Formation of weak and strong Br?nsted acid sites during alkaline treatment on MOR zeolite

Article abstract of DOI:10.1016/j.apcata.2016.08.004

Aluminum rich mordenite zeolite (Si/Al?=?10) was hierarchized by using mild (0.2?M NaOH and 70?°C) and severe (0.4?M NaOH and 85?°C) alkaline treatments. The textural characterization was made by nitrogen adsorption. Acidic properties, i.e. nature (N), concentration (C), strength (S), accessibility (A) and location (L), were drawn from adsorption of two probes molecules followed by infrared spectroscopy: pyridine for N, C, L and S properties while CO for N and S properties. The catalytic model reaction of m-xylene transformation was also used to determine C, S and L properties. The severe alkaline treatment promoted (i) silicon extraction of zeolite framework, without considerable loss of crystallinity, (ii) creation of intra-crystal mesopore and widening of micropore diameter due to formation of interconnection between main canals, and (iii) accessibility of Br?nsted acid sites (BAS) located in side pockets. The silicon removal is inevitably followed by formation of extraframework aluminum species (EFAL) that leads to formations of Lewis acid sites, weak extraframework BAS and very strong BAS due to interaction between EFAL with the bridged hydroxyl, evidenced by a correlation between turnover frequency in m-xylene transformation and concentration of LAS.

Full text of DOI:10.1016/j.apcata.2016.08.004

Applied Catalysis A: General 526 (2016) 95–104
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Formation of weak and strong Brønsted acid sites during alkaline
treatment on MOR zeolite
N. Chaouatia,b, A. Soualah , I. Hussein , J-D. Comparot , L. Pinard
b
a
a
a,∗
a
Université de Poitiers, CNRS UMR7285, Institut de Chimie des Milieux et Matériaux de Poitiers, B27, TSA 51106, 4 rue Michel Brunet, 86073 Poitiers CEDEX
9
, France
b
Laboratoire de Physico-Chimie des Matériaux et Catalyse, Faculté des Sciences Exactes, Université de Bejaia, 06000, Bejaia, Algerie
a r t i c l e i n f o
a b s t r a c t
◦
Article history:
Received 25 May 2016
Received in revised form 28 July 2016
Accepted 3 August 2016
Available online 4 August 2016
Aluminum rich mordenite zeolite (Si/Al = 10) was hierarchized by using mild (0.2 M NaOH and 70 C) and
◦
severe (0.4 M NaOH and 85 C) alkaline treatments. The textural characterization was made by nitrogen
adsorption. Acidic properties, i.e. nature (N), concentration (C), strength (S), accessibility (A) and location
L), were drawn from adsorption of two probes molecules followed by infrared spectroscopy: pyridine
for N, C, L and S properties while CO for N and S properties. The catalytic model reaction of m-xylene
(
transformation was also used to determine C, S and L properties. The severe alkaline treatment promoted
Keywords:
(i) silicon extraction of zeolite framework, without considerable loss of crystallinity, (ii) creation of intra-
Desilication
Hierarchization
CO
Pyridine
M-xylene isomerization
crystal mesopore and widening of micropore diameter due to formation of interconnection between main
canals, and (iii) accessibility of Brønsted acid sites (BAS) located in side pockets. The silicon removal is
inevitably followed by formation of extraframework aluminum species (EFAL) that leads to formations of
Lewis acid sites, weak extraframework BAS and very strong BAS due to interaction between EFAL with the
bridged hydroxyl, evidenced by a correlation between turnover frequency in m-xylene transformation
and concentration of LAS.
©
2016 Published by Elsevier B.V.
1
. Introduction
only observed on mildly steamed zeolites, since more severe dealu-
mination leads to a considerable diminution of acid sites and hence
to activity [11]. It has been proposed that this higher activity results
from interplay between active sites rather than merely a decrease
in the concentration of framework aluminum [12]. Henriques et al.
[13], explain the higher activity in o-xylene transformation on dea-
luminated modrenite by (i) the increase of accessibility of acid sites,
owing to the presence of secondary mesoporous system and by (ii)
the generation of very strong Brønsted acid sites (BAS) by interac-
tion of Lewis acid sites with BAS [14,15]. Role of extra-framework
aluminum (EFAL) was evidenced by showing on series of realumi-
nated mordenites that supplementary activity is generated when
introduced Al is on non-framework position [16]. More than that,
Hong et al. [17] demonstrate that there is a correlation between the
initial rates of o-xylene conversion per Brønsted acid site and the
number of available Lewis acid sites.
The other post-synthesis method, more efficient to generate
mesoporosity is desilication by alkaline treatment. Contrary to dea-
lumination studies, there is less attention concerning the impact
of such modification on acidity. In the majority of cases, catalytic
activity data is usually interpreted in term of textural modifica-
tions by admitting minor modification of the acidity [18]. Only the
increase in acid density and the decrease in acid strength owing to
Zeolites are some of the most catalysts used in industrial pro-
cess (oil refining, petrochemicals and chemicals manufacturing)
and their applications grow steadily [1–3]. In addition to a unique
regular microporosity and strong acid sites, their properties can be
controlled either during their synthesis or by post-synthesis treat-
ments [4]. Among the main used zeolites, mordenite is used in some
major petrochemical processes such as isomerization of C8 aro-
matics, hydroisomerization of light alkanes and cumene synthesis
[
5–7].
Dealumination post-synthesis treatment is usually performed
on zeolites to increase stability, activity, thermal stability and
control concentration of Brønsted acid sites (BAS) [8]. Zeolite dea-
lumination is a simple and efficient tool to adjust meso/micropore
network [7] in detriment to acidity owing to Al removal. Impact of
dealumination on zeolite acidity has been widely studied during the
last decades. Enhancement in activity on steamed mordenites was
observed by Mirodatos and Barthomeuf [9,10]. This benefit effect is
∗ Corresponding author.
E-mail address: ludovic.pinard@univ-poitiers.fr (L. Pinard).
http://dx.doi.org/10.1016/j.apcata.2016.08.004
0
926-860X/© 2016 Published by Elsevier B.V.

9
6
N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
Table 1
2
.3. Acidity characterization by adsorption of probe molecule
Acid and alkaline treatment conditions.
followed by IR
Treatment
Acid
HCl
Alkaline
NaOH
The possible modifications in acidic properties of starting
material (nature, concentration and strength) caused by the dif-
ferent treatments were studied by using adsorption of two probes
molecules followed by infrared spectroscopy (FTIR). Adsorption of
Agent
NaOH
Concentration (M)
Temperature ( C)
Time (min)
0.5
30
10
0.2
70
120
0.4
85
120
◦
◦
pyridine at 150 C measures the amounts of Brønsted and Lewis
−
1
−1
sites by integrating the band areas at 1545 cm and 1454 cm
respectively and by using extinction coefficients given in Ref. [21].
◦
Adsorption of carbon monoxide at −100 C evaluates acid strength
diminution of Si/Al molar ratio were highlighted [18]. Paixaoa et al.
19], explained the decrease on m-xylene conversion on deslicated
mordenite, by a loss of concentration of acid sites. But the same
authors, in another study, observed no change in acidic properties
after alkaline treatment of mordenite [20].
The aim is to study the impact of desilication on the acidity
of mordenite zeolite. For this purpose, acidity will be character-
ized by classical physiochemical methods such adsorption of probe
molecules (Pyridne and CO) followed by infra-red spectroscopy and
model reaction (xylene transformation).
of hydroxyl group by measuring value of OH band red-shift [22].
The infrared spectra were recorded in a Nicolet Magna 550-FT-IR
[
−1
spectrometer with a 2 cm optical resolution. The zeolites were
first pressed into self-supporting wafers (diameter: 1.6 cm, 18 mg)
◦
and pretreated from room temperature to 500 C (heating rate of
◦
−6
1
.5 C/min for 5 h under a pressure of 10 Torr) in a IR cell con-
nected to a vacuum line.
2.4. Acidity characterization by model reaction: m-xylene
transformation
The conversion of m-xylene (Aldrich, 99% pure) was performed
◦
at atmospheric pressure in a fixed bed reactor at 350 C, with a
2
. Experimental
N /reactant molar ratio of 13.5 and a contact time (1/WHSV) of
2
0
.006 h. Before testing, catalysts were compacted, crushed and
2
.1. Alkaline and acid treatments of HMOR zeolite
sieved to obtained homogeneous particles (0.2–0.4 mm). Samples
◦
−1
were pre-treated at 450 C under air flow (60 ml min ) over night.
Reaction products were analyzed online by gas chromatography
using a VARIAN 3800 gas chromatograph equipped with a FID
detector using a 60 m fused silica J&W DB WAX capillary column.
The zeolite used as a starting material was a commercial pro-
tonic mordenite (MOR, Süd chemie) with a molar ratio Si/Al of 10.
Alkaline treatment was carried out using NaOH as desilicating agent
following operating conditions reported in Table 1. Briefly, 10 g of
MOR were stirred at 400 rpm in 300 ml alkaline solution (0.2–0.4 M)
at 70 C and 85 C respectively for 2 h. Then, zeolite suspension was
centrifuged (6000 rpm) after being cooled down in ice–water mix-
ture. The decantation of the liquid was followed by re-suspension
with demineralized water. This procedure was repeated until pH
was 7. Samples were subsequently treated with 0.5 M of HCl under
◦
◦
2.5. Spent catalyst characterization
Coke amount was determined with a SDT Q600 TA thermo-
◦
◦
−1
gravimetric analyzer by heating up to 900 C (20 C min ) under a
1
N2 adsorption and residual acidity by pyridine adsorption followed
by FTIR.
−
00 ml min (O 10%/He) flow. Residual porosity was measured by
2
1
◦
stirring at 30 C for 10 min using the proportion of 10 ml per gram
of sample (Table 1). All samples were converted into their corre-
spondent protonic form by three consecutive exchanges with 2 M,
3
3. Results and discussion
NH NO3 using the proportion: 20 cm /g of zeolite. The exchange
4
◦
was carried out at 60 C for 4 h. After a careful wash up to pH 7, solid
was calcined under air flow (100 ml min /g of zeolite) at 550 C for
−1
◦
3.1. Physicochemical characterization
5
h.
The physicochemical properties of the parent zeolite (MOR) and
its derivatives (crystal size, crystallinity, external and BET surfaces,
micropore and mesopore volumes) are quoted in Table 2. Sam-
ples are designated as follows: the values in brackets are sodium
2.2. Physicochemical characterization
◦
hydroxide concentration (M) and temperature ( C) used for alkaline
X-ray diffraction (XRD) data were acquired on a PAN analyti-
treatment; the subscript:-ac indicates a subsequent acid leaching.
The powder XRD patterns, displayed in Fig. 1, show the preserva-
tion of mordenite structure upon alkaline and acid treatments. On
the other hand, desilication led to a loss of crystallinity which can be
estimated (from the ratio between the areas of all diffraction peaks
of the samples and of the starting material) to ca 20 and 30%. The
crystallinity slightly increased after acid leaching owing to the dis-
solution of amorphous species created during alkaline treatment
◦
◦
cal (X’Pert Pro MPD) diffractometer over a 5 –50 2␪ range using
the CuK␣ radiation (␣ = 154.05 pm). Samples were characterized
by transmission electronic microscopy (TEM) using a Philips CM
1
20 microscope equipped with a LaB filament. Chemical composi-
6
tion of samples was measured by X-ray fluorescence (XRF) and by
using inductively coupled plasma–atomic emission spectroscopy
(
ICP-AES) on an Optima 4300 DV (Perkin–Elmer). External Si/Al
molar ratio was measured by XPS using a Kratos Axis Ultra DLD
instrument.
A Micromeritics ASAP 2000 gas adsorption analyzer was used for
nitrogen sorption measurements. Prior to measurement, samples
were outgassed at 90 C for 1 h and 350 C for 10 h. Specific surfaces
areas were determined from the BET equation. Total pore volume
corresponds to the nitrogen volume adsorbed at P/P = 0.96 and t-
(EFAL). The nitrogen adsorption-desorption isotherm of starting
material displays an isotherm of type I (characteristic of a micro-
porous structure) and a hysteresis loop at high relative pressure
(
P/P > 0.9). This small hysteresis is due to mesopores formed by the
0
◦
◦
agglomeration of zeolite crystals [19]. Indeed, TEM micrograph on
Fig. 3 shows that MOR crystals are strongly aggregated and that the
void between them can be assimilated to inter-granular mesopores.
The mild desilication conditions were not enough to create signifi-
cant change in porosity (Table 2). On the other hand, both a higher
0
plot method was used to distinguish micropores from mesopores.
Mesopore size distribution was estimated by the BJH method.

N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
97
Table 2
Textural and structural properities of MOR series.
unity
MOR
MOR-ac
MOR-al(0.2/70)
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
Crystal sizea
Crystallinity
nm
%
100–500
100
524
23
0.20
0.20
0.06
–
100–500
100
530
17
0.20
0.21
0.05
–
100–500
78
544
72
0.19
0.22
0.14
4.20
5
100–500
81
649
85
0.21
0.24
0.12
4.20
5
–
72
–
77
b
c
2
−1
−1
SBET
Sext
m
m
cm
cm
cm
g
g
544
130
0.17
0.20
0.23
5.25
<10
558
170
0.18
0.21
0.26
5.25
<10
d
2
d
3
3
3
−1
−1
−1
V ultramic.
g
g
g
e
Vtotalmic.
f
Vmeso
g
Dp (BJH)
nm
nm
h
Dp (TEM)
–
–
a
Apparent size determined by using SEM and TEM images (Fig. 3).
calculated from XRD.
Specific surface area measured by BET.
External surface and micropore volume using t-plot method.
Total microporous volume estimated by DR method.
Mesopore volume = Vtotal–Vmicro (Vtotal: determined from the adsorbed volume at p/p0 = 0.96).
Pore size distribution estimated by BJH method and.
b
c
d
e
f
g
h
according to TEM images (Fig. 3).
MOR-al(0.4/85)ac
MOR-al(0.4/85)
MOR-al(0.2/70)ac
MOR-al(0.2/70)
a.u.
MOR-ac
MOR
5
10
15
20
25
30
35
40
2
theta
Fig. 1. X-ray diffraction patterns of MOR series.
temperature and a higher NaOH concentration were necessary to
generate a well pronounced hysteresis loop, which is typical of
presence of mesopores (Fig. 2). During alkaline treatment, the most
important textural modification is the generation of mesopores
decreased after alkaline treatment owing to a preferential silicon
extraction, and increased again after acid leaching due to a selective
dissolution of extra-framework Al species (EFAL). The differences
between global and external molar Si/Al ratios were similar with
all samples (except with MOR-al(0.2/70)ac)), meaning firstly an
homogenous distribution of Al on the starting material and sec-
ondly that silicon extraction occurs both on external surface and
inside zeolite crystal. The low value of (Si/Al)surf of MOR-al(0.4/85)
is probably due to a large amount of amorphous alumina on exter-
nal surface as indicates the too low global molar Si/Al ratio (6.0).
The framework Si/Al molar ratios drawn from TOT bands (SSI, [20]),
were similar whatever the used treatments, as a consequence the
theoretical concentrations of Brønsted acid sites were close on all
3
−1
(
Vmeso = 0.26 cm g ), that automatically leads to an increase of
2
−1
external surface (Sext = 170 cm g ). TEM images in Fig. 3 confirm
the presence of intra-crystal mesopores with an average diame-
ter, estimated by BJH method, to ca 5 nm (Table 2 and Fig. 2). The
mesopore distribution seems more homogeneous by using a severe
alkaline treatment than a mild one. The used operating condition
for acid leaching (Table 1) did not change the textural properties
of the starting material (Table 2); it is too mild to cause a dealu-
mination of the MOR zeolite, but sufficient to dissolve amorphous
species formed during alkaline treatment.
−
1
samples, i.e. ∼1300 ␮mol g . On the other hand, the number of
EFAL atom per unit cell of zeolite was function on the used proto-
col: alkaline treatment caused an increase of EFAL up to 2.7, while
the subsequent acid leaching dissolved them.
◦
3
.2. Acidity characterization by pyridine adsorption at 150 C
The hydroxyl stretching vibration region of IR spectra before
The acidic properties of zeolites were characterized by XRF,
ICP, XPS, FTIR and adsorption of pyridine followed by FTIR. Their
main features are summarized in Table 3. Global molar Si/Al ratio
(
continuous line) and after (dotted line) adsorption of pyridine
−
1
is given in Fig. 4. MOR exhibits an intense band at 3607 cm

9
8
N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
MOR
(
a)
MOR-ac
0
0
0
.9
.6
.3
0
MOR-al(0.2/70)ac
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
1
.2
1
(b)
0
3
6
9
12
15
MOR-al(0.4/85)ac
MOR-al(0.4/85)
MOR-al(0.2/70)ac
Dp (nm)
0
0
0
0
.8
.6
.4
.2
0
MOR-al(0.2/70)
MOR-al
MOR
0
0.2
0.4
0.6
0.8
1
P/P
0
Fig. 2. (a) BJH pore size distribution of different samples (shift 0.1), (b) N2 adsorption (open symbols) and desorption (solid symbols) isotherms at 77 K of MOR series
isotherms are shifted by +100).
(
Table 3
Acidic proprieties of MOR series.
unity
MOR
MOR-ac
MOR-al(0.2/70)
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
Si/Alglobala
Si/Alframework
Si/Al external
molar
molar
molar
wt.%
␮mol g
atom per cell
␮mol g
␮mol g
%
10.0
11.6
12.4
0.04
1329
0.6
1056
31
75
10.3
11.5
–
0.05
1333
0.4
968
38
78
8.1
11.3
8.4
0.04
1352
1.4
621
100
93
9.2
6.0
10.7
5.7
0.04
1417
2.7
337
133
65
9.8
b
11.9
15.7
0.05
1289
1.0
680
84
95
10.9
10.9
0.04
1333
0.4
533
101
80
c
a
[
[
Na]
+
d
−1
H ]theoretical
e
EFAL
+
f
−1
−1
[
[
PyH ]
PyL]f
g
PyAcc
a
Measured by XRF and ICP.
b
c
d
e
f
−1
According to TOT band at 1080–1200 cm using the correlation given in ref [20].
Determined by XPS.
Calculated from framework Al.
Extra-framework Al calculated from a and b.
Measured by pyridine adsorption on Brønsted (PyH ) and Lewis (PyL).
Drawn from subtraction of IR spectra of OH band before and after pyridine adsorption.
+
g
attributed to bridging hydroxyls groups (i.e. acidic hydroxyl
ble one part of BAS located in the side pockets, but a subsequent
acid leaching is necessary in order to totally beneficiate of this
supplementary accessibility. Indeed, EFAL species generated dur-
ing alkaline treatment can block access of pyridine to these new
sites and even to those located in the main channel (Table 3). The
concentration of Brønsted and Lewis acid sites (BAS and LAS) were
carried out using the IR bands 1545 and 1454 cm (Table 3). On
MOR sample the number of BAS probed by pyridine corresponds
to theoretical acidity multiplied with the percentage of accessible
acidic hydroxyl groups. On the other hand, some important dis-
crepancies between the theoretical and measured acidities (even
in taking account the crystallinity) appear after alkaline treatment.
These differences cannot simply explain by a blockage of access
to BAS by EFAL species, but by a modification of acidic strength
of Brønsted acid sites, as that was proposed by some authors
[19,25].
−1
groups) and a small band at 3744 cm
assigned to stretching
vibrations of terminal silanol (SiOH). Alkaline treatment led to a
decrease of the intensity of the main band concomitant both of
an increase of SiOH band and of the appearance of a small band
−
1
at 3660 cm , assigned to EFAL species. The decrease of intensity
of bridging hydroxyls groups could be due in part to the loss of
zeolite crystallinity (20 and 30% as shown previously). After pyri-
dine adsorption, the bands of SiOH and EFAL species remain almost
unchanged for all samples, as these groups do not present an acid
character. On the other hand, the main band associated to Brønsted
acid sites (BAS) decreases significantly but not totally. 75% of acid
sites were probed by pyridine (Pyacc, Table 3), suggesting that 3/4
of acid sites are located in the main channels and 1/4 in the side
pockets of MOR structure; this repartition is in accordance with
the open literature [23,24]. The alkaline treatment makes accessi-
−
1

N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
99
Fig. 3. TEM micrographs of MOR series.

1
00
N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
tion of Paukstis and Yurchenko [28]; the higher the proton affinity
value, the weaker the strength of acid sites. Thus on MOR sample,
A
0.5
−
1
OH are stronger (PA = 1145 kJ mol ) than OH (PA = 1262 kJ mol
1
1
2
2
−
1
)
. The value of PA2 is similar to that found on a silica-alumina
3
744
3
607
[
29]. The strengths of OH1 and OH2 stay unchanged regardless of
3
660
−1
treatment carried out (PA ∼ 1150 kJ mol ), contrariwise to their
1
MOR-al(0.4/85)ac
MOR-al(0.4/85)
proportions; the percentage of OH1 decreases after alkaline treat-
ment but it is restored after acid leaching. This show that EFAL
species formed during desilication are not only Lewis acid sites but
also have OH group with acid properties similar to silica alumina.
3
.4. Characterization of acidity by model reaction: m-xylene
conversion
The effect of alkaline treatment on the acidic properties of MOR
was evaluated by catalytic reaction of transformation of m-xylene,
which is commonly used as a model reaction to characterize acid-
ity [30]. At 350 C, the main reactions are m-xylene monomolecular
MOR-al(02./70)ac
◦
isomerization (I) onto o- and p-xylene and disproportionation (D)
into toluene (T) and trimethylbezene (TMB). The side reactions
are transalkylation between alkylaromatics as indicated by pres-
ence of low amounts of benzene and tretramethylbezene (TeMB),
and bimolecular m-xylene isomerization whose contribution in
isomers products can reach 16%. Fig. 5 shows the evolution of
conversion, and molar yield into I and D products as a function
of time-on-stream (TOS). All samples follow the same pattern of
deactivation, i.e. a fast initial decrease of the conversion related
to the decrease of yield into D products, the yield into I products
staying quasi stable. The active sites in disproportionation are sen-
sible to deactivation contrariwise to these in isomerization. The D
reaction passes through the formation of a reactional intermediate,
biphenylmethane, known to be a coke precursor [31] (Fig. 6).
All samples display same initial conversion (ca 40%) except on
MOR-al(0.2/70) (Table 5). The initial activities per accessible acid
sites probed by pyridine, commonly named: turn over frequency
MOR-al(02./70)
MOR
3
800 3700 3600 3500 3400
-1
Wavenumbers (cm )
Fig. 4. IR spectra of MOR series before (continuous line) and after (dotted line)
◦
pyridine adsorption at 150 C.
(TOF), increase after desilication and decrease after the subsequent
3
.3. Characterization of acidity by CO adsorption at low
acid leaching. Such changes cannot be related to modification in
textural properties of materials but in the strength of acid sites.
TOF is both proportional and inversely proportional to the con-
centrations of LAS and BAS, respectively (Fig. 7). Hong et al. [17],
also showed that initial activity per Brønsted site was proportional
to the number of Lewis sites and they concluded to a synergic
enhancement of the acidity by interaction between Brønsted and
Lewis centers. This synergic effect is related to intensification of the
strength of BAS [17]. Characterization of acidity by model reaction
allow to highlight the exaltation of BAS by EFAL at the differ-
ence of the adsorption of probe molecules followed by infrared
spectroscopy, and more particular CO which is a technique more
adapted to the characterization of weak acid sites.
The Table 6 displays the initial product distribution at iso-
conversion (X = 60%) on the MOR catalyst series. Initial selectivity
in isomers, expressed by I/D ratio, reduced on all treated materials.
Whatever post-treatment carried out, p/o and T/TMB ratios stay
constant and close to 1, implying no diffusion limitation of xylene
isomers and that T and TMB results from xylene disproportionation.
Relative amounts of the bulkiest isomers (1–3 and 1,3,5-TMB) gives
information on the available space near acidic sites owing to steric
constraint for the formation of corresponding transition states [32].
Initial ratio of 1–3/1,3,5-TMB slightly decreases after a mild alkaline
treatment, stays constant after a subsequent acid leaching, indicat-
ing a moderate modification of microporous shape. A more severe
desilication condition causes an increase of ratio between TMB iso-
mers, owing to the diminution of vacant space near the acid site
by EFAL species; their dissolution by acid washing leads to a strong
decrease of TMB ratio up to a value obtained on FAU zeolite (0.27)
[32]. In addition to the formation of mesopores, severe alkaline
temperature
Carbon monoxide is a slightly basic molecule of electronic
structure, which is extremely sensitive not only to the chemical
properties of the adsorption centres but also to their geometrical
configuration [26]. The small kinetic diameter of CO allows prob-
ing all BAS on the MOR zeolite structure, even those located on
side pockets. This probe molecule, ideal to study the strength of
zeolite acid site [27], forms hydrogen bond with acidic OH groups
of zeolite through the lone pair of electrons located on carbon
atom. Resulting elongation of O H is observed in infrared spec-
trum as lowering of the stretching frequency of hydrogen-bonded
OH group, which, in turn depends primarily on the intrinsic acid-
ity of particular hydroxyl group [28]. Fig. 5a presents subtracted
IR spectra obtained after successive introduction of CO doses: 50,
1
00, 200, 300, 400, 500 mbar, and just before appearance of CO in
gas phase i.e. saturation of the zeolite surface. On the MOR sam-
ple, interaction of CO with SiOH are limited, while with bridging
hydroxyl groups that leads to a decrease of intensity of band at
−
1
3
610 cm
concomitant of the appearance of a broad band cen-
−1
tered around 3300 cm . The deconvolution of this band, shown
in Fig. 5b, allows discriminating two types of BAS. A simple rea-
soning would be to assign the first red-shift to interaction of CO
−1
(
ꢀnOH1 = 322 cm , Table 4) with OH located in main channel and
−1
the second (ꢀꢁOH2 = 175 cm ) with OH located in the side pockets.
But, the proportion of these two families (9/1) is 3 times higher than
this found with pyridine (3/1), that means that the two families can-
not be classified in terms of location but of strength. It is possible
to calculate protonic affinity (PA) from ꢀꢂOH by using the equa-

N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
101
3
292
MOR
(b)
A
0.4
3
439
3
292
(a)
3
439
3
314
Δυ
1
Δυ
3600
2
3
800
3400
3200
3000
2800
-1
Wavenumbers (cm )
3
299
MOR-al(0.4/85)
MOR-al(0.2/70)
3304
(
b)
3473
b)
(
3
454
A
0.2
A
0.2
3
304
3
299
3473
(a)
(a)
3
454
3
610
3
613
Δυ
1
Δυ
1
Δυ
2
Δυ
2
3
800
3600
3400
3200
3000
2800
3
800
3600
3400
3200
3000
b)
2800
Wavenumbers (cm-1)
Wavenumbers (cm-1
)
MOR-al(0.2/70)ac
3302
MOR-al(0.4/85)ac
3403
3303
(
(b)
A
0.2
3
447
A
0.2
3462
3403
3303
3
302
(
a)
462
(
a)
3
3
447
3
611
3
612
Δυ
1
Δυ
1
Δυ
2
Δυ
2
3
800
3600
3400
3200
3000
2800
3800
3600
3400
3200
3000
2800
Wavenumbers (cm-1
)
Wavenumbers (cm-1
)
Fig. 5. IR bands resulting from CO adsorption at low temperature of MOR series: (a) subtracted spectra with background after adsorption of gradual amounts of CO 50, 100,
00, 300 400, 500 mbar and up to saturation (b) deconvolution of perturbed OH band after CO adsorption.
2
Table 4
Values of OH groups frequency downshift (ꢀꢂ) due to interaction with CO and proton affinities of different population of protonic sites.
0(cm 1
−
)
ꢂ
1(cm
−1
)
ꢂ
2(cm
−1
)
ꢀꢂ1(cm
−1
)
ꢀꢂ2(cm
−1
)
PA1a(kJ mol−1
)
PA2a(kJ mol−1
)
A1b
%
A2b
%
ꢂ
MOR
3614
3613
3612
3610
3611
3292
3299
3302
3304
3303
3439
3454
3447
3473
3462
322
314
310
306
308
175
159
165
137
149
1145
1150
1152
1154
1153
1262
1280
1273
1309
1293
89
75
86
60
78
11
25
14
40
22
MOR-al(0.2/70)
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
a
Proton affinity of the different population of sites calculated according to: PA = 2254.8–442.5 log((ꢀꢂ).
b
Population of acid sites (A1 and A2) corresponding to low (ꢀꢂ1 = ꢂ0–ꢂ1) and high (ꢀꢂ2 = ꢂ0–ꢂ2) downshift of OH band estimated after deconvolution of band centered
−
1
at 3300 cm
.

1
02
N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
60
50
40
30
20
10
0
MOR
0
20
40
60
TOS (min)
60
50
40
30
20
10
0
MOR-al(0.2/70)
60
50
40
30
20
10
0
MOR-al(0.4/85)
0
20
40
60
0
20
40
60
TOS (min)
TOS (min)
60
50
40
30
20
10
0
MOR-al(0.2/70)ac
60
50
40
30
20
10
0
MOR-al(0.4/85)ac
0
20
40
60
0
20
40
60
TOS (min)
TOS (min)
Fig. 6. m-Xylene conversion (♦) and yields in disproportionation (ꢀ) and isomerization (ᮀ) products as a function of time on stream.
Table 5
Conversion, activity and turn-over-frequency of transformation of m-xylene on the fresh (time-on-stream, TOS = 2 min) MOR catalyst series.
MOR
MOR-al(0.2/70)
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
X (mol%)
A (mol. h
41.6
0.96
908
56.9
1.50
2049
43.4
1.01
1488
37.5
0.84
2484
39.6
0.92
1739
−
1
−1
g )
−
1
TOF (h
)
Table 6
Initial products distribution at ca 60% conversion on the fresh (time-on-stream, TOS = 2 min) MOR series.
MOR
MOR-al(0.2/70)
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
␶
(h−1)
0.012
61.3
0.006
56.9
0.012
57.8
0.012
53.7
0.012
58.9
X (mol%)
Distribution (mol.%)
Benzene
1.6
1.1
2.0
1.2
1.5
Toluene (T)
p-xylene
o-xylene
TMB
TeMB
25.5
23.7
22.6
24.3
2.1
23.1
24.5
23.2
25.4
2.7
25.7
21.8
21.0
26.4
2.9
23.2
23.5
22.9
26.2
2.9
23.8
22.8
21.2
28.5
2.2
Isom. (%)
p/o
Disp. (%)
I/D
46.3
1.04
53.6
0.86
1.06
0.34
14.8
47.7
1.05
52.3
0.91
0.90
0.33
13.4
42.9
1.04
58.1
0.75
0.97
0.32
15.6
46.4
1.03
54.6
0.86
0.88
0.35
16.6
44.0
1.07
56.0
0.78
0.83
0.27
12.0
T/TMB
1
,2,3/1,3,5-TMB
a
Bimol. isom. (%)
a
Contribution of bimolecular isomerization, estimated by method described in Ref. [31]. TMB and TeMB: trimethylbenzene and tetramethylbenzene respectively.

N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
103
lesser extent by a subsequent acid leaching (Tox < 1). On hierarchi-
cal MOR zeolites, coke is located both on mesopores and micropores
a)
3
2
2
1
1
000
500
000
500
000
(Table 7).
MOR-al(0.4/85)
MOR-al(0.2/70)
4. Conclusion
Desilication of mordenite zeolite with Si/Al ratio of 10 is
performed by using sodium hydroxide. To obtain a significant
mesoporosity with preservation of both crystallinity and micro-
MOR-al(0.4/85)ac
MOR-al(0.2/70)ac
MOR
◦
porosity, severe conditions are necessary (0.4 M and 85 C, 2 h).
Partial silicon extraction leads to formation of extra-framework
aluminum species (EFAL) that a simple mild acid leaching can dis-
solve them. Concentration of Brønsted (BAS) acid sites, probed by
5
00
0
◦
0
200
400
600
800
1000
1200
pyridine at 150 C, decreases after an alkaline treatment. That is
+
-1
not only due to a blocking effect of EFAL species, since after acid
leaching BAS concentration is partially restored. The presence of the
extraframework aluminum and also silicon induced the formation
[
PyH ] (µmol g )
◦
of weak OH bridged evidenced by CO adsorption at −100 C.
3
2
2
1
1
000
500
000
500
000
b)
Reactivity of m-xylene reveals that presence of EFAL reduces the
space close to the acid sites, and also increases the strength of BAS.
The synergic effect between BAS and LAS is demonstrated by the
proportionality between concentration of LAS and TOF. Coke forma-
tion deactivates exclusively the reaction demanding more space as
disproportionation reactions.
MOR-al(0.4/85)
MOR-al(0.2/70)
MOR-al(0.4/85)ac
MOR-al(0.2/70)ac
Acknowledgment
MOR
5
00
0
The authors would like to acknowledge the support provided by
Algerian and French governments for funding this work through
project Tassili No. 12-MDU/859. The support of Faculty Exact
Sciences, Abderrahmane Mira University of Bejaia is gratefully
acknowledged.
0
20
40
60
80
100
120
140
-1
[
PyL] (µmol g )
Fig. 7. Turnover frequency in m xylene conversion as function of concentration of
Brønsted (a) and Lewis acid (b) sites.
References
[
1] J. Cˇ ejka, Introduction to Zeolite Science and Practice, In: H. van Bekkum, A.,
Corma, F. Schüth (Eds.), Elsevier Amsterdam, Studies in Surf. Sci. Catal. 168
(2007) 1.
treatment widens the micopores of mordenite zeolite (both large
canals and side pockets). Indeed N adsorption shows formation of
2
supermicropores (Table 2).
[2] W. Vermeiren, J.-P. Gilson, Top. Catal. 52 (2009) 1131.
[
3] M. Guisnet, J.-P. Gilson (Eds.), Zeolites for Cleaner Technologies, Imperial
College Press, London, 2002.
Main properties of spent catalyst are given in Table 6 (i.e. coke
content, residual porosity residual acidity). The amount of coke
formed on catalysts after 1 h of reaction is between 5 and 7 wt.%.
On MOR and MOR-al(0.4/85) catalysts, 70% of initial acid sites are
inaccessible to pyridine against the half on the other samples. The
toxicity of coke (Tox), number of acid sites deactivated per coke
molecules, can be estimated from the coke content and resid-
ual acidity and by assuming an average molar weight of coke of
[
4] V. Valtchev, G. Majano, S. Mintova, J. Pérez-Ramírez, Chem. Soc. Rev 42 (2013)
263.
[
[
[
5] M. Guisnet, P. Magnoux, Appl. Catal. 54 (1989) 1.
6] A. Chica, A. Corma, J. Catal. 187 (1999) 167.
7] T. Terlouw, J-P. Gilson, European Patent 458 378 (1991), to Shell Int. Res. Mij.
B.V.
[
8] M.-C. Silaghi, C. Chizallet, P. Raybaud, Microporous Mesoporous Mat. 191
(
2014) 82.
[
9] C. Mirodatos, D.J. Barthomeuf, J. Chem. Soc. Chem. Commun. (1981) 39.
−
1
2
00 g mol (value base on previous studies). On the starting mate-
[10] C. Mirandatos, B.H. Ha, K. Otsuka, D. Barthomeuf, Proceedings of the 5th
International Conference on Zeolites, In: Rees, (ed.), Heyden, London, 382
rial, Tox is higher than 1, meaning that deactivation mode occurs by
pore blockage. A simple mild alkaline treatment leads to a change in
the deactivation mode: poisoning (Tox = 1). This effect is accentu-
ated by using more severe desilication operating conditions and to a
(
1980).
[
11] R. Szostak, Introduction To Zeolite Science and Practice, In: H. van Bekkum,
E.M. Flanigen, P.A. Jacobs, J.C. Jansen (eds.), Elsevier Amsterdam, Stud. Surf.
Sci. Catal. 137 (2001) 261.
Table 7
Properties of spent catalysts.
Unity
MOR
MOR-al(0.2/70)
MOR-al(0.2/70)ac
MOR-al(0.4/85)
MOR-al(0.4/85)ac
Coke
wt.%
␮mol g
␮mol g
7.1
309
36
0.30
0.94
2.08
0.15
0.06
0.75
0.95
6.0
314
75
0.51
0.75
1.01
0.13
0.11
0.69
0.78
6.8
367
64
0.53
0.76
0.94
0.16
0.12
0.76
1
5.7
92
86
0.27
0.64
0.86
0.10
0.21
0.64
0.90
5.8
+
−1
[
[
[
[
PyH ]
PyL]
PyH ]/[PyH ]0
298
100
0.56
0.99
0.80
0.17
0.21
0.95
0.80
−
1
+
+
–
–
–
cm
cm
–
PyL]/[PyL]0
a
Tox
Vmic
Vmes
3
3
−1
−1
g
g
(
(
V/V0)mic
V/V0)mes
–
a
+
+
+
−1
Toxicity of coke Tox = (1-([PyH ]/[PyH ]0))/([coke]/[PyH ]0), [coke] is molar concentration of coke estimated by supposing an average molar weight of coke of 200 g mol .

1
04
N. Chaouati et al. / Applied Catalysis A: General 526 (2016) 95–104
[
[
12] D. Barthomeuf, J. Phys. Chem 83 (1979) 249.
13] C.A. Henriques, J.L.F. Monteiro, P. Magnoux, M. Guisnet, J. Catal. 172 (1997)
36.
14] A. Corma, In: Zeolites: Facts, Figures, Future, P.A., Jacobs, R.A. Van Santen
eds.), Elsevier Amsterdam, Stud. Surf. Sci. Catal. 49 (1989) 49.
[24] M. Maache, A. Janin, J.C. Lavalley, E. Benazzi, Zeolites 15 (1995) 507.
[25] K. Góra-Marek, K. Tarach, J. Tekla, Z. Olejniczak, P. Ku s´ trowski, L. Liu, J.
Martinez-Triguero, F. Rey, J. Phys. Chem. C 118 (2014) 28043.
[26] S. Bordiga, C. Lamberti, F. Bonino, A. Travert, F. Thibault-Starzyk, Chem. Soc.
Rev 44 (2015) 7262.
4
[
(
[
[
[
[
[
15] A. Corma, H. Garcia, Chem. Rev. 102 (2002) 3837.
16] P. Wu, T. Komatsu, T. Yashima, J. Chem, S.O.C. Faraday Trans 92 (5) (1996) 861.
17] Y. Hong, V. Gruver, J.J. Fripat, J. Catal. 150 (1994) 421.
18] D. Verboekend, J. Pérez-Ramirez, Catal. Sci. Technol. 1 (2011) 879.
19] V. Paixaoa, R. Monteiro, M. Andrade, A. Fernandes, J. Rocha, A.P. Carvalho, A.
Martins, Appl. Catal. A 402 (2011) 59.
[27] N.S. Nesterenko, F. Thibault-Starzyk, V. Montouillout, V.V. Yuschenko, C.
Fernandez, J.-P. Gilson, F. Fajula, I.I. Ivanova, Microporous Mesoporous Mater.
71 (2004) 157.
[28] E.A. Paukstis, E.N. Yurchenko, Usp. Khim. 52 (1983) 426.
[29] M. Rozwadowski, J. Datka, M. Lezanska, J. Wloch, K. Erdmann, J. Kornatowski,
Phys. Chem. Chem. Phys. 3 (2001) 5082.
[
20] V. Paixão, P. Carvalho, J. Rocha, A. Fernandes, A. Martins, Microporous
[30] M. Guisnet, N.S. Gnep, S. Morin, Microporous Mesoporous Mater. 35–36
(2000) 47.
Mesoporous Mater. 131 (2010) 350.
[
[
[
21] M. Guisnet, P. Ayrault, J. Datka, Pol. J. Chem. 71 (1997) 1455.
22] A. Zecchina, G. Spoto, S. Bordiga, Phys. Chem. Chem. Phys. 7 (2005) 1627.
23] O. Marie, P. Massiani, F. Thybault-Starezyk, J. Phys. Chem. B 108 (2004) 5073.
[31] S. Morin, N.S. Gnep, M. Guisnet, J. Catal. 159 (1996) 296.
[32] J.A. Martens, J. Perez-Pariente, E. Sastre, A. Corma, P.A. Jacobs, Appl. Catal. 45
(1988) 85.