scieee Science in your language
[en] (orig)

Activity and stability of different Fe loaded primary catalysts for tar elimination

Author: Cortázar Dueñas, María,Álvarez Gordejuela, Jon,Olazar Barona, Leire,Santamaría Moreno, Laura,López Zabalbeitia, Gartzen,Villafan Vidales, Heidi Isabel,Asueta, Asier,Olazar Aurrecoechea, Martin
Publisher: Elsevier
Year: 2022
DOI: 10.1016/j.fuel.2022.123457
Source: https://addi.ehu.eus/bitstream/10810/56715/1/1-s2.0-S0016236122003222-main.pdf
Fuel 317 (2022) 123457
A ailable online 8 Feb ua y 2022
0016-2361/© 2022 The Au ho (s). Published by Else ie L d. This is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
Ac i i y and s abili y o di e en Fe loaded p ima y ca alys s o
a elimina ion
Ma ia Co aza
a
, Jon Al a ez
b
,
*
, Lei e Olaza
a
, Lau a San ama ia
a
, Ga zen Lopez
a
,
c
,
Heidi Isabel Villa ´
an-Vidales
d
, Asie Asue a
e
, Ma in Olaza
a
a
Depa men o Chemical Enginee ing, Uni e si y o he Basque Coun y UPV/EHU, P.O. Box 644-E48080, Bilbao, Spain
b
Depa men o Chemical and En i onmen al Enginee ing, Uni e si y o he Basque Coun y UPV/EHU, Nie es Cano 12, 01006 Vi o ia-Gas eiz, Spain
c
IKERBASQUE, Basque Founda ion o Science, Bilbao, Spain
d
Uni e sidad Nacional Au onoma de Mexico, Ins i u o de Ene gias Reno ables, Temixco, Mo elos, Mexico
e
GAIKER Technology Cen e, Basque Resea ch and Technology Alliance (BRTA), Pa que Tecnol´
ogico de Bizkaia, edi icio 202, E-48170 Zamudio, Spain
ARTICLE INFO
Keywo ds:
P ima y ca alys
Biomass gasi ica ion
Ta model compound
Ta elimina ion
Fluidized bed
Ca alys deac i a ion
ABSTRACT
The pe o mance o oli ine, dolomi e and γ-alumina p ima y ca alys s was e alua ed in he con inuous a
elimina ion p ocess in which oluene was selec ed as he biomass gasi ica ion a model compound. I on was
inco po a ed in o hese ca alys s in o de o imp o e hei ca aly ic ac i i y. All he expe imen s we e pe o med
in a con inuous low luidized bed mic o- eac o , wi h a s eam/ oluene a io o 4 and a space eloci y (GHSV) o
820 h
−1
, which co esponds o a ca alys amoun o 3.8 cm
3
. The e ec o empe a u e was s udied using oli ine
in he 800–900 ◦C ange, which allowed concluding ha 850 ◦C was he bes empe a u e o a emo al. The
esh and deac i a ed ca alys s we e cha ac e ized by N
2
adso p ion–deso p ion, X- ay luo escence (XRF), X- ay
di ac ion (XRD) and empe a u e-p og ammed oxida ion (TPO). Ta con e sion e iciency was assessed by
means o ca bon con e sion, H
2
yield (based on he maximum allowed by s oichiome y), gas composi ion and
p oduc yields, wi h Fe/Al
2
O
3
leading o he highes con e sion (87.6 %) and H
2
yield (38 %). Likewise, Fe/
Al
2
O
3
also p o ided he highes s abili y, as i allowed ope a ing o long pe iods wi h high con e sion alues
(85.9 % a e 35 min on s eam), al hough i unde wen se e e deac i a ion. The analysis o he spen ca alys s
e ealed ha deac i a ion occu ed mainly by coke deposi ion on he ca alys su ace and i on phase oxida ion,
wi h Fe/oli ine and Fe/dolomi e leading o he as e deac i a ion due o hei poo e me al dispe sion ela ed o
hei educed su ace a ea. The TPO p o iles showed ha he coke deposi ed on he h ee ca alys s was amo -
phous wi h a e y small con ibu ion o highly s uc u ed ca bon.
1. In oduc ion
Biomass gasi ica ion is conside ed one o he mos e icien ou es o
con e biomass eeds ock in o gaseous uel h ough a pa ial oxida ion
p ocess a high empe a u es [1,2]. Howe e , one o he main sho -
comings o biomass gasi ica ion lies in he p esence o a s in he p oduc
s eam, which leads o ouling/clogging and co osion o downs eam
equipmen [3–5]. Hence, in o de o minimize he amoun o a and
imp o e he syngas composi ion, i s ca aly ic con e sion is one o mos
p omising ou es [6,7]. This p ocess in ol es he oxida ion o he a
componen s using s eam o p oduce a syngas iche in H
2
and, u he -
mo e, he p esence o he ca alys allows a mo e e ec i e a emo al a
lowe empe a u es han hose in he non-ca aly ic a con e sion [8].
The a is as a complex mix u e o condensable hyd oca bons,
anging om single- ing o i e- ing a oma ic compounds along wi h
o he oxygen-con aining hyd oca bons and complex polycyclic a oma ic
hyd oca bons (PAH) [9]. Ta model compounds ha e been widely used
in o de o asce ain he ca alys pe o mance and de e mine sui able
ope a ing condi ions. In his wo k, oluene was selec ed as a a model
compound because i is a s able a oma ic s uc u e, especially a ela-
i ely low empe a u es, apa om being one o he majo a species in
he biomass gasi ica ion [10–14].
Many pa allel and consecu i e eac ions in ol e a con e sion, wi h
he p oduc dis ibu ion being he esul o hei compe i ion. The main
p oduc s ob ained a e hyd ogen, ca bon monoxide and ca bon dioxide,
and he majo eac ions occu ing in he p ocess a e as ollows:
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (J. Al a ez).
Con en s lis s a ailable a ScienceDi ec
Fuel
jou nal homepage: www.else ie .com/loca e/ uel
h ps://doi.o g/10.1016/j. uel.2022.123457
Recei ed 23 No embe 2021; Recei ed in e ised o m 21 Janua y 2022; Accep ed 26 Janua y 2022
Fuel 317 (2022) 123457
2
Toluene s eam e o ming
C
7
H
8
+7 H
2
O→7 CO+11 H
2
(1)
Toluene s eam dealkyla ion
C
7
H
8
+H
2
O→C
6
H
6
+CO+2 H
2
(2)
Wa e gas shi (WGS):
CO+H
2
O↔CO
2
+H
2
(3)
The mal c acking
C
7
H
8
→x C
n
H
m
+z H
2
(4)
C
6
H
6
→x C
n
H
m
+z H
2
(5)
Hyd odealkyla ion
C
7
H
8
+H
2
→C
6
H
6
+CH
4
(6)
Toluene d y e o ming
C
7
H
8
+7 CO
2
→14 CO+4 H
2
(7)
Boudoua d eac ion:
C+CO
2
↔2 CO (8)
Me hana ion
CO+3 H
2
↔CH
4
+H
2
O (9)
Coke he e ogeneous gasi ica ion:
C+H
2
O→CO+H
2
(10)
Amongs hese eac ions, he mos impo an ones a e s eam
e o ming (Eq. (1)) and wa e gas shi (WGS) (Eq. (3)).
The selec ion o empe a u e and ca alys de e mines he ex en o
hese eac ions and he selec i i y owa ds he di e en p oduc s
[7,15–18].
Ta ca aly ic con e sion me hods a e classi ied as in-si u (o p ima y)
and pos -gasi ica ion (seconda y) ones [14,19]. In he o me , he a
educ ion occu s du ing he gasi ica ion s age, wi h he ca alys being
loca ed in he gasi ie i sel . In he seconda y app oach, he gas p oduced
in he gasi ie is ea ed downs eam in a seconda y eac o whe e he
ca alys is placed. Rega dless he s a egy ollowed, essen ial aspec s
condi ioning he gasi ica ion p ocess a e hose in ol ing he eac o
con igu a ion, ope a ing condi ions and ype o ca alys [20]. Fluidized
bed eac o s a e one o he mos de eloped echnologies o biomass
gasi ica ion, which equi e an app op ia e ca alys in e ms o ac i i y
and s abili y in o de o educe he a con en o 2 g m
−3
and a oid he
need o a mo e expensi e seconda y ca aly ic eac o downs eam
[21,22].
A wide ange o ma e ials wi h signi ican ac i i y o c acking and
e o ming o hea y a oma ic compounds ha e been in es iga ed as
p ima y ca alys s [12,14]. Na u al mine als, such as oli ine and calcined
dolomi e, ha e been widely used in he s eam gasi ica ion in luidized
beds, as hey a e ac i e o a emo al, apa om being inexpensi e
and abundan [23]. Acid ca alys s, such as alumina o zeoli es, ha e also
been used (p io o and a e me al imp egna ion) as ca alys s o a
aba emen [24]. Ne e heless, he pe o mance o all hese p ima y
ca alys s can be g ea ly imp o ed by me al phase addi ion [19,25–28].
Thus, suppo ea u es, such as mechanical ( esis ance o a i ion),
physico–chemical (su ace a ea, po osi y, acidi y, composi ion and
densi y) and ca aly ic ones (ac i i y / selec i i y and s abili y) play a
ele an ole in he me al-suppo in e ac ions, as well as in he
e o ming eac ion mechanism i sel [29].
F om a ca aly ic poin o iew, nickel is known o be he mos
in e es ing me al phase o e o ming applica ions [30]. Ni-based ca a-
lys s ha e been widely applied in he s eam e o ming o biomass a s
due o bo h hei high ac i i y o b eaking C
–
C and O
–
H bonds and
pe o mance in e ms o H
2
p oduc ion [7,12,13,31–35]. Howe e , hei
main d awbacks a e ela ed o hei apid deac i a ion, mechanical
agili y and high cos compa ed o na u al mine als o alumina [36].
Cu en ly, use o i on as an ac i e phase is gaining mo e a en ion o a
educ ion due o i s lowe cos , abundance and lowe oxici y compa ed
o nickel [37]. I on is known o be an ac i e species o a oma ic hy-
d oca bon des uc ion (b eakage o C
–
C and C
–
H bonds), as well as o
he WGS eac ion. In ac , i has been p o en e ec i e o he a o e-
men ioned eac ions in di e en oxida ion s a es [24,38–40]. The e o e,
i on imp egna ion o oli ine, dolomi e o Al
2
O
3
seems o be in e es ing
o syn he izing in-bed p ima y ca alys s o gas–solid con ac eac o s,
such as luidized o spou ed beds, om bo h economic and en i on-
men al pe spec i es. Ne e heless, he lowe ac i i y o he i on species
wi h espec o he Ni ones equi es highe amoun s o dopan , gene ally
in he 10–30 w % ange [41].
Acco dingly, he aim o his wo k is o analyse he pe o mance o
oli ine, dolomi e and γ –Al
2
O
3
as p ima y ca alys s, as well as he e ec
he imp egna ion o each ca alys wi h 10 w % Fe has on he elimina ion
o oluene, which has been selec ed as he model compound o biomass
gasi ica ion a . Fu he mo e, a de ailed cha ac e iza ion o he esh
and deac i a ed Fe-doped ca alys s has been ca ied ou in o de o
de e mine he main deac i a ion mechanisms in his p ocess. The esul s
ob ained will p o ide essen ial in o ma ion o he selec ion o op imal
p ima y ca alys s o biomass gasi ica ion in he bench-scale uni
equipped wi h an imp o ed spou ed bed eac o de eloped by ou
esea ch g oup [21,42,43]. Fu he mo e, he esul s ob ained may also
be ex apola ed o indus ial gasi ica ion eac o s, which a e mainly
luidized beds. This s udy add esses mul iple aspec s ha ha e no been
join ly app oached in he li e a u e, as a e ca alys s p epa a ion and
cha ac e iza ion, in luence o empe a u e, ca alys pe o mance a ze o
ime on s eam, s abili y o Fe-loaded ca alys s and he main deac i a-
ion causes.
2. Expe imen al
2.1. Ca alys p epa a ion and cha ac e iza ion
Six ca alys s ha e been es ed in he oluene s eam e o ming p o-
cess. Th ee o hem (oli ine, dolomi e and γ –alumina) a e p ima y
ca alys s, whe eas he o he h ee a e hose ob ained by imp egna ing
he a o emen ioned p ima y ca alys s wi h Fe, i.e., Fe/oli ine, Fe/
dolomi e and Fe/Al
2
O
3
. Besides, uns wi h silica sand we e ca ied ou
o compa ison pu poses. Mine als Sibelco supplied he oli ine and
dolomi e, and Al a Aesa he γ-Al
2
O
3
. These h ee p ima y ca alys s
p o ided sa is ac o y esul s in a p e ious s udy o biomass gasi ica ion
in a oun ain con ined conical spou ed bed eac o (CSBR), as hey
allowed educing a o ma ion, as well imp o ing he yield and
composi ion o he syngas [21,42,43]. The ca alys pa icles we e sie ed
in o de o e ain he ac ions wi hin he anges o 90–150
μ
m o
oli ine, 150–250
μ
m o dolomi e and 250–400
μ
m o γ - Al
2
O
3
, which
allow a aining simila luidiza ion egimes wi h hese ma e ials o
di e en densi ies. P io o use, dolomi e was calcined a 900 ◦C o 4 h
in a mu le o en in o de o comple e he deca boxyla ion o calcium and
magnesium ca bona es.
The Fe loaded ca alys s we e p epa ed by we imp egna ion o he
suppo s wi h an aqueous solu ion o Fe(NO
3
)
3
⋅9H
2
O (Pan eac Appli-
Chem, 98 %). The amoun o saline p ecu so added was ha co e-
sponding o he desi ed inal ca alys composi ion. The concen a ion o
Fe was ixed a 10 w % in o de o compa e he ca aly ic ac i i y and
selec i i y o he h ee ca alys s o same amoun o me al loaded.
Subsequen o he imp egna ion p ocess, he p epa ed ca alys s we e
d ied a 100 ◦C o 24 h and hen calcined a 1000 ◦C o 4 h. Gi en ha
he ca aly ic ac i i y o i on species gene ally inc eases wi h hei
educ ion s a e (Fe
2
O
3
<Fe
3
O
4
<FeO <Fe(0)) [26], hese i on-
imp egna ed ca alys s we e used once hey had been subjec ed o an
M. Co aza e al.
Fuel 317 (2022) 123457
3
ex si u educ ion p ocess a 850 ◦C o 4 h unde 10 ol% H
2
s eam,
which ensu ed ull educ ion o e ic oxides in o hei me allic phase.
The pa icle sizes o he Fe loaded ca alys s we e he same as hose o
hei p ima y coun e pa s.
The physical p ope ies o he ca alys s we e de e mined by N
2
adso p ion–deso p ion in a Mic ome i ics ASAP 2010 ins umen . Based
on he in o ma ion o hese iso he ms, he ca alys s ea u es, such as
hose in ol ing speci ic su ace a ea and po ous s uc u e (a e age po e
size and po e olume), we e calcula ed by he B unaue –Emme –Telle
(BET) me hod. P io o he analysis, and in o de o emo e any impu-
i y, he samples we e degassed a 150 ◦C un il a p essu e below 2⋅10
−3
mmHg was eached. The chemical composi ion (w %) o each ca alys
was measu ed by X- ay luo escence (XRF). Mo e de ailed in o ma ion
abou he XRF me hodology can be ound elsewhe e [43].
The empe a u e-p og ammed educ ion (TPR) o he ca alys s was
ca ied ou in an Au oChem II 2920 Mic ome i ics, which allowed
de e mining he ca alys educ ion empe a u e be o e using i . This
me hod consis s in exposing he solid o a educing gas low o 10 ol%
H
2
/A , while empe a u e is inc eased wi h a cons an hea ing a e o
5 ◦C min
−1
om ambien one o 900 ◦C. The educ ion empe a u e o
each ca alys was asce ained by moni o ing he H
2
consumed.
The c ys alline s uc u e o he esh and deac i a ed ca alys s was
analyzed using X- ay powde di ac ion (XRD) pa e ns. A B uke D8
Ad ance di ac ome e wi h Cu K
α
1 adia ion was used o conduc XRD.
The de ailed p ocedu e ollowed is desc ibed elsewhe e [44]. The me al
c ys alli e size was calcula ed by using he Sche e o mula. Me al
dispe sion was calcula ed om me al c ys alli e size using he equa ion
D (%) =97.1/d (nm) and assuming ha he size o Fe a om is he same as
ha o Ni a om, as epo ed elsewhe e [45,46].
The alues o o al acidi y o he ca alys s ha e been ob ained by
moni o ing he di e en ial adso p ion o NH
3
a 150 ◦C using simul a-
neously calo ime y and he mog a ime y in a Se a am TG-DSC 111
equipmen .
The amoun o coke deposi ed on he used ca alys s was de e mined
by empe a u e-p og ammed oxida ion (TPO) in a he mobalance (TGA
Q5000TA The mo Scien i ic). This TGA is connec ed on-line o a Blaze
Ins umen s Mass Spec ome e (The mos a ) and he p ocedu e ol-
lowed o de e mine he coke deposi ed on each sample is as ollows: (i)
signal s abiliza ion wi h He s eam a 100 ◦C o 30 min, and (ii) a amp
o 5 ◦C min
−1
o 800 ◦C in a s eam o O
2
dilu ed in He, wi h his em-
pe a u e emaining cons an o 30 min in o de o ensu e ull coke
combus ion.
2.2. Expe imen al equipmen and p ocedu e
The expe imen s o oluene con e sion wi h he di e en ca alys s
we e pe o med in an Inconel luidized bed eac o (300 mm in leng h
and 10 mm in in e nal diame e ), as shown in Fig. 1. The eac o is
loca ed wi hin a adian o en, which p o ides he hea o ope a ing up
o 900 ◦C. The empe a u e was measu ed and eco ded by means o wo
K- ype he mocouples, wi h one being loca ed inside he eac o ,
app oxima ely in he middle zone o he bed, and he o he one close o
he wall o he elec ic o en.
The wa e o gene a ing he s eam and oluene we e in oduced by
means o a high-p essu e pump (ASI 521) and a sy inge pump (PHD
4400), espec i ely. Thei pumping low a es we e main ained cons an
in all he uns, wi h he alues being 0.24 mL min
−1
o wa e and 0.06
mL min
−1
o oluene, which co espond o a s eam/ oluene a io (S/T)
o 4 and a mola s eam/ca bon (S/C) a io o 3.35. P io o eeding in o
he eac o , hese wo compounds we e pumped sepa a ely in o an
e apo a ion sys em a 350 ◦C, which ensu es hei ull apo iza ion. This
plan is also p o ided wi h a ni ogen mass low me e (B ooks SLA5800)
ha allows eeding up o 1 L min
−1
. In ac , a ni ogen low a e o 300
mL min
−1
was used as luidizing agen du ing he hea ing p ocess p io
o he eac ion.
The gaseous s eam lea ing he eac o was passed h ough a hea e ,
whose empe a u e was kep a 300 ◦C in o de o p e en he conden-
sa ion o he p oduc s be o e en e ing he on-line analysis sys em. Then,
he ola ile s eam ci cula ed h ough a condensa ion de ice consis ing
o wo coalescence il e s, which ensu ed o al condensa ion and
e en ion o he non- eac ed s eam and oluene, as well as oluene
de i ed p oduc s.
This s udy deals wi h he e ec o ope a ing condi ions in a ca aly ic
p ocess o a elimina ion p ocess, i.e., e o ming empe a u e (in he
800–900 ◦C ange), ca alys s ype (oli ine, dolomi e and alumina, as
well as hei coun e pa s wi h Fe imp egna ion) and ca alys s abili y.
Oli ine was chosen o analyse he e ec o empe a u e, whe eas 850 ◦C
was es ablished as he sui able ope a ing empe a u e o s udy he in-
luence o ca alys ype and ime on s eam. The e ec o eac ion ime
was s udied o he Fe loaded ca alys s in he 5–115 min ange in o de
o assess he e olu ion o ca alys ac i i y and s abili y.
Gi en ha he densi y o he p ima y ca alys s di e s signi ican ly
(3300 kg m
−3
o oli ine, 1666 kg m
−3
o alumina and 1275 kg m
−3
o
dolomi e), and in o de o ope a e unde he same hyd odynamic con-
di ions in he luidized bed eac o , he same bed olume was used in all
expe imen s. Acco dingly, as men ioned abo e, sui able pa icle sizes
we e chosen. Thus, 3.8 cm
3
o he co esponding ca alys (o sand in case
he expe imen was ca ied ou wi hou ca alys ) we e placed in he bed
in all cases, co esponding o a gas hou ly space eloci y (GHSV) o 820
h
−1
. Expe imen s a ze o ime on s eam we e epea ed a leas 3 imes o
ensu e ep oducibili y o he esul s and he ca bon mass balance closu e
was abo e 95 % in all uns.
2.3. P oduc analysis
The analysis o he ola ile s eam lea ing he eac o was conduc ed
on-line by means o a GC (Agilen 7890) p o ided wi h a lame ioniza-
ion de ec o (FID). The sample was injec ed in o he GC p io o
condensa ion by means o a line main ained a 280 ◦C in o de o a oid
he condensa ion o hea y a compounds. The analysis o he non-
condensable gases (a e sepa a ing he a s om he gaseous s eam
in he condensa ion sys em) was ca ied ou by means o a mic o GC
Fig. 1. Schema ic diag am o he oluene s eam e o ming labo a o y uni .
M. Co aza e al.
Fuel 317 (2022) 123457
4
(Agilen 4900). The h ee independen modules wi h di e en columns
(molecula sie e, po apak and plo alumina) allowed iden i ying and
quan i ying he gaseous p oduc s p e iously calib a ed. This analysis
me hodology allowed a de ailed quan i ica ion o he en i e p oduc
s eam.
2.4. Reac ion indices
The con e sion and p oduc yields we e aken as eac ion indices o
moni o and assess p ocess pe o mance. The ca bon con e sion o
oluene was de ined as he moles o ca bon in he gaseous p oduc
s eam di ided by he moles o ca bon in he oluene eed (Eq. (11)).
No e ha he moles o CO, CO
2
and C
1
-C
4
hyd oca bons o med (co -
esponding o he o al amoun o ca bon moles in he gas) ha e been
de e mined om he mic o-GC analysis, whe eas he moles o ca bon in
he eed we e calcula ed based on he o al amoun o oluene in o-
duced in o he eac o ( o al olume o oluene injec ed in he un).
Ccon e sion(%) = moles o ca bon in he p oduc gas
moles o ca bon in he eed ⋅100 (11)
The p oduc yields we e calcula ed as he a io be ween he g ams o
each p oduc (H
2
, CO, CO
2
and CH
4
) in he gaseous s eam and he
g ams o he model compound in he eed:
Yield(w %) = g o he compound in he p oduc gas
g o model compound in he eed ⋅100 (12)
Mo eo e , H
2
po en ial was also de e mined as he a io be ween he
concen a ion o H
2
in he e luen gas and he maximum allowed by
s oichiome y:
H2po en ial(%) = moles o H2in he p oduc gas
maximum moles o H2allowed by s oichiome y (13)
The maximum numbe o H
2
moles allowed by s oichiome y was
calcula ed by conside ing oluene e o ming eac ion and ha o WGS.
Thus, H
2
po en ial is de ined based on he maximum numbe o H
2
moles
ob ained when oluene is ully e o med o CO
2
and H
2
.
3. Resul s and discussion
3.1. F esh ca alys cha ac e iza ion
Table 1 shows he physical p ope ies (speci ic su ace a ea, po e
olume and a e age po e diame e ) and chemical composi ion o he
p ima y ca alys s and hose imp egna ed wi h Fe. As obse ed, oli ine
has he lowes speci ic su ace a ea (1.92 m
2
g
−1
) and po e olume
(0.002 cm
3
g
−1
) due o i s non-po ous s uc u e. A e imp egna ion
wi h Fe(NO
3
)
3
⋅9H
2
O solu ion, he speci ic su ace a ea and he po e
olume o dolomi e and Al
2
O
3
dec eased mainly due o me al
deposi ion, as i blocks some o he mic opo es o he ca alys s. Ac-
co ding o Kuma e al. [47], he p esence o i on on alumina accele a es
he sh inkage o alumina and ans o ms he alumina om gamma in o
o he phases, which dec eases he su ace a ea because Fe
2
O
3
pa icles
ac as he e ogeneous nuclea ion si es o
α
-Al
2
O
3
pa icles a high em-
pe a u e. Ne e heless, he opposi e end was obse ed in he Fe/
oli ine, i.e., he speci ic su ace a ea inc eased due o he deposi ion o
Fe on he ex e nal su ace a ea, and he po e olume and a e age po e
size became la ge , which sugges s he collapse o he in e -po e s uc-
u e o oli ine. No e ha he same end has been obse ed o me al
imp egna ion on suppo s wi h low po osi y su aces [22,48,49]. Apa
om he imp egna ion p ocess, he high calcina ion empe a u e also
con ibu es o educing he BET su ace a ea and po osi y o he Fe/
Al
2
O
3
ca alys , al hough o a lesse ex en . In a p e ious s udy [50], he
same Al
2
O
3
used in his s udy was calcined wi h ai a 1000 ◦C du ing 5 h
and i s BET su ace a ea and po e olume educed o 87 m
2
g
−1
and 0.38
cm
3
g
−1
, espec i ely.
Dolomi e is a calcium magnesium ca bona e, i.e., CaMg(CO
3
)
2
, and
he e o e ca bona es a e decomposed in o CaO and MgO in he calci-
na ion, which a e he main cons i uen s in he calcined dolomi e, as
shown in Table 1. Mo eo e , he XRF e ealed ha he e is a high
con en o Fe in he Fe/oli ine. In ac , he con en o Fe in he aw
oli ine was o a ound 5.3 w % and a e imp egna ion, he Fe amoun in
he ca alys inc eased signi ican ly o 17 w %, which con i med ha he
me al con en was close o ha co esponding o he imp egna ion (~10
w %) plus ha in he o iginal oli ine. In he o he wo ca alys s, namely
Fe/Al
2
O
3
and Fe/dolomi e, he ini ial Fe con en was negligible and
a e he imp egna ion inc eased up o 9.9 and 9.3 w %, espec i ely.
Thus, he Fe con en o he h ee s udied ca alys s is consis en wi h he
a ge ed me al loading o 10 w %.
Table 2 shows he me al dispe sion o each ca alys which was es i-
ma ed based on he me al c ys alli e size ob ained by XRD analysis (by
applying Debye-Sche e equa ion). As obse ed, he highes me al
dispe sion is a ained o Fe/Al
2
O
3
(2.6 %), whe eas he poo es alue is
o dolomi e (0.5 %). This esul con i ms ha he physical s uc u e o
he suppo plays an essen ial ole in he dispe sion o he me al phase;
ha is, he suppo wi h he highes su ace a ea as ha o Al
2
O
3
leads o
he highes me al dispe sion.
Table 1
P ope ies o p ima y and Fe imp egna ed ca alys s.
Oli ine Fe/Oli ine γ-Al
2
O
3
Fe/Al
2
O
3
Calcined Dolomi e Fe /Dolomi e
Physical p ope ies
S
BET
(m
2
g
−1
) 1.92 3.75 100.00 12.48 17.42 3.55
V
po e
(cm
3
g
−1
) 0.002 0.017 0.42 0.059 0.05 0.009
d
po e
(Å) 78 234 167 206 113 162
Chemical p ope ies
MgO (w %) 48.79 36.98 – 0.23 43.61 32.15
SiO
2
(w %) 43.18 37.20 0.02 – 0.12 0.11
Fe
2
O
3
(w %) 7.68 24.39 – 14.13 0.02 13.21
CaO (w %) 0.12 0.17 – 0.14 56.07 48.53
Al
2
O
3
(w %) 0.04 0.43 99.98 81.02 0.15 0.26
Na
2
O (w %) 0.06 0.06 – – 0.01 0.03
TiO
2
(w %) 0.02 0.03 – 0.12 0.02 0.03
MnO (w %) 0.11 0.10 – – – –
Acidi y
To al acidi y (µmol NH
3
g
ca −1
) 2.4 8.8 80.0 11.4 8.7 10.5
Table 2
Me al dispe sion (%) calcula ed om me al c ys alli e o he h ee Fe imp eg-
na ed ca alys s.
d
Fe
XRD
a
(nm) Fe dispe sion (%)
Fe/Al
2
O
3
38 2.6
Fe/oli ine 68 1.4
Fe/dolomi e 193 0.5
a
Calcula ed om he ull wid h a hal heigh o he Fe
0
(110) di ac ion peak
a 2θ =44◦in he XRD p o iles using he Sche e equa ion.
M. Co aza e al.
Fuel 317 (2022) 123457
5
The XRD pa e ns o he p ima y ca alys s and Fe educed ca alys s
a e shown in Fig. 2a and 2b, espec i ely. As obse ed, he h ee Fe
doped ca alys s show an in ense peak o he me al i on phase a 2θ =44◦
and wo smalle ones a 2θ =65◦and 82◦. No e ha i on oxide phases
we e no de ec ed in hese ca alys s, which is e idence o hei ull
educ ion. In bo h oli ine (Fig. 2a) and Fe/oli ine (Fig. 2b), he main
c ys alline phases obse ed a e hose co esponding o oli ine
(Mg
1.81
Fe
0.19
⋅(SiO
4
)) and ens a i e (MgSiO
3
). Fu he di ac og am o
un educed Fe/oli ine ca alys can be ound elsewhe e [22]. Rega ding
Fe/dolomi e (Fig. 2b), apa om he me al i on phase, hose o Ca
(OH)
2
, CaO and MgO we e also obse ed, wi h all o hem being de i ed
om he calcina ion o calcium magnesium ca bona e, which is he main
mine al species in he dolomi e [51]. These las h ee phases (Ca(OH)
2
,
CaO and MgO) we e also obse ed in he XRD di ac og am o calcined
dolomi e (Fig. 2a). These alkaline ea h oxides (CaO and MgO) con-
aining Lewis basic si es may p omo e adso p ion and mig a ion o H
2
O
and OH g oups on he ca alys su ace, and he e o e p omo e ca bon
gasi ica ion and educe ca bon deposi ion [52]. The Ca(OH)
2
di ac ion
peaks a e e idence ha CaO (a highly hyg oscopic compound) abso bed
humidi y om he ambien and o med Ca(OH)
2
. In he Fe/Al
2
O
3
ca alys , ypical di ac ion peaks co esponding o he Al
2
O
3
suppo
we e de ec ed, as well as he cyni e (FeAl
2
O
4
), whose di ac ion lines a e
loca ed a 2θ =31◦, 36◦, 51◦, 59◦and 64◦. The high calcina ion
empe a u e used (1000 ◦C) allowed he o ma ion o he cyni e spinel
(FeAl
2
O
4
), which occu s a empe a u es abo e 600 ◦C by he in e ac ion
be ween Fe species (Fe
0
, FeO and Fe
3
O
4
) and Al
2
O
3
, ollowing he e-
ac ion mechanism epo ed in he li e a u e [53,54]. Mo eo e ,
compa ing he XRD pa e ns o Al
2
O
3
be o e and a e imp egna ion and
calcina ion s ages, he e is a phase change om γ-Al
2
O
3
o a mo e s able
one, which is p obably he mos s able one (
α
-phase) due o he high
empe a u e o calcina ion used (1000 ◦C). The peaks assigned o Al
2
O
3
in he Fe loaded ca alys in Fig. 2b a e clea and sha p, which is e idence
o i s high c ys alliza ion deg ee, whe eas he peaks assigned o Al
2
O
3
in
Fig. 2a a e b oad and low, he eby sugges ing an amo phous s uc u e
wi h a small c ys alliza ion deg ee. No e ha he same di ac ion peaks
han hose obse ed o Al
2
O
3
c ys alline phases in Fig. 2a and b ha e
been epo ed in he li e a u e and co espond o γ-Al
2
O
3
and
α
-Al
2
O
3
,
espec i ely [55,56]. The e o e, phase ans o ma ion is he conse-
quence o he he mal deg ada ion o he suppo , which a ec s
ad e sely he physical p ope ies o he ca alys by educing ca alys
su ace a ea, he eby educing ca alys ac i i y. Se e al au ho s ha e
called his p ocess suppo sin e ing [35,57].
The empe a u e p og ammed educ ion (TPR) p o iles o calcined
Fe/oli ine, Fe/dolomi e and Fe/Al
2
O
3
ca alys s a e shown in Fig. 3.
Gi en ha me al i on is expec ed o be he ac i e phase o b eaking
C
–
C and C
–
H bonds [24,58], he educibili y o he ca alys s is o g ea
Fig. 2. XRD pa e ns o p ima y ca alys (a) and Fe imp egna ed ones (b). C ys alline phases: (+) (Mg
1.81
Fe
0.19
(SiO
4
)); (o) Ens a i e (MgSiO
3
); (∇) Fe
0
; (◆) Al
2
O
3
;
(□) MgO; (•) CaO; (❖) He cyni e (FeAl
2
O
4
); (✦) Calcium hid oxide (CaOH
2
).
M. Co aza e al.

Fuel 317 (2022) 123457
6
ele ance. Acco ding o he li e a u e [26,59] he educ ion o Fe
2
O
3
gene ally p oceeds in wo s eps, as a e: he educ ion o Fe
2
O
3
o Fe
3
O
4
in he 350–500 ◦C ange and he educ ion o Fe
3
O
4
o me al Fe in he
500–900 ◦C ange. Howe e , acco ding o ce ain s udies, he in e me-
dia e FeO is o med in he educ ion om Fe
3
O
4
o Fe
0
[60,61]. These
wo egions associa ed wi h wo o h ee educ ion s eps om Fe
2
O
3
a e
obse ed in he h ee educed ca alys s, al hough di e ences in he in-
e ac ions be ween he i on and he suppo s shi ed he loca ion o he
peaks. In he TPR p o ile o he Fe/oli ine, a b oad educ ion zone be-
ween 350 and 700 ◦C is obse ed wi h 3 peaks. The i s wo (a 470 and
530 ◦C) a e associa ed wi h he educ ion o Fe
2
O
3
and Fe
3
O
4
/FeO,
espec i ely, whe eas he la e peak abo e 600 ◦C is due o he Fe a oms
ha mig a ed in o he oli ine suppo o o m a e y s able MgFe
2
O
4
spinel phase [62]. Peaks a 380 ◦C and 500 ◦C appea in he Fe/dolomi e,
which a e cha ac e is ic o i on species educ ion, bu he e is also a
b oad peak a 750 ◦C, which co esponds o he educ ion o Fe
3+
om
he calcium i on oxide (s eb odolski e, Ca
2
Fe
2
O
5
) o Fe, as was sug-
ges ed by Zamboni e al. [63,64]. These au ho s obse ed he o ma ion
o his phase when i on ni a e was used in he we imp egna ion o
dolomi e. In his s udy, no e idences o Ca
2
Fe
2
O
5
a e obse ed in he
XRD di ac og am (Fig. 2), p obably due o i s low c ys allini y
.
In he
Fe/Al
2
O
3
ca alys , apa om he wo peaks iden i ied a 380 and
580 ◦C, which a e associa ed wi h he educ ion o i on species (Fe
2
O
3
,
Fe
3
O
4
and FeO) , a hi d b oad educ ion zone appea s be ween 700 and
900 ◦C, which is a ibu ed o he educ ion o i on alumina es
(FeAl
2
O
4
), also iden i ied in he XRD spec a [65]. Di e en au ho s
sugges ed ha he p esence o alumina s abilizes Fe
2
O
3
phase and he
educ ion goes h ough he o ma ion o FeAl
2
O
4
spinel, whose educ-
ion occu s abo e 700 ◦C [66,67].
3.2. Role o empe a u e in he a con e sion on oli ine ca alys
The in luence o empe a u e on oluene aba emen on oli ine ca -
alys s is displayed in Fig. 4. Fig. 4a shows he e olu ion o ca bon con-
e sion and H
2
po en ial. As obse ed, empe a u e has a g ea in luence
on ca bon con e sion and H
2
po en ial, since hei alues inc ease om
3.6 and 2.6 % a 800 ◦C o 46.0 and 23.6 % a 900 ◦C, espec i ely. This
inc ease in bo h pa ame e s is a ibu ed o he endo he mic na u e o
he oluene e o ming eac ions, as well as o hose in ol ing decom-
posi ion and dehyd ogena ion, as all o hem a e p omo ed a high
empe a u es [68]. The same end o ca bon con e sion and H
2
po-
en ial wi h empe a u e on oli ine ca alys s was obse ed by o he
au ho s in he a s eam e o ming [58,69].
The yields o he compounds in he p oduc s eam is displayed in
Fig. 4b. An inc ease in empe a u e leads o highe yields in gaseous
compounds (including benzene) due o he p omo ion o bo h e o ming
and c acking eac ions, wi h he highes yields being hose o CO and
CO
2
a 900 ◦C (49.8 and 26.8 w %, espec i ely). The yield o CH
4
in-
c eases wi h empe a u e, bu i is lowe han 2.2 w % a he h ee
empe a u es s udied. I should be no ed ha he yield o C
2
-C
4
hyd o-
ca bons is ha dly no iceable (below 0.01 w %), and has no he e o e
been included in Fig. 4b. CH
4
is mainly o med om dealkyla ion o he
me hyl g oup in he oluene s uc u e and, o a mino ex en , om he
me hana ion o CO [8]. Howe e , s eam e o ming o CH
4
p e ails o e
hese eac ions, since he con en o CH
4
in he p oduc s is e y low [12].
The p esence o an undesi ed compound (benzene) is due o incomple e
decomposi ion o oluene [70], which is con i med in Fig. 4b, whe e
benzene yield inc eases om 0.6 w % a 800 ◦C o 12.6 w % a 900 ◦C a
he expense o a dec ease in oluene yield. Se e al eac ions, such as
s eam dealkyla ion (Eq. 2), he mal c acking (Eqs. 4–5) o hyd o-
dealkyla ion o oluene (Eq. 6) lead o he o ma ion o benzene (all o
hem enhanced a high empe a u es) [11,15,17,71]. Howe e , he
small amoun o CH
4
in he p oduc s eam is e idence ha hyd o-
dealkyla ion eac ion (Eq. 6) is no signi ican [72]. I should be no ed
ha he benzene p oduced om he a o emen ioned eac ions can un-
de go e o ming eac ions o p oduce u he CO and H
2,
al hough hese
eac ions a e limi ed due o benzene s abili y [11,12]. The yield o
polycyclic a oma ic hyd oca bons (PAHs, e e ed o he compounds
hea ie han oluene) also inc eases wi h empe a u e due o he p o-
mo ion o condensa ion eac ions o ligh e a s. Howe e , he low yield
o hese PAHs (below 1.2 w % in he whole ange o empe a u es
s udied) is e idence ha he ex en o hese eac ions is almos negli-
gible, p obably due o he p esence o s eam [21,73]. Swie czynski e al.
[3] also obse ed a yield o a ound 6 w % o benzene and 14 w % o
polya oma ics in he p oduc s eam o oluene s eam e o ming a
850 ◦C when hey used oli ine as p ima y ca alys .
Fig. 4c displays he gas composi ion in he 800–900 ◦C ange. I can
be obse ed ha he e ec o empe a u e on he gas composi ion is no
e y p onounce abo e 850 ◦C, i.e., he concen a ion ha dly changes
abo e his empe a u e. Be ween 800 and 850 ◦C, ce ain ends a e
obse ed when empe a u e is inc eased, as a e: a sligh dec ease in H
2
and CO
2
concen a ions ( om 69.1 o 66.2 ol% and om 8.1 o 6.8 ol
%, espec i ely) and an inc ease in ha o CO ( om 21.8 o 25.5 ol%).
This esul is explained by he p omo ion o he e e se WGS eac ion
due o i s exo he mic na u e. The same end wi h empe a u e was
obse ed in o he s udies o ca aly ic e o ming o a model compounds,
wi h his e ec being a ibu ing o he exo he mic na u e o he WGS
eac ion [68,74].
3.3. Compa ison o p ima y ca alys s pe o mance
In o de o s udy he pe o mance o p ima y ca alys s, oluene
con e sion on oli ine, dolomi e and alumina was moni o ed a 850 ◦C
and he esul s ob ained a e displayed in Fig. 5. The e ec o he mal
c acking was asce ained by compa ing he esul s o ca bon con e sion
(Fig. 5a), p oduc yields (Fig. 5b) and concen a ion o gaseous
Fig. 3. TPR p o iles o Fe/Al
2
O
3
, Fe/dolomi e and Fe/oli ine ca alys s.
M. Co aza e al.
Fuel 317 (2022) 123457
7
compounds (Fig. 5c) ob ained wi h he ca alys s and hose ob ained wi h
ine sand. As obse ed, he p esence o any ca alys imp o es he
o e all e iciency o he p ocess by inc easing ca bon con e sion and he
yields o gaseous compounds, especially hose o H
2
, CO and CO
2
, as well
as educing ha o oluene. This imp o emen o e he esul s ob ained
wi h ine sand is associa ed wi h he p omo ion o s eam e o ming (Eq.
1), c acking (Eqs. 4–5) and WGS eac ions (Eq. 3). The p esence o
p ima y ca alys s also p omo es s eam dealkyla ion (Eq. 2) and he mal
c acking (Eq. 4) eac ions, since he concen a ion o benzene in he
p oduc s eam is highe han ha ob ained wi h sand.
Compa ing he e iciency o he p ima y ca alys s (Fig. 5a), Al
2
O
3
leads o he highes con e sion (58.4 %) ollowed by dolomi e (39.1 %).
Howe e , he H
2
po en ial wi h bo h ca alys s is simila (28.5 % o
Al
2
O
3
and 28.9 % o dolomi e). This la e esul can be explained by
he lowe ac i i y o Al
2
O
3
and he highe o dolomi e in he WGS. Thus,
he highe ac i i y o dolomi e in he WGS eac ion is ela ed o CaO and
MgO basic si es, wi h ac i i y being highe as he Ca/Mg a io is
inc eased [75,76]. Fu he mo e, he p esence o CaO and MgO also
explains he highe yield o benzene a he expense o lowe ing ha o
oluene [77,78]. Mo eo e , oli ine has he smalles in luence on he
oluene s eam e o ming, since i p o ided he lowes ca bon con e sion
and H
2
po en ial alues. In his case, al hough he p esence o Fe p o-
mo es e o ming eac ions, he low BET su ace a ea (1.91 m
2
g
−1
) and
po e olume (0.002 g cm
−3
) a e he ac o s leading o he low e iciency
o his ca alys in he oluene elimina ion p ocess. S udies epo ed in he
li e a u e con i m ha dolomi e and Al
2
O
3
we e mo e ac i e han
oli ine o educing he amoun o a de i ed om biomass gasi ica ion,
as he ex en o he WGS eac ion is enhanced wi h dolomi e [43,79].
3.4. E ec o Fe inco po a ion in o he p ima y ca alys s
Fig. 6 compa es he pa ame e s in ol ing oluene con e sion (ca -
bon con e sion and H
2
po en ial (a), p oduc yields in he ou le s eam
(b) and he concen a ion o gaseous compounds (c)) o he Fe loaded
ca alys s. Fig. 6a e eals ha Fe inco po a ion in o he p ima y ca alys s
leads o highe ca bon con e sion and H
2
po en ial han hose on he
p ima y ca alys s in all cases, Fig. 5a, which is e idence o hei highe
ca aly ic ac i i y o oluene e o ming. Thus, on he one hand, i is well
s ablished ha me al i on is ac i e o C
–
C and C
–
H bond b eakdown,
which enhances hyd oca bon e o ming and c acking eac ions [58,80].
On he o he , he addi ion o Fe p omo es he WGS eac ion because he
adso p ion o wa e molecules on he ca alys ac i e si es is a ou ed,
hus leading o highe H
2
yields [81]. This imp o emen is especially
ema kable wi h oli ine, whose ca bon con e sion and H
2
po en ial
inc eases om 18 and 10.5 % o 73 and 31.9 %, espec i ely.
As occu ed wi h p ima y ca alys s, ha o Fe/Al
2
O
3
p o ided he
bes esul s in e ms o ca bon con e sion (87.6 %) and H
2
po en ial (38
%) (Fig. 6a). Howe e , he ends we e e e sed o Fe/oli ine and Fe/
dolomi e a e Fe inco po a ion, a aining highe ca bon con e sion in
he o me . This esul is closely ela ed o he change in he su ace a ea
o he ca alys s caused by he imp egna ion, which de ini ely a ec s
me al dispe sion. As obse ed in Table 1, he su ace a ea inc eased in
he oli ine when Fe was in oduced, whe eas i signi ican ly dec eased
in he dolomi e ( om 17.42 o 3.55 m
2
g
−1
). Fu he mo e, he esul s in
Fig. 4. E ec o empe a u e on ca bon con e sion and H
2
po en ial (a), p oduc yields and un eac ed oluene ac ion in he ou le s eam (b), and concen a ion o
he gaseous s eam (c).
M. Co aza e al.
Fuel 317 (2022) 123457
8
Table 2 con i m he be e dispe sion o Fe on he oli ine han on he Fe/
dolomi e, which sugges s ha he ac i e si es a e mo e accessible o he
eac an s in he o me , as all he i on is loca ed on he ca alys su ace.
This implies a highe ca aly ic ac i i y o Fe/oli ine, which explains he
be e esul s o ca bon con e sion on his ca alys han on Fe/dolomi e,
whose me al dispe sion is he poo es .
Compa ing he p oduc yields shown in Fig. 6b, he highes yields o
CO and CO
2
(mainly de i ed om he e o ming and WGS eac ions,
espec i ely) and benzene (a c acking p oduc ) a e ob ained on he Fe/
Al
2
O
3
ca alys , whe eas ha o oluene is he lowes (below hal o hose
ob ained on Fe/oli ine o Fe/dolomi e). No e ha Fe ac s as he ac i e
phase o he e o ming and WGS eac ions, whe eas he alumina sup-
po p o ides he acidi y equi ed o c acking eac ions, i.e., he com-
bina ion o bo h p o ides Fe/Al
2
O
3
ca alys wi h he highes ac i i y o
hese eac ions. Mo eo e , a compa ison o Fig. 6b wi h Fig. 5b shows
ha he yield o benzene inc eases g ea ly when Fe is added o he
p ima y ca alys s. I seems ha he p esence o Fe mainly ca alyzed he
con e sion o oluene o benzene. Some s udies sugges ed ha empe -
a u es highe han 800 ◦C inc ease he hyd odealkyla ion ac i i y o he
s eam e o ming o oluene on i on-based ma e ials [72,82], whe eas
o he esea ches concluded ha he ac i i y o i on-based ma e ials
leads o he decomposi ion o la ge a compounds in o small agmen s
o ca bon species, which subsequen ly o m benzene [83]. The e o e, i
can be concluded ha he highe benzene con en is a combined e ec o
c acking and hyd odealkyla ion o oluene molecules on Fe ac i e si es.
The highe CH
4
yields obse ed on Fe loaded ca alys han on p ima y
ca alys s also con i ms his hypo hesis.
Fig. 6c displays he gas composi ion ob ained wi h he h ee Fe-
imp egna ed ca alys s. A compa ison o hese esul s wi h hose o
p ima y ca alys s (Fig. 5c) shows he ele ance o me al i on in he WGS
eac ion (Eq. 3), since he concen a ion o CO
2
g ea ly inc eased in all
he cases, whe eas ha o CO educed. This is consis en wi h p e ious
s udies in he li e a u e, in which a high ac i i y o Fe is epo ed in he
WGS eac ion [38,40,84]. Analysing Fe loaded ca alys s, Fe/Al
2
O
3
led o
he lowes concen a ion o CH
4
and CO
2
and he highes o H
2
and CO,
which is e idence o a high ex en o s eam and d y e o ming o hy-
d oca bons (Eq. 1 and 7). Acco ding o Adnan e al. [85,86], his ac is
a ibu ed o he basic si es o Fe/Al
2
O
3
ca alys s, which p omo e
endo he mic CO
2
e o ming o hyd oca bons.
The di e ences obse ed among hese Fe-imp egna ed ca alys s a e
he consequence o a ious ac o s. As p e iously s a ed, one o he mos
in luen ial ac o is ela ed o he me al dispe sion on he ca alys sup-
po , which plays a key ole in he ini ial ca alys ac i i y. A sui able
me al-suppo in e ac ion enhances he mig a ion o me al c ys alli es,
he eby ob aining a be e dispe sion o Fe on he suppo [26].
Fu he mo e, he physical s uc u e o he suppo g ea ly in luences he
dispe sion o he me al phase, as shown in Table 2, in which he highes
Fe dispe sion was ob ained o Al
2
O
3
( he suppo wi h he highes BET
su ace a ea and po e olume). The esul s in Fig. 6 con i m ha he
be e su ace p ope ies o he Al
2
O
3
suppo p omo e he dispe sion o
Fig. 5. E ec o p ima y ca alys s on ca bon con e sion and H
2
po en ial (a), p oduc yields and un eac ed oluene ac ion in he ou le s eam (b), and concen-
a ion o gaseous compounds (c).
M. Co aza e al.
Fuel 317 (2022) 123457
9
he ac i e phase, and he e o e lead o highe ca alys ac i i y. Besides,
he highe dispe sion o Fe on oli ine also explains he highe ca bon
con e sion han on Fe/dolomi e.
Ano he ac o is ela ed o he ac i i y o he suppo o c acking
and/o e o ming eac ions, which is di ec ly linked o i s acidi y
[30,57]. Thus, he po ous s uc u e o oli ine and dolomi e ba ely ha e
mic o o mesopo es, whe eas alumina has a mo e de eloped po ous
s uc u e, as shown in Table 1. Adnan e al. [85] sugges ed ha oluene
con e sion eac i i y is domina ed by s ong acid si es in he ca alys ,
which a e di ec ly a ached o he su ace o he ca alys . Thus, a highe
su ace a ea o he ca alys inc eases he numbe o s ong si es a ailable
o con ac wi h oluene, he eby leading o a highe acidi y o he ca -
alys s, and consequen ly o a highe con e sion o oluene, as is he case
o Fe/Al
2
O
3
, which has he highes acidi y (Table 1) o he h ee Fe
loaded ca alys s [87]. Besides, Adnan e al. [88] s a ed ha a highe
con en o Fe in he ca alys also p omo es ca alys acidi y, and he e o e
oluene con e sion. Compa ing he acidi y o p ima y and Fe doped
ca alys s (Table 1), he p esence o Fe inc eases he acidi y o Fe/oli ine
and Fe/dolomi e ca alys s om 2.4 o 8.8 and om 8.7 o 10.5 µmol NH
3
g
ca −1
, espec i ely, which explains he highe oluene c acking capa-
bili y o Fe doped ones. Rega ding he acidi y alue o Fe/Al
2
O
3
(11.4
µmol NH
3
g
ca
−1
), i is much lowe han ha o he aw γ-Al
2
O
3
. Indeed,
as p e iously s a ed, he educ ion in BET su ace a ea caused by he
calcina ion and imp egna ion s ages leads o he blockage o some po es
and educes he numbe o acid si es a ailable, hus educing he o al
acidi y o he ca alys . Howe e , compa ing Fig. 5b and 6b, benzene
yield is highe when Fe/Al
2
O
3
is used han when he p ima y Al
2
O
3
is
used, which sugges s ha he c acking ac i i y o Fe/Al
2
O
3
is highe .
This esul is explained by he combina ion o wo issues. On he one
hand, as was p e iously s a ed, he be e pe o mance o Fe o
e o ming and WGS eac ions leads o highe H
2
pa ial p essu es in he
eac ion en i onmen , hus p omo ing hyd odealkyla ion eac ions (Eq.
6) which lead o highe benzene con en s. On he o he hand, he eal
acidi y o γ-Al
2
O
3
unde eac ion condi ions is much lowe han ha
gi en in Table 1, as he high empe a u es used in his s udy (850 ◦C)
and he p esence o s eam accele a e he collapse o he po ous s uc u e
and he ans o ma ion o γ-alumina in o o he mo e s able phases, as
s a ed elsewhe e [47]. Thus, he blockage o po es and he ans-
o ma ion o γ-phase in o o he ones (δ, θ o
α
) educes he numbe o
acid si es a ailable, and he e o e i s c acking ac i i y.
O he impo an issue in ol ing ca aly ic ac i i y is he educ ion
s a e o he i on species, wi h ac i i y being highe as Fe species a e
u he educed (me al Fe is he mos ac i e phase). Thus, he XRD
pa e ns in he h ee esh ca alys s e eal he p esence o me al Fe,
whe eas he p esence o o he species wi h di e en educ ion s a es,
such as Fe
2
O
3
, Fe
3
O
4
o FeO, was no ini ially obse ed (Fig. 2b). This is
an e idence ha he di e ence in he ca aly ic ac i i ies o Fe imp eg-
na ed ca alys s is mos ly a ibu ed o he in e ac ions be ween he me al
i on and he suppo s, as well as hei physical s uc u e. Thus, he be e
p ope ies o Al
2
O
3
(i ac s as a ex u al p omo e p e en ing he as
sin e ing o he i on me al, as well as s abilizing ac i e si es on i s su -
ace) lead o be e dispe sion o he Fe oxide phase, and he e o e be e
Fig. 6. E ec o Fe imp egna ed ca alys s on ca bon con e sion and H
2
po en ial (a), p oduc yields and un eac ed oluene ac ion in he ou le s eam (b), and
concen a ion o gaseous compounds (c).
M. Co aza e al.