspect to be considered regarding the supported Pd-based mem-branes is the coefficient “n” in the Eq.(9.1), which is strictly correlated to the hydro-gen permeation characteristics of the composite membrane As stated above, thiscoefficient can vary between 0.5 (typical value for expressing the Sieverts-Ficklaw, regulating fully hydrogen permselective Pd membranes) and 1.0 (low hydrogenpermselectivity) In case of supported membranes showing a thickness of the dense Pd

cost-or Pd-alloy layer around 5μm, the values of the coefficient “n” can describe the anism responsible of the hydrogen permeation through the membrane

mech-For example,Fig 9.3shows the graphic calculation of the coefficient “n” for aPd/Al2O3membrane (dense Pd layer of 5μm) (Iulianelli et al., 2015), and the value

ofn ¼ 0.6 associated to the highest linear regression factor (R2) means that the controlling step is in the intermediate range within the mass transfer (n ¼ 1.0) and thebulk diffusion (n ¼ 0.5) Apart from the extremities of this range, it is possible to findseveral exponent “n” describing an intermediate behavior during the hydrogen perme-ation process Obviously, “n” values closer to 0.5 describe more hydrogenpermselective membranes than values closer to 1.0

rate-Fig 9.3 Graphic calculation of coefficient “n” from the linear regression factor (R2) for aPd/Al2O3membrane with a thin Pd layer of 5μm (adapted from Iulianelli, A., Liguori, S.,Huang, Y., Basile, A., 2015 Model biogas steam reforming in a thin Pd-supported membranereactor to generate clean hydrogen for fuel cells, J Power Sou 273, 25-32), merged to the rate-controlling step for the hydrogen permeation through the membrane as a function of exponent

“n” versus top layer thickness (adapted from Yun, S., Oyama, S T., 2011 Correlations inpalladium membranes for hydrogen separation: A review, J Membrane Sci 375, 28–45)

9.3 Hydrogen production in MRs from bio-alcohols

reforming

As reported in the Introduction section of this chapter, in the last decade, MR nology has been particularly utilized in reforming reactions of biofuels to generate

tech-as much tech-as pure hydrogen Among the biofeedstocks useful tech-as reactants, bioalcoholssuch as ethanol emerged as a more competitive source for generating hydrogen sinceproducible renewably from biomass Furthermore, ethanol results to be a more com-petitive renewable biofeedstock because it is less toxic than alcohols such as metha-nol, glycerol, or other emergent biosources such as acetic acid and diethyl ether(Iulianelli and Basile, 2011) Ethanol and/or bioethanol steam reforming (ESR and/

or BESR) reactions have been extensively studied in CRs to produce hydrogen From

a stoichiometric point of view, hydrogen generation from ESR reaction is represented

by Eq.(9.12)as reported below:

C2H5OH + 3H2O¼ 2CO2+ 6H2 ΔH°

Besides hydrogen, other undesirable by-products can be produced owing to a complexreaction system, which depends on the catalyst utilized to carry out ESR reaction.Consequently, an ESR reformed stream coming out from a CR contains not onlyhydrogen but also CO2, CO, and CH4, besides acetaldehyde, ethylene, ethane, etc

In this chapter, one of the main intents is to give a short overview on ESR reactioncarried out in MRs, describing their superior performance owing to the special char-acteristics of Pd-based membranes in terms of hydrogen permselectivity Therefore,Table 9.1summarizes operating conditions and membrane characteristics of Pd-basedmembranes housed in the MRs, besides the most significant experimental perfor-mance in terms of conversion, hydrogen recovery, and hydrogen permeate purity

In the upper part ofTable 9.1, the most recent results about ESR and BESR reaction

in supported Pd-based MRs are reported, while the downer part refers to the literaturedata involving unsupported Pd-based MRs, and the restant part is related to notPd-based MRs applications Then, according to the need of decreasing the Pd content

in the membranes to favor lower membrane cost, several results show that a fewmicrons of dense Pd layer make almost complete ethanol conversion feasible withthe further advantage of collecting as much as pure hydrogen with recovery rangingfrom 40% to 70%

Concerning the supported Pd-based MRs,Ma et al (2016)reached complete anol conversion at 500°C, recovering a hydrogen stream with a purity of 99.9% Theyhoused inside the MR a supported membrane in which a first layer of palladium andgold was doubled on PSS support.Yun et al (2012)prepared an ultrathin supported Pdmembrane (2μm thick) allocated in a MR to perform ESR reaction In this work, highfeed molar ratio, moderate reaction temperature (360°C) and ambient pressureallowed to reach more than 70% ethanol conversion Other interesting results wereachieved by Espinal et al (2014), who used a thin Pd-Ag layer (4/5μm) deposited

eth-H 2 production from bioalcohols and biomethane steam reforming in membrane reactors 329

Table 9.1 Experimental data from literature about ethanol reforming reactions in MRs

Membrane

typology in

Palladiumlayer (μm)

H2O/

C2H5OH

T(°C) p(bar)

Conversion(%)

H2recovery(%)

H2purity

Pd-Cu

2012Supported Pd

on Al2O3

et al., 2015Supported

Pd-Ag on

PSS

et al., 2014Supported Pd

on Al2O3

et al., 2016a,2016bSupported

Ni-Pd-Ag

2008, 2010Supported Pd

on PSS

2011cSupported Pd

on PSS

et al., 2012Unsupported

Pd-Ag

and Basile,2010;

Pd-Ag

2008aUnsupported

Pd-Ag

2008bUnsupported

Pd-Ag

et al., 2009Composite

silica-alumina

2010Supported

on PSS as a composite membrane With this solution, at 500°C, the MR packed with aCo-based catalyst reached 100% of conversion and a hydrogen recovery of 70% Atmilder temperature conditions (400°C),Iulianelli et al (2016a)obtained 98% of con-version and 65% of hydrogen recovery, even though its purity was around 95% due tothe presence of pinholes and microcracks on the Pd layer formed under operations.Papadias et al (2010)performed ESR reaction at steam-to-carbon molar ratio equal

to 12/1, using a Rh-LaAl2O3catalyst packed in a 30μm thick Pd-Ag supported MR

As best result, at 700°C and 6.9 bar, they obtained 75% of hydrogen yield.Lin et al.(2008, 2010)used two kinds of supported membranes, Ni-Pd-Ag (8μm thick) andPd-Ag (20μm thick) Then, in MR, they obtained an ethanol conversion of around80% with the Ni-Pd-Ag MR operated at 450°C and 3.0 bar

However, a previous literature about ESR and/or BESR reaction in MRs involvesdense unsupported Pd-based MRs For example, Basile and coworkers utilized fullhydrogen permselective unsupported Pd-Ag membranes (50μm thick) (Basile et al.,2008a,b, 2011a; Seelam et al., 2012; Iulianelli and Basile, 2010; Iulianelli et al.,

2009, 2010a,b) in MRs to produce pure hydrogen from ESR or BESR reaction,reaching various results depending on the catalysts and operating conditions utilizedduring the experiments The best result was achieved using a synthetic bioethanolmixture and working with a Co-based catalyst at 400°C In these conditions, com-plete ethanol conversion and 90% of hydrogen recovered in the permeate side (100%pure) were the most relevant findings of their efforts (Iulianelli and Basile, 2010;Iulianelli et al., 2010a,b) Other options to the Pd-based MRs were considered forperforming ESR or BESR reaction In particular, Lim et al (2010) housed asilica-alumina composite membrane (not hydrogen fully permselective and pre-pared by chemical vapor deposition technique) in the MR at 350°C, reaching an eth-anol conversion of 90% and a hydrogen yield of around 20% AlsoYu et al (2009)used a SiO2-based membrane synthesized on PSS Thus, the reaction was carried out

in the MR packed with a Ru-SiO2catalyst The stream permeated through the brane is rich in hydrogen and CO, which is further converted via water-gas shiftreaction to decrease the CO content below 1% As shown inTable 9.1, at 600°Cand water/ethanol feed ratio equal to 5/1, complete ethanol conversion was reachedwith a hydrogen recovery around 85% (although not highly concentrated inhydrogen)

Methanol steam reforming (MSR) reaction to produce hydrogen has been extensivelystudied in CRs also because methanol represents a further renewable source havingseveral benefits as a hydrogen carrier for fuel-cell applications MSR reaction is con-ventionally carried out at relatively low temperatures between 240°C and 320°C andcan be represented by the following reaction mechanism (Iulianelli et al., 2014):