1
Expe imen al implemen a ion o a ca aly ic memb ane
eac o o he di ec syn hesis o DME om
H2+CO/CO2
Pablo Rod iguez-Vegaa, Aina a A ekaa*, Izumi Kumaki ib, Hec o Vicen ea, Ja ie
E eñaa, And es T. Aguayoa, Ja ie Bilbaoa
a Depa men o Chemical Enginee ing, Uni e si y o he Basque Coun y UPV/EHU,
P.O. Box 644, 48080 Bilbao, Spain
b G adua e School o Science and Technology o Inno a ion, G adua e School Science
and Enginee ing, Yamaguchi Uni e si y, Ube, 755-8611, Japan
*Co esponding au ho . Tel.: 34-94-6015361. E-mail add ess: aina a.a ek[email p o ec ed]
This is he accep ed manusc ip o he a icle ha appea ed in inal o m in Chemical Enginee ing Science 234 : (2021) //
A icle ID 116396, which has been published in inal o m a h ps://doi.o g/10.1016/j.ces.2020.116396. © 2021 Else ie unde
CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/)
2
Abs ac
The di ec syn hesis o dime hyl e he (DME) by he hyd ogena ion o CO2 and
CO2/COx mix u es has been s udied in an o iginal packed bed memb ane eac o
(PBMR). The ole o he hyd ophilic LTA zeoli e memb ane is o emo e H2O om he
eac ion medium, educing he e o e he he modynamic limi a ions o me hanol
syn hesis and dehyd a ion s ages. LTA zeoli e has he bes pe mea ion p ope ies among
he s udied zeoli es (LTX and SOD). The expe imen s we e ca ied ou using a CuO-
ZnO-Z O2/SAPO-11 ca alys a 275-325 ºC, 10-40 ba , space ime o 10 gca h (molC)-1
and using in he pe mea e sec ion a sweeping gas low a e o he same composi ion as
ha ed o he eac ion sec ion. The esul s (DME yield, CO2 con e sion and p oduc
dis ibu ion) o he PBMR a e compa ed wi h hose ob ained in PBR wi hou
memb ane. In he hyd ogena ion o CO2, a DME yield o 12 % and a CO2 con e sion
o 20 % a e ob ained a 275 ºC, 40 ba and space ime o 10 gca h (molC)-1 wi h a g ea
ca alys s abili y.
Keywo ds: dime hyl e he , CO2, syngas, memb ane eac o , LTA zeoli e, deac i a ion
3
1. In oduc ion
The u iliza ion o a memb ane in ca aly ic p ocess eac o s pu sues educing bo h
ene gy consump ion and p oduc ion cos s, as well as minimizing he en i onmen al
impac . This s a egy is pa o he enginee ing app oaches o p ocess in ensi ica ion
and has a g owing implan a ion in uels and chemical p oduc s syn hesis p ocesses om
sus ainable sou ces (N Diban e al., 2013; Cannilla e al., 2017; Tian e al., 2018).
Among hese p ocesses, he di ec syn hesis o dime hyl e he (DME) is conside ed one
o he mos a ac i e ou es o he la ge scale CO2 alo iza ion (Olah e al., 2009;
Ra iee e al., 2018; Leonzio, 2018). DME has upwa d comme cial in e es as
au omo i e and domes ic uel (has a ce ane numbe o 56) (A coumanis e al., 2008;
Semelsbe ge e al., 2006) and as eac an (al e na i e o me hanol) o he p oduc ion o
aw chemicals (ole ins and a oma ics) (Pé ez-U ia e e al., 2016; Co de o-Lanzac e al.,
2018) and he p oduc ion o H2 in ehicles ( h ough s eam e o ming) (Shimoda e al.,
2011; Oa -A e a e al., 2016). I is also widely used as e ige an and sp ay (Good and
F ancisco, 2003), and in oil ex ac ion (Ja anma d e al., 2019). The eac ion sys em o
i s di ec syn hesis om CO and CO2 hyd ogena ion in ol es:
Me hanol syn hesis: CO + 2H2 ↔ CH3OH (1)
CO2 + 3H2 ↔ CH3OH+ H2O (2)
Re e se wa e gas shi ( WGS): CO2 + H2 ↔ CO + H2O (3)
Me hanol dehyd a ion o DME: 2CH3OH ↔ CH3OCH3 + H2O (4)
Hyd oca bons o ma ion (undesi ed): CO +3H2 ↔ CH4 + H2O (5)
Using bi unc ional ca alys s, me hanol syn hesis and WGS eac ions a e ca alyzed by
he me allic unc ion, while me hanol dehyd a ion by he acid unc ion. Pe o ming
4
me hanol dehyd a ion in si u in he same eac o displaces he he modynamic
equilib ium o he me hanol syn hesis eac ions. The he modynamic ad an ages o e
me hanol syn hesis and he syn hesis o DME in wo sepa a e eac ion s ages (me hanol
syn hesis and i s dehyd a ion) ha e been quan i ied in he li e a u e, and a e pa icula ly
in e es ing o a o CO2 con e sion when co- ed wi h syngas (Chen e al., 2016; A eka
e al., 2017).
Ca alys p epa a ion has ecei ed g ea a en ion (Sun e al., 2014; Ca izzone e al.,
2018; Mondal and Yada , 2019), in pa icula , seeking o p omo e CO2 con e sion and
imp o e s abili y. The mos s udied me allic unc ion is CuO-ZnO-Al2O3 wi h di e en
me allic oxides such as MgO, CeO2, MnO o Z O2 among o he s and hei well-
es ablished beha io in he syn hesis o me hanol is epo ed. The me allic oxides we e
used o eplace Al2O3 pa ially o o ally o as p omo e s o inc ease he s abili y o Cu,
a enua ing i s sin e ing (F us e i e al., 2015; Bonu a e al., 2016; Zhou e al., 2016).
The con en ional acid unc ion used o me hanol dehyd a ion (γ-Al2O3) has been
p og essi ely eplaced by less hyd ophilic ma e ials, such as some zeoli es (HZSM-5
and e ie i e ha e been widely s udied) (Ga cía-T enco and Ma ínez, 2012; Cai e al.,
2016; F us e i e al., 2017) and silicoaluminophospha es (SAPOs), as SAPO-18 and
SAPO-11 (A eka e al., 2016, 2017). These acidic unc ions mus ha e a high densi y o
si es, bu o mode a e acidic s eng h o minimize he o ma ion o coke. The
con en ional p epa a ion me hod o bi unc ional ca alys s consis s o physically mixing
bo h unc ions (hyb id ca alys s), and subsequen pelle izing o achie e an adequa e
pa icle size and he equi ed mechanical esis ance o i s use in he eac o . As o he
co e-shell con igu a ion o he pa icles ega ds, e en i i can a oid nega i e e ec s
esul ing om he con ac be ween he me allic and acid unc ions (Ga cía-T enco e al.,
2012; Bonu a e al., 2020), he p epa a ion me hod has g ea e di icul ies han hose o
5
he hyb id ca alys (Sánchez-Con ado e al., 2018a). Conside ing he eac ion sys em
o he di ec syn hesis o DME (Eqs. (1)-(5)), he p esence o H2O in he eac ion
medium condi ions he he modynamics o he p ocess, limi ing he con e sion o CO
and CO2 yielding DME. Consequen ly, he p oposal o a hyd ophilic memb ane eac o
ha a o s he sepa a ion o H2O om he medium is a challenge o ele an in e es .
Iliu a e al. (2010, 2011) ha e s udied by simula ion he emo al o H2O om he
eac ion medium using a hyd ophilic memb ane o by means o adso p ion in o de o
inc ease me hanol and DME yields. Diban e al. (2013) ha e p oposed a ma hema ical
model o simula e an iso he mal packed bed memb ane eac o sys em, in o de o
de e mine he anspo cha ac e is ics o he mos sui able memb ane o he di ec
syn hesis o DME, co- eeding CO2 oge he wi h CO and H2. In his model
unidi ec ional low o he eed and he sweeping gas in coun e -cu en mode a e
conside ed. These au ho s emphasize he need o a s able hyd ophilic memb ane (ZSM-
5, MOR, SIL) as hose used o Fische -T opsch (FT) syn hesis. They also poin ed ou
he es ic ions o H2O pe mea ion selec i i y a high eac ion empe a u es. Using o
he simula ion an ideal memb ane (impe meable o me hanol and DME) hese au ho s
(Diban e al., 2014) achie ed an upg ade o DME yield o o e 30 %, compa ed o ha
ob ained wi hou a memb ane. Fa si e al. (2016) ha e compa ed, by simula ion o a
noniso he mal one-dimensional eac o , he pe o mance o he di ec syn hesis o
DME o a double memb ane eac o (hyd ogen-wa e ), p e iously p oposed o
me hanol syn hesis (Fa si and Jahanmi i, 2012), wi h ha o o he eac o s wi h
hyd ophilic memb anes, hyd ogen selec i e memb anes and wi hou memb ane. De
Falco e al. (2016) ha e demons a ed by simula ion he ad an ages o wo al e na i e
eac ion sys ems, hus, a zeoli e memb ane eac o and wo se ies uni s, which consis ed
o a packed bed eac o and a wa e sepa a ion module. Subsequen ly, hese au ho s (De
6
Falco e al., 2017a) ha e s udied he e ec o he ope a ion condi ions (CO2/COx and
H2/COx a ios in he eed, empe a u e, p essu e, space ime) on CO2 con e sion and
DME yield, a indus ial scale condi ions wi h a non-iso he mal model in he simula ion
o he memb ane eac o . Among he simula ion esul s, he achie emen o he
ollowing esul s is o be highligh ed: a DME yield o 0.75 (0.57 in he con en ional
eac o ), a DME selec i i y close o 100 % and COx and CO2 con e sions o 0.75 and
0.69, espec i ely, enhancing in 15.4 % and 30.2 % he esul s in he con en ional
eac o . Addi ionally, he esul s ha e been imp o ed by hese au ho s (De Falco e al.,
2017b), wi h an ope a ion s a egy called Double Recycling Loop DME (DRL-DME).
This design consis s o he u iliza ion o a pu e CO2 s eam as sweeping gas in he
pe mea ion zone and eci cula ion o he eac o .
Despi e he ad ances in he p epa a ion o zeoli e memb anes and in hei applica ion o
memb ane eac o s in di e en ca aly ic p ocesses (Beda d and Liu, 2018), he
applica ion o his ype o memb anes in me hanol and DME syn hesis is s ill a he
p elimina y s age, consis ing o he p oposal o H2O pe m-selec i e memb anes o be
used in se e e eac ion condi ions (Gallucci, 2018). Galluci e al. (2004) ha e
expe imen ally upg aded he yield o me hanol syn hesis using a LTA memb ane, and
Fedoso e al. (2015) ha e used a LTA memb ane (NaA Zeoli e) o he dehyd a ion o
me hanol owa ds DME. Go be e . al. (2018) ha e ob ained esul s on he capaci y o A
zeoli e o selec i ely sepa a e wa e and me hanol. The expe imen s ha e been ca ied
ou measu ing he pe mea ion o a H2, CO2 and H2O mix u e, wi hin he ange o
in e es o me hanol syn hesis (160-240 ºC, 10-27 ba ). Recen ly, Lee e al. (2021) ha e
used a polyimide hollow ibe memb ane eac o o me hanol syn hesis.
In his wo k, a labo a o y scale packed bed eac o p o ided wi h a LTA zeoli e
memb ane (selec ed om a g oup o ma e ials acco ding o i s pe mselec i i y) has been
7
used in he di ec syn hesis o DME. The s udy has been conduc ed in a wide ange o
ope a ing condi ions, analyzing he e ec s o eac ion empe a u e, p essu e and
CO2/COx a io in he eed on a ious eac ion indices (DME and me hanol yield and
selec i i y, COx con e sion and CO2 con e sion) and on he s abili y o he ca alys .
The expe imen al esul s a e compa ed wi h hose ob ained in a eac o wi hou
memb ane o quan i y he shi a ained in he yield o DME and he con e sion o CO2
o e he he modynamic limi a ions. The composi ion o he used ca alys (CuO-ZnO-
Z O2/SAPO-11) has been op imized in p e ious wo ks (Sánchez-Con ado e al., 2018a,
2018b, 2018c).
2. Expe imen al
2.1. Ca alys p epa a ion and cha ac e iza ion
CuO-ZnO-Z O2/SAPO-11 ca alys was p epa ed by physical mix u e and subsequen
pelle izing o he me allic and acid unc ions in a 1/2 mass a io. The a omic Cu:Zn:Z
a io o he me allic unc ion (de e mined by ICP-OES analysis) is 2:0.75:1.21. The
sui abili y o his composi ion and he p epa a ion condi ions o each o he unc ions
ha e been es ablished in p e ious wo ks (Sánchez-Con ado e al., 2018a, 2018b,
2018c). Likewise, he cha ac e iza ion me hodologies o each o he unc ions
comp ising he ca alys and he bi unc ional ca alys ha e been desc ibed in de ail in
hese wo ks. The mos signi ican physico-chemical p ope ies o he inal ca alys a e
lis ed in Table 1.
8
Table 1. Physico-chemical p ope ies o he CuO-ZnO-Z O2/SAPO-11 bi unc ional
ca alys .
Physical p ope ies
SBET (m2 g-1)
Vmic o (cm3 g-1)
Vmeso (cm3 g-1)
dp (Å)
122
0.029
0.170
97.9
Me allic p ope ies
SCu (m
2
gCu
-1
)
S´Cu (m
2
gca
-1
)
Cu dispe sion (%)
53.5
6.3
8.2
Acid p ope ies
To al acidi y (mmolNH3 g-1)
Acid s eng h (kJ molNH3-1)
0.17
85
2.2. Memb ane p epa a ion and cha ac e iza ion
Th ee ypes o mic opo ous zeoli e memb anes, hus, Linde Type A (LTA), Linde Type
X (LTX) and Sodali e (SOD) memb anes we e syn hesized on s ainless s eel suppo s
pu chased om Mo Co po a ion. Table 2 shows he syn heses empe a u es and imes
o he di e en memb anes. The p epa a ion condi ions and ma e ials a e de ailed in
he Appendix (Tables A4-A6).
Table 2. Hyd o he mal condi ions o he syn hesis o he di e en memb anes.
Zeoli e memb ane
Tempe a u e (ºC)
Time (h)
Linde ype A (LTA)
110
12
Linde ype X (LTX)
90
8
Sodali e (SOD)
130
48
The mo phology o he memb anes was analyzed by scanning elec on mic oscopy
(SEM, in a JEOL/JSM-7000 equipmen , equipped wi h a W ilamen , 3.5 eV
esolu ion, a ached o ene gy dispe si e X- ay analyze EDX, Ox o d, 133 eV
esolu ion).
9
Pe apo a ion and apo pe mea ion we e pe o med o examine he dehyd a ion
pe o mances o he memb anes. The pe apo a ion (PV) o he di e en memb anes
was e alua ed as desc ibed in he Appendix. The sepa a ion ac o s (Eqs. (A1) and
(A2)) and luxes (Eq. (A3)) o he PV expe imen s we e calcula ed o E OH/H2O and
MeOH/H2O mix u es a 75 ºC and 60 ºC, espec i ely.
In addi ion, pe mea ion o di e en gases (He, H2, CO2 N2, CH4 and SF6) was
measu ed a 100-200 ºC. Ideal selec i i y was calcula ed om he pe mea ion a io o
wo gases. The mal s abili y es s we e pe o med by cha ac e izing he memb anes
be o e and a e he mally ea ing he memb anes a 300 ºC wi h PV, VP, XRD and
SEM. The empe a u e o he he mal ea men ep esen s he DME syn hesis
empe a u e.
The used echniques and analysis condi ions ha e been desc ibed in he Appendix
(sec ion A3).
2.3. Packed bed memb ane eac o (PBMR)
The s ainless s eel packed bed memb ane eac o (PBMR), shown in Fig. 1, s uc u ally
p esen s wo concen ic sec ions: he eac ion sec ion, in which he ca aly ic bed is
loca ed, su ounded by ano he concen ic sys em ha allows he sweep o he pe mea e
low (pe mea e sec ion).
16
Table 3. Resul s o H2O pe apo a ion om E OH /H2O and MeOH/H2O mix u es
h ough he di e en memb anes a 75 ºC and 60 ºC, espec i ely.
Memb ane
Composi ion
Feed
(w %)
Pe mea e
(w %)
Q
(kg m
-2
h
-1
)
α
LTA
E OH
H2O
90
10
0.15
99.8
1.59 9329
LTA
MeOH
H2O
90
10
0.45
99.5
0.92 1486
LTX
E OH
H2O
90
10
55.4
44.6
2.40 7.9
LTX
MeOH
H2O
90
10
69.4
30.6
2.00 2.9
SOD
E OH
H2O
90
10
89.7
10.3
0.80 1.1
3.2. Memb ane mo phology
Physical and mo phological p ope ies. The po ous ex u e o he LTA zeoli e powde
has been cha ac e ized by CO2 adso p ion a 0 ºC (in ASAP 2020 equipmen ,
Mic ome i ics) (Fig. (A3)). To ca y ou he analysis, LTA has been p epa ed ollowing
he same p ocedu e desc ibed in he Appendix, wi hou in oducing he me allic suppo
in he au ocla e. Well-de ined CO2 adso p ion iso he m has been ob ained wi h a
p ac ically sa u a ed shape in he ange o low ela i e p essu e. Po e wid h and
mic opo e su ace, calcula ed using he Dubinin-Radushke ich equa ion and by he
Ho a h-Kawazoe me hod, a e 3.98 Å and 458 m2 g-1, espec i ely. These esul s
sugges ha he c ys alline zeoli e p esen s na ow mic opo es (3.98 Å).
S uc u al p ope ies. X ay di ac ion (XRD) measu emen s ha e been used o s udy
he c ys allini y and pu i y o he LTA memb ane. F om he XRD pa e n o he LTA
memb ane (Fig. 3) i can be con i med, on one hand, ha he zeoli e has been
app op ia ely syn hesized (Bel iso e al., 2018) and, on he o he hand, i s co ec
17
c ys alliza ion on he s ainless s eel suppo ( undamen al s ep wi hin he memb ane
eac o con igu a ion). Nea and well-de ined peaks can be obse ed, indica ing an
absence o amo phous phase in he ma e ial. Acco ding o he da abase, he mos in ense
peaks on he di ac og am clea ly iden i ied in Fig. 3 sugges ha a c ys alline phase
wi h high pu i y suppo ed on s ainless s eel can be achie ed h ough his p epa a ion
me hod. The XRD pa e n o s eel can also be obse ed in Fig. 3, consis ing o h ee
peaks a highe alues o 2θ angles (67.1, 79.5 and 129.1 deg ees) a ibu ed o he
me allic suppo o he memb ane (ob ained in a PAN analy ical Xpe P o de ice).
Fig. 3. XRD pa e n o he syn hesized LTA zeoli e c ys allized on s ainless-s eel
suppo .
In Fig. 4 SEM mic og aphs o he syn hesized LTA zeoli e a e depic ed. A he on
iew o he memb ane c oss-sec ion (Fig. 4a) bo h phases (zeoli e and suppo ) a e
clea ly di e en ia ed. A zeoli e laye o egula hickness has been deposi ed on he
su ace o he non-polished s ainless s eel suppo . The homogeneous ex u e o he
18
LTA memb ane (Fig. 4b) sugges s a high e ec i eness o he p epa a ion me hod o
he pu pose o achie ing a comple e co e age o he suppo wi h zeoli e c ys als, in
acco dance wi h ha epo ed by Bel iso e al. (2018). As i is obse ed in Fig. 4c LTA
zeoli e mo phology consis s o cubic agg ega es. SEM cha ac e iza ion has
demons a ed he co ec syn hesis me hod o LTA memb anes, highligh ing he
o ma ion o uni o m hickness zeoli e laye and i s homogeneous dis ibu ion on he
suppo su ace.
19
Fig. 4. SEM images o he LTA zeoli e memb ane. C oss-sec ion (a) and op iew
a di e en scales (b and c).
3.3. S abili y o LTA memb ane
Pe apo a ion. The PV p ope ies o he LTA memb ane a e i s he mal ea men a e
shown in Table 4. Compa ing wi h he esul s p io o he he mal ea men (Table 3)
he sepa a ion ac o s (α), bo h o E OH/H2O and MeOH/H2O mix u es, dec ease a e
he ea men (2905 and 1011 espec i ely). Ne e heless, i main ains i s hyd ophilici y
a)
b)
c)
20
owa ds bo h mix u es. Conce ning he luxes (Q) o he wo eed composi ions h ough
he memb ane, hey a e sligh ly aised om 1.59 o 1.70 kg m-2 h-1 o E OH/H2O (≈ 6
%) and om 0.92 o 1.05 kg m-2 h-1 o MeOH/H2O (≈ 14 %).
Table 4. E OH/H2O and MeOH/H2O mix u es pe apo a ion esul s h ough he
he mally ea ed LTA memb ane a 75 ºC and 60 ºC, espec i ely.
Composi ion
Feed
(w %)
Pe mea e
(w %)
Q
(kg m
-2
h
-1
)
α
E OH
H
2
O
90
10
0.47
99.5
1.70 2905
MeOH
H2O
90
10
0.75
99.2
1.05 1011
Vapo pe mea ion. VP pe o mance o he LTA memb ane a e i s he mal ea men is
summa ized in Table 5. LTA memb ane H2O pe mselec i i y emains a e i s ea men
(highe han 3000) and he apo lux h ough i ises up o 1.32 kg m-2 h-1.
Table 5. E OH/H2O mix u e apo pe mea ion esul s h ough he he mally ea ed
LTA memb ane a 125 ºC.
Composi ion
Feed
(w %)
Pe mea e
(w %)
Q
(kg m
-2
h
-1
)
α
E OH
H2O
90
10
0.19
99.8
1.32 3232
Single gas pe mea ion. The pe meances o di e en gases a di e en empe a u es
(100, 150 and 200 ºC) h ough he he mally ea ed LTA memb ane a e shown in Table
6. Mo eo e , in Table 7 he ideal selec i i ies o di e en gases a 100, 150 and 200 ºC
h ough he he mally ea ed LTA memb ane ha e been ga he ed. Inc easing
empe a u e up o 200 ºC, a sligh inc emen in he gas pe meances h ough he
21
he mally ea ed LTA memb ane is obse ed. In spi e o his pe meance inc ease, he
analysis demons a es ha e en a high empe a u es, he LTA memb ane has low gas
pe meabili y and, he e o e, is able o ope a e a empe a u es nea o hose equi ed in
he syn hesis o DME.
Table 6. Gas pe meances h ough he he mally ea ed LTA memb ane a 100, 150
and 200 ºC (in 10-9 mol m-2 s-1 Pa-1).
Tempe a u e (ºC) He H2 CO2 N2 CH4 SF6
100 2.39 2.90 0.56 0.78 0.93 0.31
150 3.47 4.60 0.94 1.24 1.61 0.55
200 9.24 13.2 3.18 3.61 4.34 1.66
Table 7. Ideal selec i i ies o di e en gas mix u es h ough he mally ea ed LTA
memb ane a 100, 150 and 200 ºC.
Tempe a u e (ºC)
H
2
/N
2
H
2
/CH
4
CO
2
/N
2
CO
2
/CH
4
He/SF6
H
2
/SF
6
100 3.7 3.1 0.3 0.2 10 11
150
3.7
2.8
0.8
0.6
6
8
200 3.7 3.0 0.9 0.7 6 8
Ine cha ac e . The null ac i i y o he memb ane in he eac ions in ol ed in he
syn hesis o DME (syn hesis o me hanol and i s dehyd a ion o DME) has been
asce ained by means o expe imen s eeding syngas and co- eeding me hanol wi h
syngas, espec i ely. In bo h cases he con e sion and he o ma ion o pa a ins and
coke is null.
3.4. PBMR, E ec o eac ion empe a u e
Fig. 5 shows an example o he e olu ion wi h ime on s eam (TOS) o he mola
ac ions o he eac ion p oduc s (DME, MeOH, H2O) a he eac o ou le in he
22
eac ion sec ion (RS) and in he pe mea e sec ion (PS). These esul s, shown as an
example, co espond o CO2 hyd ogena ion a 275 ºC and 40 ba . Unde hese
condi ions, he ad ance o he e e se WGS eac ion is acili a ed, wha leads o a high
o ma ion o H2O. I should be no ed ha his is a key eac ion o he p oduc ion o
DME since CO is mo e e ec i e han CO2 unde hese condi ions (Aguayo e al., 2007;
A eka e al., 2018). The low concen a ions o DME and me hanol in he PS, show ha
he memb ane is pe m-selec i e owa ds hese molecules (kine ic diame e o ca. 4.3 Å
and 3.8 Å, espec i ely, and 2.6 Å o H2O), specially o DME. In Fig. A2, he
concen a ion alues o all he gaseous componen s in he eac ion and pe mea e
sec ions a e shown o di e en eac ion empe a u e and di e en CO2/COx a ios in
he eed.
The simila H2O concen a ion alues in he eac ion and pe mea ion sec ions obse ed
in Figs. 5 and A2 e eal a good deg ee o H2O sepa a ion a ained wi h he memb ane.
None heless, he undesi ed pa ial pe mea ion o me hanol and DME also akes place.
The high p essu e and empe a u e equi ed o his eac ion ha e a g ea esponsibili y
on limi ing he pe mea ion selec i i y o H2O wi h espec o oxygena es.
23
30 60 90 120 150 180
0
0.005
0.010
0.015
0.040
0.045
0.050
30 60 90 120 150 1800
0.005
0.010
0.015
0.040
0.045
0.050
b) PS
TOS (min)TOS (min)
a) RS
H2O
DME
MeOH
Mola F ac ion
Fig. 5. E olu ion o H2O, DME and me hanol mola ac ions wi h ime on s eam.
(a) esul s in he eac ion sec ion (RS) and (b) esul s in he pe mea e sec ion
(PS). Reac ion condi ions: 275 ºC; 40 ba ; 10 gca h (molC)-1; CO2/COx, 1;
H2/COx, 3. Pe mea e condi ions: equal composi ion and low a e
(60 cm3 min-1) as in he eac ion sec ion.
The common empe a u e ange o DME syn hesis s udied in he li e a u e is 250-
300 ºC, since he empe a u e is limi ed by he he modynamic equilib ium (A eka e al.,
2017) and also o p ese e he hyd o he mal s abili y o Cu in he ca alys . Ne e heless,
he u iliza ion o he memb ane eac o (PBMR) diminishes he he modynamic
limi a ions, due o he lowe H2O concen a ion in he eac ion medium, pe mi ing he
displacemen o he he modynamic equilib ium o H2O o ma ion eac ions (me hanol
syn hesis, e e se WGS and me hanol dehyd a ion). Consequen ly, i allows ope a ing
a highe empe a u es achie ing a highe con e sion. Fu he mo e, he composi ion and
condi ions used o he p epa a ion o he ca alys a e adequa e o con e an accep able
hyd o he mal s abili y a 325 ºC. The ac ha H2O concen a ion in he eac ion
24
medium is lowe han in he con en ional eac o (PBR) is, a p io i, ano he ad an age
o he PBMR, in which less Cu sin e ing is expec ed. A ending o his ad an age he
s udied empe a u e ange has been 275-325 ºC.
Fig. 6 shows he e ec o empe a u e in he di e en eac ion indices a ze o TOS
(Figs. 6a,b) and on he s abili y o he ca alys (Fig. 6c) o ce ain eac ion condi ions
(30 ba ; 10 gca h (molC)-1; CO2/COx, 0.5; H2/COx, 3) and H2+CO/CO2 eeds ock. Fig.
6a shows ha empe a u e has a g ea ele ance on COx con e sion, imp o ing om
13.3 % a 275 ºC o 19.3 % a 325 ºC. DME yield enhances om 275 ºC un il 300 ºC,
achie ing a maximum o 14.8 % a his empe a u e. Fu he inc easing eac ion
empe a u e, a sligh decay is obse ed due o, p edic ably, he he modynamic
limi a ion. Howe e , CO2 con e sion is a o ed upon ising empe a u e, eaching
17.5 % a 325 ºC. MeOH and pa a ins yields con inue ising wi h a maximum o 3.7 %
and 1.1 %, espec i ely, a 325 ºC.
As a consequence o he e olu ion o he indi idual yields wi h empe a u e, DME
selec i i y (Fig. 6b) d ops when inc easing empe a u e om 275 o 325 ºC, emaining
quasi-cons an a highe empe a u e. On he o he hand, pa a ins selec i i y should be
poin ed ou since i ises up o 5.6 % a 325 ºC. A his empe a u e, pa a ins o ma ion
mechanisms a e p omo ed, ei he om DME and me hanol h ough he hyd oca bon
pool mechanism (ac i a ed by he acid unc ion) o h ough me hana ion o Fische -
T opsch mechanisms om CO and CO2 on he me allic unc ion.
The o ma ion o hyd oca bons acili a es deac i a ion by coke deposi ion, due o hei
ole as in e media es o a oma ics o ma ion, which condensa e o coke. The e ec o
empe a u e on he e olu ion o DME yield wi h ime on s eam (TOS) depic ed in Fig.
6c demons a es how empe a u e a ec s he s abili y o he ca alys , which is lowe he
25
highe he empe a u e is wi hin he ange o 275-300 ºC. Ne e heless, a 325 ºC
ca alys s abili y is g ea e han a 300 ºC, which can be ela ed o ca alys sin e ing. The
lowe DME yield a 325 ºC han a 300 ºC, a ze o ime on s eam, has he e o e lowe
deac i a ion as a o able coun e pa . I should also be poin ed ou ha deac i a ion a
275 ºC is signi ican ly slow.
275 300 325
0
5
10
15
20
Yi (%)
Tempe a u e (ºC)
DME
MeOH
HC
0
5
10
15
20
a)
X
CO
x
X
CO
2
X
CO
2
, XCO
x
(%)
275 300 325
0
20
40
60
80
100 b)
S
i (%)
Tempe a u e (ºC)
DME MeOH HC
32
20 30 40
0
20
40
60
80
100 b)
DME MeOH HC
Si
(%)
P essu e (ba )
020 40 60 80 100 120 140 160 180
0
5
10
15
20
25 c)
TOS (min)
YDME (%)
40 ba
30 ba
20 ba
Fig. 8. E ec o p essu e on DME, MeOH and hyd oca bons yield, and COx and
CO2 con e sion (a) and on p oduc selec i i y (b) a ze o ime on s eam;
and he e olu ion o DME yield wi h ime on s eam a di e en eac ion
p essu es (c). Reac ion condi ions: 300 º; 10 gca h (molC)-1; CO2/COx, 0.5;
H2/CO, 3. Pe mea e condi ions: equal composi ion and low a e
(60 cm3 min-1) as in he eac ion sec ion.
33
The e ec o p essu e on he eac ion indices eeding H2+CO2 has been s udied a
275 ºC (Fig. 9). This empe a u e has been de e mined in Sec ion 3.1 o be he mos
app op ia e o achie e maximum DME yield wi h his eeds ock. The inc ease o he
ope a ing p essu e a o s he linea ise o bo h COx con e sion ( om 3.0 o 14.9 %)
and CO2 con e sion ( om 9.4 o 25.0 %) wi hin 10-40 ba ange (Fig. 9a). Conce ning
DME and MeOH yields, he enhancemen wi h p essu e is clea ly obse ed whe eas
pa a ins o ma ion emains quasi-negligible, achie ing a maximum o 0.01 % unde
40 ba . Due he cons an ise o he p oduc s yield wi hin he s udied p essu e ange,
MeOH and DME selec i i y (Fig. 9b) a e ba ely a ec ed by he inc ease o he
ope a ing p essu e ( om 23.9 o 20.0 % and om 76.1 o 80.0 %, espec i ely). Fig. 9c
displays he deac i a ion o he ca alys wi h ime on s eam. I demons a es ha , a his
empe a u e and eeding H2+CO2, he deac i a ion o he ca alys is p ac ically
negligible h oughou 3 h o eac ion a he s udied p essu e ange.
A ending o hese esul s unde di e en p essu es, e en hough empe a u e is di e en
o he wo eeds, he e ec o p essu e o H2+CO/CO2 and H2+CO2 eeds is
quali a i ely simila . The d op o DME yield dec easing he p essu e (in e es ing o
educing ope a ing cos s) is p opo ionally simila o bo h eeds, as well as he decay in
COx con e sion. Rega ding DME selec i i y, i is cons an wi h p essu e in bo h cases.
None heless, eeding syngas, a p essu e dec ease implies a ise o pa a ins selec i i y o
he de imen o ha o me hanol (being a o ed CH4 syn hesis and Fische -T opsch
eac ions wi h espec o me hanol syn hesis). Feeding CO2 a 275 ºC, on he con a y,
pa a ins o ma ion is insigni ican e en unde 10 ba . I is also ema kable ha ca alys
deac i a ion is sligh ly a o ed by he inc ease o p essu e o syngas eeds a 325 ºC,
and e y slow o H2 +CO2 eeds a 275 ºC.
34
10 20 30 40
0
5
10
15
20
25 a)
XCO
2
, XCOx (%)
Yi (%)
P essu e (ba )
DME
MeOH
HC
0
5
10
15
20
25
XCO2
XCOx
10 20 30 40
0
20
40
60
80
100 b)
DME MeOH HC
S
i (%)
P essu e (ba )
35
020 40 60 80 100 120 140 160 180
0
5
10
15 c)
TOS (min)
YDME (%)
40 ba
30 ba
20 ba
10 ba
Fig. 9. E ec o p essu e on DME, MeOH and hyd oca bons yield, and COx and
CO2 con e sion (a) and on p oduc selec i i y (b) a ze o ime on s eam;
and he e olu ion o DME yield wi h ime on s eam a di e en p essu es
(c). Reac ion condi ions: 275 º; 10 gca h (molC)-1; CO2/COx, 1; H2/CO, 3.
Pe mea e condi ions: equal composi ion and low a e (60 cm3 min-1) as in
he eac ion sec ion.
3.6. Compa ison be ween PBMR and PBR
The elimina ion o H2O om he eac ion medium has wo po en ial e ec s: i) he
al e a ion o he heo e ical he modynamic equilib ium owa ds an appa en
equilib ium, and ii) he inc ease o he eac ion a e o some s ages o he eac ion. As
a o emen ioned, acco ding o he he modynamics (A eka e al., 2017), when CO2
concen a ion in he eeds ock inc eases, he he modynamic limi a ions o he eac ion
augmen , due o he inc emen o H2O concen a ion in he eac ion medium ( o med by
me hanol syn hesis (Eq. (2)) and WGS (Eq. (3)) eac ions). Besides, he H2O is
36
adso bed on he acid si es o he ca alys compe ing wi h he eac an s and educing i s
capaci y o dehyd a e me hanol (Jun e al., 2002) on he one hand, and, on he o he
hand, on he me allic si es limi ing he me hanol syn hesis a e (Dadga e al., 2016).
Fig. 10 shows oxygena es yields (MeOH + DME) ob ained o ce ain ope a ing
condi ions a di e en empe a u es. These esul s aim o e i y how he emo al o
H2O om he eac ion medium using a PBMR pe mi s o e aking he he modynamic
equilib ium o he p ocess p edic ed o PBR (wi hou memb ane o H2O emo al). I
can be obse ed ha inc easing empe a u e up o 325 ºC, he yield enhancemen using a
PBMR compa ed wi h he p ocess he modynamic equilib ium o PBR becomes mo e
signi ican .
275 300 325
0
5
10
15
YOxyg. (%)
Tempe a u e (ºC)
STD The modynamic Equilib ium
Expe imen al PBMR
STD The modynamics
O e aking
Fig. 10. Compa ison o he expe imen al alues o oxygena es yields ob ained in
PBMR wi h he he modynamic equilib ium alues p edic ed wi hou
memb ane. Reac ion condi ions: 20 ba ; 10 gca h (molC)-1; CO2/COx, 1;
H2/COx, 3. Pe mea e condi ions o he PBMR: equal composi ion and low
a e (60 cm3 min-1) as in he eac ion sec ion.
37
Fo s udying quan i a i ely he in luence o inco po a ing a memb ane on he yield o
oxygena es and on CO2 con e sion Fig. 11 is p esen ed. The compa ison be ween
oxygena es (DME and MeOH) yield e olu ion wi h empe a u e ob ained in PBR and
PBMR eac o s eeding H2+CO2 is depic ed in Fig. 11a. In all he cases, wi hin he
s udied empe a u e ange (275-325 ºC) and a 30 ba , oxygena es yield is enhanced
using a PBMR. This imp o emen is g ea e when inc easing empe a u e, om 2.80 o
6.57 % a 325 ºC. The con e sion o CO2 is also g ea e in he PBMR wi hin he s udied
empe a u e ange, obse ing he highes enhancemen (37 %) a 325 ºC (Fig. 11b). The
esul s in Figs. 10 and 11 highligh ha in he ope a ion in he PBMR, he sepa a ion o
H2O has a g ea e ec on inc easing he yield o oxygena es and CO2 con e sion. As a
consequence o he pa ial sepa a ion o H2O om he eac ion medium, a pseudo-
equilib ium s a e is eached, whose con e sion also dec eases wi h inc easing
empe a u e. In his pseudo-equilib ium, he dec ease o limi con e sion wi h ising
empe a u e is a enua ed wi h espec o he he modynamics p edic ion.
Fig. 11c shows he e ec o CO2/COx a io in he eed om 0 o 1 on oxygena es yield
a 325 ºC. Highe yield is obse ed in PBMR wha e e he CO2 con en in he eed.
Being he mos ema kable, he g ea es upg ade ob ained wi h he highes CO2 con en
in he eed, imp o ing oxygena es yield by 86 % when eeding H2+CO2.
38
275 300 325
0
5
10 a)
YOxyg. (%)
Tempe a u e (ºC)
PBR
PBMR
275 300 325
0
5
10
15
20
25
30 b)
PBMR
PBR
X
CO2
(%)
Tempe a u e (ºC)
39
0.0 0.5 1.0
0
10
20
30
40
50
YOxyg. (%)
CO2/COx
PBR
PBMR
c)
Fig. 11. Compa ison o he PBMR and PBR. Oxygena es yield (a) and CO2
con e sion (b) a di e en empe a u es; and oxygena es yield o di e en
CO2/COx a ios in he eed (c). Reac ion condi ions: 325 ºC; 30 ba ;
10 gca h (molC)-1; H2/COx, 3; CO2/COx, 1. Pe mea e condi ions o he
PBMR: equal composi ion and low a e (60 cm3 min-1) as in he eac ion
sec ion.
The e ec o memb ane u iliza ion is mo e ele an wi h inc easing space ime. In Fig.
12, he yield o DME and he con e sion o CO2 a ze o ime on s eam can be obse ed
o di e en space ime alues. As i can be obse ed, he be e pe o mance o he
PBMR o e he PBR is g ea e upon inc easing space ime (indus ially ope a ing
condi ions). Indeed, o he s udied condi ions (325 ºC, 30 ba ) and a H2+COx eeds ock
wi h CO2/COx= 0.5, using a memb ane eac o doubles he yield o DME ob ained wi h
a PBR o a space ime o 100 g h (molC)-1, and boos s CO2 con e sion in a majo
ex en . Tha is, a la ge-scale p oduc ion condi ions, hus, a high space ime alues, he
40
sepa a ion o H2O om he eac ion medium has a g ea e e ec because i shi s he
alues o DME yield and CO2 con e sion wi h espec o he he modynamic
equilib ium in he PBR.
020 40 60 80 100
0
10
20
30
40
Space ime (g h (molC)-1)
PBR PRMR
YDME
XCO2
YDME , XCO2 (%)
Fig. 12. Compa ison o he e olu ion wi h space ime o DME yield and CO2
con e sion in PBMR (con inuous lines) and PBR (dashed lines). Reac ion
condi ions: 325 ºC; 30 ba ; H2/COx, 3; CO2/COx, 0.5.
Al hough he emo al o H2O om he eac ion medium imp o es he eac ion a es, i
has an undesi ed side e ec , because i is well es ablished ha he p esence o H2O in
he eac ion medium educes coke o ma ion. Sie a e al. (2011) explain his e ec by
he a enua ion o he o ma ion o me hoxy ions om me hanol and DME. I is also
well es ablished ha hese ions a e he in e media es p ecu so s o hyd oca bons
o ma ion in he acid unc ion (Bjø gen e al., 2007). In addi ion, he g ea e capaci y o
DME o he o ma ion o me hoxy ions is conside ed o be he main cause o he highe
a e o hyd oca bons and coke o ma ion compa ed o ha o me hanol (Co de o-Lanzac
41
e al., 2018; Ibáñez e al., 2017). This e ec has been s udied by de e mining he coke
con en in he used ca alys s by means o he mog a ime ic analyses o hei
combus ion wi h ai . Fo he condi ions in Fig. 12 and CO2/COx = 0.5, coke con en in
3 h is 0.32 w % in he PBR and 0.61 w % in he PBMR. This deposi ion o coke esul s
in a pa ial de e io a ion o he p ope ies o he ca alys . The ex u al, me allic and acid
p ope ies o he ca alys a e i s use unde di e en eac ion condi ions ha e been
ga he ed in Table A3. Consequen ly, and as expec ed, he emo al o H2O om he
eac ion medium has he un a o able e ec o up u ning coke con en . Howe e , his
e ec is no e y ele an and has li le consequence in he e olu ion o he eac ion
indices wi h TOS. To explain his low incidence o H2O sepa a ion on deac i a ion, i
mus be aken in o accoun ha he g ea e ad ance o he eac ion in he PBMR also
yields highe H2O con en . Thus, despi e he sepa a ion, a su icien H2O concen a ion
is p esen o limi he o ma ion o coke. In addi ion, his p oblem is less ele an when
eeding CO2, since he o ma ion o H2O is a o ed and he e o e, he a enua ion o
deac i a ion.
The s abili y o he a o emen ioned esul s has been expe imen ally asce ained by
means o epea ed uns wi h he same memb ane. I should be no ed ha he esul s in
his wo k co espond o pe -pass con e sions. Indeed, o assess he e ec o he eac ion
a iables in he pe o mance o he ca alys , he s udy has been ca ied ou unde kine ic
egime, ensu ing he e is no he modynamic es ic ions condi ioning he esul s;
enabling he e o e he compa ison o PBMR and PBR. The scaling up his p ocess
should be ca ied ou ecycling he noncon e ed eac an s (H2, CO and CO2) a e he
condensa ion o he oxygena es and H2O. This eci cula ion s a egy is used indus ially
in he syn hesis o me hanol (Bozzano and Manen i, 2016) and has also been p oposed
48
polyimide hollow ibe memb ane eac o . Chem. Eng. J. 403, 126457–126466.
h ps://doi.o g/10.1016/j.cej.2020.126457
Leonzio, G., 2018. S a e o a and pe spec i es abou he p oduc ion o me hanol,
dime hyl e he and syngas by ca bon dioxide hyd ogena ion. J. CO2 U il. 27, 326–
354. h ps://doi.o g/10.1016/j.jcou.2018.08.005
Mondal, U., Yada , G.D., 2019. Pe spec i e o dime hyl e he as uel: Pa I. Ca alysis.
J. CO2 U il. 32, 299–320. h ps://doi.o g/10.1016/j.jcou.2019.02.003
Oa -A e a, L., Remi o, A., Ep on, F., Bion, N., Aguayo, A.T., Bilbao, J., Gayubo,
A.G., 2016. Compa ison o Noble Me al- and Coppe -Based Ca alys s o he S ep
o Me hanol S eam Re o ming in he Dime hyl E he S eam Re o ming P ocess.
Ind. Eng. Chem. Res. 55, 3546–3555. h ps://doi.o g/10.1021/acs.iec .6b00126
Olah, G.A., Goeppe , A., P akash, G.K.S., 2009. Chemical ecycling o ca bon dioxide
o me hanol and dime hyl e he : F om g eenhouse gas o enewable,
en i onmen ally ca bon neu al uels and syn he ic hyd oca bons. J. O g. Chem.
74, 487–498. h ps://doi.o g/10.1021/jo801260
Osa iash iani, A., Pué olas, B., Oli ei a, C.C.S., Manayil, J.C., Ba be o, B., Isaacs, M.,
Michailo , C., He acleous, E., Pé ez-Ramí ez, J., Lee, A.F., Wilson, K., 2017. On
he in luence o Si:Al a io and hie a chical po osi y o FAU zeoli es in solid acid
ca alysed es e i ica ion p e ea men o bio-oil. Biomass Con e s. Bio e ine y 7,
331–342. h ps://doi.o g/10.1007/s13399-017-0254-x
Pe ez-Ca bajo, J., Bales a, S.R.G., Cale o, S., Me kling, P.J., 2020. E ec o la ice
sh inking on he mig a ion o wa e wi hin zeoli e LTA. Mic opo ous Mesopo ous
Ma e . 293, 109808. h ps://doi.o g/10.1016/j.mic omeso.2019.109808
Pé ez-U ia e, P., A eka, A., Aguayo, A.T., Gayubo, A.G., Bilbao, J., 2016. Kine ic
model o he eac ion o DME o ole ins o e a HZSM-5 zeoli e ca alys . Chem.
Eng. J. 302, 801–810. h ps://doi.o g/10.1016/j.cej.2016.05.096
Ra iee, A., Rajab Khalilpou , K., Milani, D., Panahi, M., 2018. T ends in CO2
con e sion and u iliza ion: A e iew om p ocess sys ems pe spec i e. J. En i on.
Chem. Eng. 6, 5771–5794. h ps://doi.o g/10.1016/j.jece.2018.08.065
Sánchez-Con ado , M., A eka, A., Aguayo, A.T., Bilbao, J., 2018a. Di ec syn hesis o
dime hyl e he om CO and CO2 o e a co e-shell s uc u ed CuO-ZnO-
Z O2@SAPO-11 ca alys . Fuel P ocess. Technol. 179, 258–268.
h ps://doi.o g/10.1016/j. up oc.2018.07.009
Sánchez-Con ado , M., A eka, A., Aguayo, A.T., Bilbao, J., 2018b. Beha io o SAPO-
11 as acid unc ion in he di ec syn hesis o dime hyl e he om syngas and CO2.
J. Ind. Eng. Chem. 63, 245–254. h ps://doi.o g/10.1016/j.jiec.2018.02.022
Sánchez-Con ado , M., A eka, A., Rod iguez-Vega, P., Bilbao, J., Aguayo, A.T., 2018c.
Op imiza ion o he Z Con en in he CuO-ZnO-Z O2/SAPO-11 Ca alys o he
Selec i e Hyd ogena ion o CO+CO2 Mix u es in he Di ec Syn hesis o Dime hyl
E he . Ind. Eng. Chem. Res. 57, 1169–1178.
h ps://doi.o g/10.1021/acs.iec .7b04345
Semelsbe ge , T.A., Bo up, R.L., G eene, H.L., 2006. Dime hyl e he (DME) as an
al e na i e uel. J. Powe Sou ces 156, 497–511.
49
h ps://doi.o g/10.1016/j.jpowsou .2005.05.082
Shimoda, N., Faungnawakij, K., Kikuchi, R., Eguchi, K., 2011. A s udy o a ious
zeoli es and CuFe2O4 spinel composi e ca alys s in s eam e o ming and hyd olysis
o dime hyl e he . In . J. Hyd ogen Ene gy 36, 1433–1441.
h ps://doi.o g/h p://dx.doi.o g/10.1016/j.ijhydene.2010.10.088
Sie a, I., E eña, J., Aguayo, A.T., A andes, J.M., Olaza , M., Bilbao, J., 2011. Co-
eeding wa e o a enua e deac i a ion o he ca alys me allic unc ion (CuO–
ZnO–Al2O3) by coke in he di ec syn hesis o dime hyl e he . Appl. Ca al. B
En i on. 106, 167–173.
h ps://doi.o g/h p://dx.doi.o g/10.1016/j.apca b.2011.05.021
Sie a, I., E eña, J., Aguayo, A.T., A eka, A., Bilbao, J., 2013. Kine ic modelling o he
dehyd a ion o me hanol o dime hyl e he o e γ-Al2O3. Chem. Eng. T ans. 32,
613–618. h ps://doi.o g/10.3303/ce 1332103
Sun, J., Yang, G., Yoneyama, Y., Tsubaki, N., 2014. Ca alysis chemis y o dime hyl
e he syn hesis. ACS Ca al. 4, 3346–3356. h ps://doi.o g/10.1021/cs500967j
Tian, Y., Demi el, S.E., Hasan, M.M.F., Pis ikopoulos, E.N., 2018. An o e iew o
p ocess sys ems enginee ing app oaches o p ocess in ensi ica ion: S a e o he a .
Chem. Eng. P ocess. - P ocess In ensi . 133, 160-210.
h ps://doi.o g/10.1016/j.cep.2018.07.014
Zhou, X., Su, T., Jiang, Y., Qin, Z., Ji, H., Guo, Z., 2016. CuO-Fe2O3-CeO2/HZSM-5
bi unc ional ca alys hyd ogena ed CO2 o enhanced dime hyl e he syn hesis.
Chem. Eng. Sci. 153, 10–20.
h ps://doi.o g/h ps://doi.o g/10.1016/j.ces.2016.07.007
