Phenol methylation on acid catalysts: Study of the catalyst deactivation kinetics and mechanism

Article abstract of DOI:10.1016/j.apcata.2014.01.039

The kinetics and mechanism of coke formation and catalyst deactivation during the synthesis of cresols from phenol methylation were investigated on SiO₂-Al₂O₃, tungstophosphoric acid (HPA) supported on silica, and zeolites HBEA, HZSM5, HMCM22 and HY. The nature, density and strength of surface acid sites were probed by temperature programmed desorption of NH₃ coupled with infrared spectra of adsorbed pyridine. Coke formed on the catalysts during reaction was characterized by temperature programmed oxidation. All the samples deactivated on stream. A linear correlation was observed between the initial catalyst deactivation and the amount of coke, thereby indicating that coke formation was responsible for the activity decay. Coking kinetic studies showed that a significant part of coke was rapidly formed from the reactants. When methanol was feeding alone, significant amounts of carbon were formed on the catalysts (between 0.6% and 6.2%C), particularly on samples containing mainly strong Br?nsted acid sites. Nevertheless, the coke amounts formed during phenol methylation were clearly higher (between 3.7% and 14.9%C), which showed that phenol was also responsible for coke formation. More insight on the role of phenol and methanol on coke formation was obtained by characterizing the coke nature using infrared spectroscopy. Coked samples recovered after methanol decomposition reaction exhibited IR absorption bands characteristics of olefinic species formed on Br?nsted and Lewis acid sites. The IR spectra of coked samples recovered after phenol methylation showed the presence of phenolate, aromatic and polyaromatic species adsorbed mainly on Lewis acid sites.

Full text of DOI:10.1016/j.apcata.2014.01.039

Applied Catalysis A: General 475 (2014) 305–313
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Phenol methylation on acid catalysts: Study of the catalyst
deactivation kinetics and mechanism
∗
M.E. Sad, C.L. Padró, C.R. Apesteguía
Catalysis Science and Engineering Research Group (GICIC), Instituto de Investigaciones en Catálisis y Petroquímica—INCAPE-(UNL-CONICET), Santiago del
Estero 2654, Santa Fe 3000, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 October 2013
Received in revised form 12 January 2014
Accepted 17 January 2014
Available online 26 January 2014
The kinetics and mechanism of coke formation and catalyst deactivation during the synthesis of cresols
from phenol methylation were investigated on SiO2–Al2O3, tungstophosphoric acid (HPA) supported on
silica, and zeolites HBEA, HZSM5, HMCM22 and HY. The nature, density and strength of surface acid sites
were probed by temperature programmed desorption of NH3 coupled with infrared spectra of adsorbed
pyridine. Coke formed on the catalysts during reaction was characterized by temperature programmed
oxidation. All the samples deactivated on stream. A linear correlation was observed between the initial
catalyst deactivation and the amount of coke, thereby indicating that coke formation was responsible
for the activity decay. Coking kinetic studies showed that a significant part of coke was rapidly formed
from the reactants. When methanol was feeding alone, significant amounts of carbon were formed on
the catalysts (between 0.6% and 6.2%C), particularly on samples containing mainly strong Brønsted acid
sites. Nevertheless, the coke amounts formed during phenol methylation were clearly higher (between
Keywords:
Phenol methylation
Cresol synthesis
Acid catalysts
Coking mechanism
Coking kinetics
3
.7% and 14.9%C), which showed that phenol was also responsible for coke formation. More insight on
the role of phenol and methanol on coke formation was obtained by characterizing the coke nature
using infrared spectroscopy. Coked samples recovered after methanol decomposition reaction exhibited
IR absorption bands characteristics of olefinic species formed on Brønsted and Lewis acid sites. The IR
spectra of coked samples recovered after phenol methylation showed the presence of phenolate, aromatic
and polyaromatic species adsorbed mainly on Lewis acid sites.
©
2014 Elsevier B.V. All rights reserved.
1
. Introduction
cresols, xylenols and methylanisoles at relatively low temperatures
473–523 K), but the product distribution greatly depend on the
(
Cresol isomers are useful intermediates for the synthesis of
nature and strength of surface acid sites and on the catalyst pore
microstructure. However, the catalysts suffer from strong deacti-
vation on stream as occurs in similar alkylation reactions [10–12],
and thus the potential use of solid acids for obtaining cresol isomers
via phenol methylation is hampered because of the rapid activity
decay. In spite of that, no studies have been reported yet dealing
specifically with the deactivation of solid acids employed for the
phenol methylation reaction. In particular, our understanding of
the deactivation kinetics and mechanism under reaction conditions
is lacking.
In previous works, we studied the synthesis of cresol isomers via
the alkylation of phenol with methanol using acid catalysts [13–15].
Research efforts were devoted to elucidate the exact requirements
of density, nature and strength of surface acid sites for selectively
improving the reaction pathways leading from phenol to a desired
cresol isomer.
antioxidants, pharmaceuticals, herbicides, agrochemicals and dyes.
o-Cresol is commercially produced by gas-phase alkylation of phe-
nol with methanol on solid bases at temperatures between 673 K
and 773 K [1–3]. p-Cresol is obtained by toluene dehydration via
sulfonation with sulphuric acid in a four reaction step process
while m-cresol is produced by alkaline chlorotoluene hydrolysis
or cymene hydroperoxide cleavage [4]. The current syntheses of
m- and p-cresol entail concerns related to corrosion, and disposal
of spent and waste materials, and there is therefore a clear inter-
est for developing more ecofriendly synthesis routes, in particular
using solid catalysts. In this regard, the alkylation of phenol with
methanol has been widely studied at atmospheric pressure on acid
catalysts for producing cresols [5–9]. Reports showed that on acid
zeolites, silica–alumina or Nafion–H, phenol converts to anisole,
In particular, we investigated the selective synthesis of p-
cresol on zeolites HMCM22, HBEA and HZSM5 [14]. We found
that the particular pore structure of zeolite HMCM22 promotes
by shape selectivity the formation of p-cresol at the expense of
∗ Corresponding author. Tel.: +54 3424555279.
E-mail addresses: capesteg@fiq.unl.edu.ar,
capesteg@gmail.com (C.R. Apesteguía).
0
926-860X/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2014.01.039

3
06
M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
anisole and o-cresol among the primary products and drastically
suppresses the consecutive reactions forming secondary prod-
ucts. Thus, the p-cresol yield and the para-/ortho-cresol ratio on
HMCM22 for 93% phenol conversion were about 58% and 3.4,
respectively, the highest values reported up to now for p-cresol
formation from methylation of phenol. However, the activity decay
on HMCM22 during the progress of the reaction was considerable,
thereby suggesting that the reactions forming coke intermedi-
ates were not suppressed to any significant extent. In basis of
this observation, we decided to extend our studies for the phe-
nol methylation reaction focusing specifically on catalyst activity
decay. Thus, in this paper we study the kinetics and mechanism of
coke formation and catalyst deactivation during phenol methyla-
tion on samples containing essentially strong Brønsted acid sites
sample wafers were initially outgassed in vacuum at 723 K dur-
ing 2 h and then a background spectrum was recorded after
cooled down to room temperature. Coked sample wafers were out-
gassed 2 h at 473 K. Spectra were recorded at room temperature,
after admission of pyridine, adsorption at room temperature and
sequential evacuation at 298 K, 423 K, 573 K, and 723 K.
Coke formed on the catalysts during reaction was measured by
temperature programmed oxidation (TPO). Samples (20–100 mg)
were heated in a 2% O /N stream from room temperature to 1073 K
2
2
at 10 K/min. The evolved CO2 was converted into methane passing
through a methanation catalyst (Ni/kieselghur) at 673 K. Methane
was detected and quantified in a SRI 8610C gas cromatograph
equipped with a flame ionization detector.
(
HPA/SiO ) and catalysts containing both Lewis and Brønsted acid
2
2
.3. Catalyst testing
sites of either strong (zeolites HZSM5, HBEA, HMCM22 and HY)
or moderate (SiO –Al O ) strength. The aim was to identify the
species responsible for coke formation and to establish the relation-
ship between catalyst deactivation and the nature of surface acid
sites.
2
2
3
The gas phase alkylation of phenol (Merck, >99%) with methanol
(Merck, 99.8%) was carried out in a fixed bed tubular reactor at
4
73 K and 101.3 kPa in continuous flow of N . Samples (0.025–0.6 g,
2
particles with 0.35–0.42 mm diameter) were pretreated in-situ, at
23 K in air flow (75 ml/min) for 2 h before reaction. Methanol (M)
7
and phenol (P) were fed (M/P = 2:1 molar) using a syringe pump
2
. Experimental
and vaporized into flowing N2 to give a N /(P + M) molar ratio of
2
2
6.8. Catalytic runs were carried out at different contact times
2.1. Catalyst preparation
0
(W/F ), between 5.6 g h/mol and 560 g h/mol. Samples were col-
P
lected every 20 min during 4 h and the exit gases were analyzed
on-line using an Agilent 6850 gas chromatograph equipped with a
flame ionization detector and a 30 m Innowax column (inner diam-
eter: 0.32 mm, film thickness: 0.5 ␮m). After reaction, catalysts
The HY zeolite was prepared by triple ion exchange of a
commercial NaY zeolite (UOP-Y 54) with ammonium chloride
Merk, 99.8%) at 353 K and subsequent calcination in air. Zeolite
HMCM22 was synthesized according to [16], by using sodium alu-
minate (Alfa Aesar, Technical Grade), silica (Aerosil Degussa 380),
sodium hydroxide (Merck, >99%), hexamethyleneimine (Aldrich,
(
were kept in N flow at 473 K during 45 min to desorb reactants and
2
products weakly adsorbed. Phenol conversion (XP) and methanol
0
− YP,M)/Y0
conversion (XM) were calculated as: XP,M = (Y
a
9
9%) and deionized water as reagents. The molar composi-
P,M
P,M
where Y P0 and Y0 are the molar fraction of phenol and methanol
tion of the synthesis gel was SiO /Al O = 30, OH/SiO = 0.18,
2
2
3
2
M
hexamethyleneimine/SiO = 0.35 and H O/SiO = 45. The gel was
at the entrance of the reactor while YP and YM are the molar frac-
2
2
2
transferred to a Teflon lined stainless steel autoclave, rotated at
0 rpm, and heated to 423 K in an oven for 7–10 days. Commer-
tion of phenol and methanol at the exit. The selectivity to product
5
i formed from phenol molecules (S , mol of product i/mol of phe-
i
cial samples were zeolite HBEA (Zeocat PB), zeolite HZSM5 (Zeocat
Pentasil PZ-2/54) and SiO –Al O (Ketjen LA-LPV). HPA (28%)/SiO2
nol reacted) was determined as: S = [Y /ꢀY ] where Y is the molar
fraction of products formed from phenol.
i
i
i
i
2
2
3
was obtained by wet impregnation. A suspension of SiO2 powder
2
(
Grace G62, 99.7%; 230 m /g) was stirred in an aqueous solution of
3
. Results and discussion
HPA (H PW12O40.6H O, Merck P.A) at room temperature for 24 h.
Before reaction, all the samples were treated in air (60 cm /min) at
3
2
3
3.1. Catalyst characterization
7
23 K during 3 h, except HPA/SiO2 that was treated at 573 K for 2 h.
The chemical composition and surface areas of the samples used
in this work are shown in Table 1. Zeolite HY presented the highest
2
.2. Catalyst characterization
2
2
SBET value (660 m /g) followed by HBEA (560 m /g) and SiO –Al O
2
2
3
2
The crystal structure of the samples was determined by pow-
(540 m /g) samples. In the case of HPA/SiO2 the addition of bulky
der X-ray diffraction methods (XRD) using a Shimadzu XD-D1
diffractometer and Ni-filtered CuK␣ radiation. Surface areas were
measured by N2 physisorption at its normal boiling point using
the BET method in an Autosorb Quantochrome Instrument 1-C
sorptometer. The chemical compositions were measured by atomic
absorption spectroscopy.
HPA molecules caused only a slight diminution of the SiO surface
2
2
2
area, from 230 m /g to 205 m /g. The crystalline structure of zeolite
HMCM22 synthesized in our laboratory was checked by XRD tech-
nique and the positions and intensities of the diffraction peaks were
in good agreement with the HMCM22 diffractograms reported in
literature [17].
The density and strength of acid sites were determined by
temperature-programmed desorption (TPD) of NH3 preadsorbed
The sample acid properties (nature, density and strength of acid
sites) determined by TPD of NH3 and FTIR of adsorbed pyridine are
also presented in Table 1. The total amount of desorbed NH3 was
measured by integration of the NH3 TPD curves (not shown here)
and it was taken as an indication of the total acid site density. Data
in Table 1 show that zeolite HY presented the highest surface acid
density on a weight basis (1380 ␮mol/g) whereas on an areal basis
at 373 K. Samples (100 mg) were treated at 723 K for 2 h in He
3
(
60 cm /min) and then exposed to a 1% NH /He stream for 40 min
3
at 373 K. Weakly adsorbed NH3 was removed by flushing with He
at 373 K (2 h). The temperature was then increased at 10 K/min
and the NH3 concentration in the effluent was measured by mass
spectrometry in a Baltzers Omnistar unit.
The nature of surface acid sites of fresh and coked catalysts
was determined by infrared spectroscopy (IR) using pyridine as
probe molecule and a Shimadzu FTIR Prestige-21 spectrophotome-
ter according to the procedure detailed elsewhere [14]. Fresh
2
2
zeolites HZSM5 (2.2 mmol/m ) and HY (2.1 mmol/m ) exhibited
2
the highest acid density followed by SiO –Al O (1.8 mmol/m ).
2
2
3
The nature and strength of surface acid sites were determined
from the FTIR spectra obtained after admission of pyridine at
room temperature and sequential evacuation at 423 K, 573 K, and

M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
307
Table 1
Physicochemical and acid properties of the samples.
2
TPD of NH3
FTIR of pyridinea
Sample
Si/Al
HPA (wt%)
Na (wt%)
SBET (m /g)
Pore size (Å)
Brønsted sites (B)
Lewis sites (L)
L/B
(
␮mol/g)
(␮mol/m2)
(area/g)
(area/g)
HBEA
HZSM5
HY
HMCM22
SiO2-Al2O3
HPA/SiO2
25.0
20.0
2.4
15.0
11.3
–
–
–
–
–
–
28
0.03
0.43
0.33
–
–
–
560
350
660
400
540
205
6.6 × 6.7; 5.6 × 5.6
5.1 × 5.5; 5.3 × 5.6
7.4 × 7.4
500
770
1380
470
1030
160
0.9
2.2
2.1
1.2
1.8
0.8
387
337
310
560
68
309
341
465
176
204
15
0.8
1.0
1.5
0.3
3.0
0.2
4.0 × 5.5; 4.1 × 5.1
45
225
91
a
FTIR spectra obtained after adsorption of pyridine at 298 K and evacuation at 423 K (B: Brønsted sites; L: Lewis sites).
sample showed the highest proportion of Lewis sites (L/B = 3). The
L/B ratios on zeolites HY, HZSM5 and HBEA were 1.5, 1.0 and 0.8,
respectively. Zeolite HMCM22 exhibited mainly Brønsted acidity
25
HPA/SiO2
HMCM22
(L/B = 0.3) which is consistent with previous work reporting L/B
ratios between 0.2 and 0.5 for zeolites HMCM22 with Si/Al atomic
ratios of about 15 [22,23]. The L/B ratio on HPA/SiO2 was 0.2. The
HBEA
presence of pyridine adsorbed on Lewis acid sites of HPA/SiO sug-
2
gests that upon impregnation of HPA on SiO , the heteropolyacid
2
partially transforms giving rise to lacunary or unsaturated species
of Lewis acid character formed by interaction with the silica support
HY
[
24].
Fig. 2 shows the evolution of the amount of pyridine adsorbed
SiO -Al O
3
2
2
on surface Lewis and Brønsted sites as a function of the evacuation
temperature and gives insight on the acid site strength. Significant
amounts of pyridine remained adsorbed on HMCM22, HZSM5 and
HY after evacuation at 723 K thereby showing that these zeolites
contain strong acid sites. The L/B ratio on HY increased with the
evacuation temperature from 1.5 (423 K) to 2.1 (723 K) and proba-
bly reflects the fact that a part of the Brønsted sites corresponding
to the bridged hydroxyl groups of the structure are of medium
strength [25]. The very strong Lewis sites of zeolite HY have been
associated with aluminium atoms that are either extra-framework
(tetracoordinated) or associated with framework defects (tricoor-
dinated) [25]. In contrast to zeolite HY, the L/B ratio on zeolite
HZSM5 did not vary significantly by increasing the desorption tem-
perature. In fact, after degassing at 423 K and 723 K the L/B ratio
were 1.0 and 0.9, respectively. In the case of zeolite HMCM22, the
L/B ratio increased from 0.3 to 0.5 when the evacuation tempera-
ture was increased from 423 K to 723 K. The L/B ratio determined
on HBEA evacuated at 423 K (L/B = 0.8) diminished to 0.3 upon
HZSM5
1
700 1650 1600 1550 1500 1450 1400
-
1
Wavenumber (cm )
Fig. 1. FTIR spectra obtained after pyridine adsorption at 298 K and sample evacu-
ation at 423 K.
7
23 K. FTIR spectra after pyridine adsorption at 298 K and evac-
uation at 423 K are displayed in Fig. 1. The absorption bands at
−1
−1
−1
around 1540 cm and between 1440 cm and 1460 cm arise
from pyridine adsorbed on Brønsted and Lewis acid sites, respec-
tively [18–21]. The relative contribution of Lewis and Brønsted acid
sites after evacuation of pyridine at 423 K was obtained then by
integration of the bands at 1445–1460 cm (Lewis) and 1540 cm
Brønsted); results are presented in Table 1. The SiO –Al O
3
−
1
−1
(
2
2
500
400
300
200
100
0
600
500
400
300
200
100
0
A
B
400
500
600
700
800
400
500
600
700
800
Temperature (K)
Temperature (K)
Fig. 2. Lewis (A) and Brønsted (B) peak areas as a function of the evacuation temperature. HY (ꢀ), HBEA (ꢁ), HZSM5 (ꢀ), SiO2–Al2O3 (᭹), HMCM22 (ꢂ), HPA/SiO2 (ꢃ).

3
08
M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
Anisole may in turn produce o- and p-cresol by alkylating phenol.
m-Cresol is formed by consecutive isomerization of o- and p-cresol
isomers. Alkylation of o- and p-cresol isomers with methanol forms
Scresols
6
4
2
0
0
0
0
2
,4 and 2,6 xylenols, while anisole may reacts with methanol to
produce methylanisoles.
Sanisole
Fig. 3 shows that phenol conversion diminished with time-on-
stream, reflecting the in-situ sample deactivation. Similar catalyst
activity decay was observed on all the samples. Due to catalyst
deactivation, initialcatalyticresultswereobtainedby extrapolation
of reactant and product concentration curves to zero time-on-
stream. The phenol conversions obtained at zero time and after 3 h
on stream for all the samples are shown in Table 2. Zeolites HBEA
and HY showed the highest initial phenol conversions which is con-
sistent with previous work reporting that the formation of cresols
by phenol methylation is better promoted on wide pore zeolites
containing strong surface Lewis and Brønsted acid sites [13]. Zeo-
Sxylenols
XP
SMA
0
50
100
150
200
250
Time (min)
∼
lites HZSM5 and HMCM22 presented similar activity (XP = 35%) but
Fig. 3. Phenol conversion (XP ) and selectivities as a function of time-on-stream on
HPA/SiO2 [473 K, PT = 101.3 kPa, W/FP = 112 g h/mol, PP = 0.96 kPa, PM = 1.92 kPa].
were clearly less active than HBEA or HY, in spite of their strong
acidity. As proposed elsewhere [14], the phenol methylation rate
is probably controlled by intracrystalline diffusion on HZSM5 and
HMCM22 taking into account that these two zeolites present the
smallest pore sizes (Table 1). HPA/SiO2 showed the lowest ini-
tial activity (XP = 20%) suggesting that phenol methylation is not
efficiently promoted on catalysts containing essentially Brønsted
acid sites. Methanol conversion was higher than phenol conversion
for all the catalysts because methanol was involved in additional
parallel reactions forming dimethyleter and hydrocarbons. Nev-
ertheless, methanol was always present among reaction products
indicating that the reaction was not limited by total methanol con-
sumption.
0
evacuation at 723 K. The L/B ratio on SiO –Al O evacuated at 423 K
2
2
3
was 3 (Table 1) and increased to about 4 when the evacuation tem-
perature was increased to 723 K. Finally, after evacuation at 423 K
and 473 K, HPA/SiO2 contained almost exclusively Brønsted sites.
Overall, the results in Table 1 and Figs. 1 and 2 show that the sam-
ples used in this work have different acid properties; i.e. different
acid site density, nature and strength.
3
.2. Catalytic tests
3
.2.1. Phenol methylation: activity and product distribution
Main products in the alkylation of phenol with methanol were
The product selectivities obtained at XP = 20% are also reported
in Table 2. On all the catalysts, primary monoalkylated products
(anisole and cresols) accounted for more than 85% of total selec-
tivity; minor amounts of secondary dialkylated products (xylenols
and methylanisoles) were also obtained. The cresol selectivity on
HBEA, HZSM5, HY and SiO2–Al2O3 was about 54%, being o-cresol
isomer formed in major proportion. In contrast, anisole was the
main product on HPA/SiO2 (Sanisole = 65%) suggesting that catalysts
exhibiting essentially Brønsted acidity will preferentially promote
the O-alkylation of phenol. The cresol selectivity on HMCM22 was
about 90%; the narrow channels of this zeolite completely sup-
pressed the formation of secondary dialkylated compounds. In
general, xylenol formation was favored on wide pore zeolites con-
taining strong surface Lewis and Brønsted acid sites (zeolites HBEA
and HY) while the selectivity to methylanisoles was lower than 4%
on all the samples.
cresols, anisole, xylenols and methylanisoles (MA). Dimethylether
DME) and some light hydrocarbons were also formed by dehy-
(
dration and further condensation of methanol on acid sites. Fig. 3
shows the evolution of phenol conversion (XP) and selectivities (Si)
as a function of time obtained at 473 K on HPA/SiO2 and typically
illustrates the time-on-stream behavior of the catalysts for phenol
methylation. Results obtained in previous works [13,14] allowed
us to propose the reaction network of Fig. 4 for the methylation of
phenol on solid acids. Phenol initially reacts with methanol via two
parallel alkylation reactions: by O-alkylation phenol is transformed
to anisole and by C-alkylation yields directly o- and p-cresols.
3.2.2. Catalyst deactivation and coke formation
All the catalysts deactivated during the alkylation of phenol with
methanol. In Fig. 5 we observe the evolution of the activity (aP) as a
function of time for all the catalysts obtained at similar initial phe-
0
0
nol conversions (25–35%); aP is defined as aP = rP/r , where r and
P
P
rP are the rates of phenol conversion at t = 0 and t = t, respectively.
From the initial slopes of curves aP vs. t in Fig. 5 we determined
the parameter d0 = −[daP/dt]
that accounts for the initial cat-
t=0
alyst deactivation [26]; the obtained d values are included in
0
−
3
−1
Table 3. The highest d0 value (27.0 × 10 min ) was determined
on zeolite HY while the lowest initial deactivation was observed for
−
3
−1
SiO –Al O (d = 2.1 × 10 min ). HPA/SiO and HMCM22 pre-
2
2
3
0
2
−
3
−1
sented similar d values of about 18 × 10 min that were higher
0
−
3
−1
than those corresponding to HZSM5 (14.6 × 10 min ) and HBEA
−
3
−1
(
10.5 × 10 min ).
The catalyst deactivation observed in Fig. 5 could be caused
by coke formation from parallel/consecutive reactions of reac-
tants/products on surface active sites. We characterized then the
carbonaceous deposits of samples recovered from the catalytic
Fig. 4. Reaction network for methylation of phenol on solid acids.

M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
309
Table 2
Catalytic results for the alkylation of phenol with methanol.
Phenol conversióna
XP (%)
Methanol
conversión XM (%)
Selectivity (XP = 20%, t = 0)b Si (%)
Catalyst
a
t = 0 h
t = 3 h
t = 0 h
t = 3 h
Anisole
Cresols
Xylenols
MA
HBEA
HZSM5
HY
HMCM22
HPA/SiO2
SiO2–Al2O3
85
35
73
34
20
30
32.0
7.0
0.3
6.0
2.0
88
75
80
83
82
55
70
50
72
70
65
50
33
41
31
8.5
65
34
53
54
54
90
32
54
11.5
3
13
0.5
0.5
9
2.5
2
2
1
2.5
3
17.5
a
473 K, PT = 101.3 kPa, PP = 0.96 kPa, PM = 1.92 kPa, W/F0 = 112 g h/mol].
473 K, PT = 101.3 kPa, PP = 0.96 kPa, PM = 1.92 kPa].
[
[
P
b
HY and HZSM5 exhibited two superimposed bands at 690–750 K
and 880–920 K, respectively.
The amounts of carbon were calculated from the areas under
the TPO curves of Fig. 6 and are reported in Table 3 as % Carbon.
Significant coke amounts were formed on the catalysts, between
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
3
.7% (HPA/SiO ) and 14.9% (HMCM22). In Table 3 we also included
2
as
the values of nC , number of moles of C deposited per mol of acid
sites (acid site density data obtained from TPD of NH ). The highest
3
as
C
as
n
(
values were obtained on HMCM22 (nC = 26.5) and HPA/SiO2
=19.3), probably due to the predominant presence of strong
as
n
C
Brønsted acid sites on both samples. HBEA and HZSM5 that contain
L/B acid site ratios of about 1 (Table 1) formed similar amounts
as
as
of C per acid site (nC between 14 and 16). The lowest nC value
as
was determined on SiO –Al O (n =4.8), probably reflecting its
2
2
3
C
moderate acid strength.
In order to get more insight on the relationship between cat-
alyst deactivation and coke formation, we plotted in Fig. 7 initial
as
deactivation d0 as a function of nC for samples recovered from
the reactor after performing catalytic tests for 40 min and 4 h. A
as
linear correlation between d0 and nC was obtained for all the cat-
alysts (excepting HY) in both cases, thereby suggesting that coke
formation is responsible for the activity decay observed in Fig. 5
by blocking the surface active sites required for the reaction to
proceed. The different deactivation behavior observed for zeolite
HY is consistent with previous work showing that the activity decay
is usually higher on HY than on HZSM5 or HBEA [27]. This is prob-
ably due to fact that zeolite HY contains large cages only accessible
0
50
100
150
200
250
Time [min]
Fig. 5. Activity decay for phenol conversion (aP ) as a function of time on stream.
Initial phenol conversions between 25% and 35% [473 K, PT = 101.3 kPa, PP = 0.96 kPa,
PM = 1.92 kPa]. SiO2–Al2O3 (᭹), HBEA (ꢁ), HZSM5 (ꢀ), HPA/SiO2 (ꢃ), HMCM22 (ꢂ),
HY (ꢀ)
runs showed in Fig. 5 by performing TPO experiments. Prior the
TPO characterization, samples were treated at 473 K in N2 dur-
ing 45 min. The TPO profiles are presented in Fig. 6. The HPA/SiO2
sample, containing almost exclusively strong Brønsted acid sites,
presented a single oxidation peak with a maximum at around 790 K.
The TPO curve for HMCM22 exhibited a low-temperature peak at
2
00
SiO -Al O
3
2
2
5
90 K and a broad combustion band between 600 K and 1030 K. The
low-temperature peak probably arises from carbon deposits with
high H/C ratio retained in the small channels of this zeolite. HBEA
and SiO –Al O showed a broad combustion band that begins at
HY
2
2
3
HBEA
6
00–620 K and ends at about 1050 K. Finally, the TPO profiles for
HZSM5
Table 3
Phenol methylation: initial deactivation d0 and coke formed after 4 h reaction time.
HMCM22
3
−1
as
C
Catalyst
d0 × 10 (min
)
(%) Carbon
n
(mol C/mol acid site)
HBEA
HZSM5
HY
HMCM22
HPA/SiO2
SiO2–Al2O3
10.5
14.6
27.0
18.8
17.5
2.1
9.5
6.2
13.2
14.9
3.7
15.8
14.2
7.9
26.5
19.3
4.8
HPA/SiO
2
400
600
800
1000
Temperature [K]
6.0
[473 K, PT = 101.3 kPa, PP = 0.96 kPa, PM = 1.92 kPa, XP = 25–35%].
Fig. 6. TPO profiles of samples recovered from the catalytic runs showed in Fig. 5.

3
10
M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
2
1
4
6
8
0
(
a)
HY
400
HMCM22
HPA
HZSM5
(b)
(
c)
3
2
00
00
HBEA
SiO -Al O
3
2
2
0
10
20
30
as
nC (mol C/mol acid sites)
100
0
Fig. 7. Initial catalyst deactivation (d0) versus nas (moles of C per mol of acid site).
C
Samples recovered after 40 min (ꢀ) and 4 h (ꢂ) on stream for phenol methylation
[
473 K, PT = 101.3 kPa, PP = 0.96 kPa, PM = 1.92 kPa, W/F P0 = 112 g h/mol].
4
00
600
800
100 0
through small pores. Coke easily formed in large cages produces
then severe stereochemical blockage and pronounced deactivation
Temperature (K)
[
28].
Fig. 9. TPO profiles of HBEA samples recovered after using in 4-h phenol methylation
runs at different contact times: (a) 56 g h/mol phenol, (b) 112 g h/mol phenol and (c)
168 g h/mol phenol [473 K, PT = 101.3 kPa, PP = 0.96 kPa, PM = 1.92 kPa].
3.2.3. Catalyst deactivation kinetics and mechanism
Several catalytic runs of different length were performed in
and SiO –Al O samples (results not shown here), thereby sug-
gesting that a significant part of coke is rapidly formed from
reactants.
To obtain more insight on the species responsible for catalyst
deactivation, we performed additional catalytic tests on zeolite
HBEA at different contact times. In Fig. 9 we compared the TPO
profiles of HBEA samples recovered after using in phenol methyl-
ation runs of 4-h length at contact times between 56 g h/mol and
2
2
3
order to attain deeper knowledge on the kinetics of coke forma-
tion. Specifically, phenol methylation was carried out on zeolite
HBEA during 20 min, 40 min and 240 min, and the catalysts recov-
ered from these runs were characterized by TPO experiments as
shown in Fig. 8. The TPO profile shapes were qualitatively similar,
thereby suggesting that the coke nature did not change signifi-
cantly with time on stream. The amounts of coke determined by
integrating the TPO curves of Fig. 8 were 5.7%, 7.0% and 9.5% after
1
68 g h/mol. The TPO profile shapes in Fig. 9 were qualitatively sim-
2
6
0 min, 40 min, and 240 min of reaction, respectively. Thus, about
0% of the total coke determined after the 240-min catalytic run
ilar but the area under the curves (i.e. the amount of coke) increased
was formed during the initial 20 min on stream. Qualitatively, sim-
ilar coke formation kinetics was obtained on HZSM5, HY, HMCM22
16
600
500
400
300
200
100
0
1200
2
40 min
12
8
1
8
6
4
000
40 min
00
20 min
00
4
00
0
200
0
50
100
150
200
0
20
40
60
80
Time (min)
W/F 0 (g h/mol)
P
Fig. 8. TPO profiles of HBEA samples recovered after using in phenol methyla-
tion runs of 20 min, 40 min and 240 min length [473 K, PT = 101.3 kPa, PP = 0.96 kPa,
PM = 1.92 kPa, W/F P0 = 112 g h/mol].
Fig. 10. Phenol methylation: Coke formed after 4 h reaction as a function of contact
time [473 K, PT = 101.3 kPa, PP = 0.96 kPa, PM = 1.92 kPa].HMCM22 (ꢂ), HY (ꢀ), HBEA
(ꢁ), HZSM5 (ꢀ), SiO -Al O (᭹)
2
2
3

M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
311
A
B
6
0
2
1
4
3
4
5
8
2
1
3
(c)
(c)
(b)
(b)
(a)
(a)
3200
3100
3000
2900
2800
2700 1700
1600
1500
1400
-1
-1
Wavenumber (cm )
Wavenumber (cm )
Fig. 11. 2700–3200 cm 1 (A) and 1350–1700 cm (B) FTIR spectra regions obtained for: (a) fresh HBEA sample; (b) HBEA sample recovered after 4-h methanol decomposition
−
−1
reaction; (c) HBEA sample recovered after 4-h phenol methylation reaction.
when contact time (and phenol conversion) decreased. Actually,
the effect of contact time on coke formation was studied on all the
catalysts used in this work. In Fig. 10 we have plotted for all the
catalysts the % carbon formed after the 4-h catalytic runs as a func-
tion of contact time. It is observed that in all the cases the amount
of coke increased when contact time was decreased thereby con-
firming that catalyst deactivation is related with the presence of
the reactants, methanol and/or phenol.
Although methanol conversion remained constant with time on
stream, the catalysts changed their color at the end of the runs
suggesting the presence of carbonaceous deposits on the catalyst
surface. The amounts of carbon formed on all the samples were
determined by TPO technique and the results are given in Table 4.
In general, the amount of coke increased with catalyst activity for
the methanol decomposition reaction. For example, the highest
amountof coke (6.2%) was determinedon HPA/SiO , the most active
2
Therefore, we decided to study the formation of coke on our cat-
alysts when only methanol is feeding to the reactor at the standard
operation conditions used for phenol methylation. We analyzed
the methanol decomposition products (DME and hydrocarbons)
and determined the carbon amounts formed on the catalysts by
TPO technique. Results are shown in Table 4. On all the catalysts,
methanol conversion (XM) was higher than 70% and no noticeable
catalyst deactivation was observed during the 4-h reaction time.
catalyst, while SiO –Al O formed only 0.6% carbon. What is impor-
2 2 3
tant to remark here is that in spite of that methanol decomposition
formed significant amounts of coke on the catalysts, the %C val-
ues presented in Table 4 were clearly lower than those formed on
the corresponding catalysts during phenol methylation (Table 3).
These results strongly suggested that phenol is also responsible for
coke formation in the alkylation of phenol with methanol, either
because of its strong adsorption on surface active sites or because
of secondary phenol/methanol reactions producing coke interme-
diates. To obtain more insight on the role of phenol and methanol
in the formation of coke, we performed additional experiences by
characterizing the nature of coke using FTIR spectroscopy.
The highest XM value was obtained on HPA/SiO (96%) that contains
2
essentially strong Brønsted acid sites (Table 1). Zeolites HY, HBEA,
HZSM5 and HMCM22 (containing both Brønsted and Lewis acid
sites of relatively high strength) were also very active for convert-
ing methanol (XM = 85–90%). Finally, methanol conversion was only
7
0% on SiO –Al O that exhibits moderate acid strength. Formation
2 2 3
3.2.4. Study of nature of coke by FTIR spectroscopy
of coke from methanol decomposition on solid acids has been stud-
ied by other authors [29–33]. It has been proposed that methanol is
initially dehydrated to dimethylether on either Brønsted or Lewis
acid sites and then light hydrocarbons such as methane, ethene,
ethane, propane, propene, and aromatics may be formed preferen-
tially on Brønsted acid sites [31]. Consistent with this proposal, we
observed here that all our catalysts produced DME which suggests
that methanol dehydration is promoted on Lewis and Brønsted
acid sites indistinctively. Besides, we also observed that hydro-
carbon formation was particularly favored on samples containing
We used FTIR spectroscopy to characterize the coke formed
on HBEA during phenol methylation and methanol decomposition
reactions. Samples were recovered after carrying out both reac-
tions under the same operation conditions for 4 h. Fig. 11 shows
the FTIR spectra obtained on fresh and coked HBEA samples in the
two most important frequency regions to characterize coke, i.e.
−
1
−1
1
350–1700 cm and 2700–3200 cm regions [34]. Coked HBEA
sample recovered after methanol decomposition reaction showed
−
1
in the 2700–3200 cm region (Fig. 11A, spectrum b) several bands
−
1
−1
−1
and 2957 cm (–CH3
at 2854 cm
(–CH groups), 2872 cm
groups), 2927 cm (–CH and –CH groups) and 2982 cm (olefins
2
predominantly strong Brønsted acid sites such as HPA/SiO , HY or
2
−1
−1
2
HMCM22.
groups) [35]. This sample also exhibited a very broad absorption
−
1
−1
band between 1695 cm
and 1550 cm (Fig. 11B, spectrum
b) attributed mainly to C–C double bond stretch of olefins [36].
Coked HBEA sample recovered after phenol methylation displayed
Table 4
Methanol decomposition reaction: conversion and coke formation.
−
1
several IR bands in the 1350–1700 cm region (Fig. 11B, spectrum
Catalyst
XM (%)a
(%) Carbonb
−
1
−1
are consistent with
c). The bands at 1505 cm
and 1487 cm
HBEA
HZSM5
HY
HMCM22
HPA/SiO2
SiO2–Al2O3
89
85
85
86
96
70
3.8
3.4
5.5
5.3
1.9
0.6
the presence of adsorbed phenol species [36] whereas the bands
−
1
−1
−1
at 1602 cm , 1585 cm and 1539 cm are characteristic of C–C
stretching vibrations of polyaromatics compounds [35]. The IR
−
1
band appearing at 1382 cm
supports the presence of methyl
−1
groups linked to aromatic compounds [36]. In the 2700–3200 cm
region (Fig. 11A, spectrum c), this coked HBEA sample presented
all the IR absorption bands associated with carbonaceous species
formed from methanol decomposition, and additional bands at
[
473 K, PT = 101.3 kPa, PM = 1.92 kPa, W/F0 = 57.9 g h/mol].
M
a
Methanol conversion was constant during the 4-h catalytic tests.
b
Determined by TPO technique after 4 h reaction.

3
12
M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
phenol conversion rate on both samples is lower than that on HBEA
and HY because the reaction is probably controlled by intracrys-
talline diffusion. Phenol methylation is not favored on samples
that present moderate Lewis and Brønsted acidity (SiO –Al O ) or
(c)
2
2
3
contain only strong Brønsted acid sites (HPA/SiO ).
2
(
b)
All the samples deactivated on stream because of coke forma-
tion. The initial catalyst deactivation decreased linearly with the
amount of coke, excepting for zeolite HY. The major part of coke
is formed rapidly from methanol/phenol reactants and therefore
the amount of coke formed on the catalysts decreases with contact
time (and phenol conversion). Methanol reactions form significant
carbonaceous deposits, particularly on samples containing pre-
dominantly strong Brønsted acid sites such as zeolite HMCM22
(a)
and HPA/SiO . Nevertheless, the coke amount formed in the phenol
2
1
700
1650
1600
1550
1500
1450
1400
methylation reaction is significantly higher than that formed on the
corresponding catalyst during methanol decomposition showing
that phenol is the main responsible for coke formation. Methanol
decomposition builds carbonaceous olefinic species on the cat-
alyst surface while cofeeding methanol/phenol forms additional
adsorbed phenolate, aromatic and polyaromatic species, mainly on
Lewis acid sites. The preferential adsorption of phenol on Lewis acid
sites required to promote phenol methylation explains the fact that
phenol conversion decay is clearly higher than that of methanol
conversion during the progress of this reaction. In fact, methanol
decomposition can simultaneously proceed on Brønsted acid sites
via parallel dehydration, dehydrogenation and condensation reac-
tions.
-
1
Wavenumber (cm )
Fig. 12. FTIR spectra of pyridine adsorbed at 298 K and evacuated at 423 K on: (a)
fresh HBEA sample; (b) HBEA sample recovered after 4-h methanol decomposition
reaction; (c) HBEA sample recovered after 4-h phenol methylation reaction.
3
050 cm 1, 3076 cm−1, 3106 cm attributable to adsorbed phenol,
− −1
olefinic CH2 groups, and benzene groups, respectively [34,35]. In
summary, IR spectra in Fig. 11 show that methanol decomposition
forms carbonaceous olefinic species on the catalyst surface while
cofeeding methanol/phenol causes the formation of additional
adsorbed phenolate, aromatic and polyaromatic species.
In an attempt to ascertain the relationship between coke forma-
tion and the nature of surface acid sites, we characterized the coked
HBEA samples of Fig. 11 by FTIR of adsorbed pyridine. Coked HBEA
sample wafers were outgassed 2 h at 473 K and then a background
spectrum was recorded after cooled down to room temperature.
Then pyridine was admitted at room temperature and desorbed
at 423 K before recording the IR spectrum. The obtained pyridine
IR spectra presented in Fig. 12 are difference spectra where the
coked HBEA spectra served as the reference. Spectrum (a) shows the
Acknowledgments
We thank the Universidad Nacional del Litoral (UNL), the Agen-
cia Nacional de Promoción Científica y Tecnológica (ANPCyT) and
the Consejo Nacional de Investigaciones Científicas y Técnicas
(CONICET), Argentina, for the financial support of this work.
−
1
bands characteristics of pyridine absorbed on Lewis (1450 cm
)
References
−
1
and Brønsted sites (1540 cm ) on fresh HBEA zeolite; the L/B ratio
is 0.8 as reported in Table 1. In spectrum (b) corresponding to HBEA
sample recovered after 4-h methanol decomposition reaction it is
[
[
[
1] G.L. Gregory, US Patent 4933509, General Electric Co., 1989.
2] S. Sato, K. Koizumi, F. Nozaki, Appl. Catal., A 133 (1995) L7–L10.
3] F. Cavani, L. Maselli, S. Passeri, J.A. Lercher, J. Catal. 269 (2010) 340–350.
−
1
−1
observed that the bands at 1450 cm and 1540 cm decreased
in about the same proportion as compared to those of fresh HBEA
sample. In contrast, spectrum (c) shows that the Lewis sites band
decreased in higher proportion than that of Brønsted sites on HBEA
sample recovered after 4-h phenol methylation reaction. As a con-
sequence, the L/B ratio on this sample diminished to 0.3. This late
result is consistent with previous reports showing that phenol is
adsorbed preferentially on Lewis acid centers, predominantly in
vertical orientation [37,38]. The preferential adsorption of phenol
on Lewis acid sites allows us to explain data in Table 2 showing that
the phenol conversion decay during the progress of phenol methyl-
ation reaction was clearly higher than that of methanol conversion.
In fact, the IR spectra of Fig. 11 indicated that part of coke formed on
the catalysts are adsorbed phenolate species that would deactivate
Lewis acid sites required to promote phenol methylation. In con-
trast, methanol decomposition may proceed on Brønsted acid sites
via dehydration, dehydrogenation and condensation reactions that
will be not deactivated by coking on Lewis acid sites.
[4] Willey-VCH (Ed.), Ullman’s Encyclopedia of Industrial Chemistry, seventh ed.,
Willey-VCH, Wenheim, 2011.
[
5] S. Balsama, P. Beltrame, P.L. Beltrame, P. Carniti, L. Forni, G. Zuretti, Appl. Catal.
3 (1984) 161–170.
[6] M. Renaud, P. Chantal, S. Kaliaguine, Can. J. Chem. Eng. 64 (1986) 787–791.
1
[
7] M. Marczewski, G. Perot, M. Guisnet, in: M. Guisnet, J. Barrault, C. Bouchoule, D.
Duprez, C. Montassier, G. Pérot (Eds.), Studies in Surface Science and Catalysis,
41, Elsevier, Amsterdam, 1988, p. 273.
[8] M. Marczewski, J.P. Bodibo, G. Perot, M. Guisnet, J. Mol. Catal. 50 (1989)
211–218.
[
9] E. Santacesaria, D. Grasso, D. Gelosa, S. Carrá, Appl. Catal. 64 (1990) 83–99.
[
10] G. Mirth, J. Cejkab, J.A. Jan Krtilb, Lercher, in: B. Delmon, G.F. Froment (Eds.),
Studies in Surface Science and Catalysis, 88, Elsevier, Amsterdam, 1994, p. 241.
[11] M. Han, S. Lin, E. Roduner, Appl. Catal., A 243 (2003) 175–184.
[
12] M.N. Vanoy-Villamil, C.R. Apesteguía, C.L. Padró, Microporous Mesoporous
Mater. 181 (2013) 54–60.
[
13] M.E. Sad, C.L. Padró, C.R. Apesteguía, Catal. Today 133–135 (2008) 720–728.
[14] M.E. Sad, C.L. Padró, C.R. Apesteguía, Appl. Catal., A 342 (2008) 40–48.
[
[
[
15] M.E. Sad, C.L. Padró, C.R. Apesteguía, J. Mol. Catal. A: Chem. 327 (2010) 63–72.
16] M.K. Rubin, P. Chu, US Patent 4,954,325, 1990.
17] S.L. Lawton, A.S. Fung, G.J. Kennedy, L.B. Alemany, C.D. Chang, G.H. Hatzikos,
D.N. Lissy, M.K. Rubin, H.K.C. Timken, S. Steuernagel, D.E. Woesner, J. Phys.
Chem. 100 (1996) 3788–3798.
[
[
[
[
[
18] E.P. Parry, J. Catal. 2 (1963) 371–379.
19] J.W. Ward, J. Catal. 10 (1968) 34–46.
20] H. Knözinger, Adv. Catal. 25 (1976) 184–271.
21] G. Busca, Catal. Today 41 (1998) 191–200.
22] D. Meloni, S. Laforge, D. Martin, M. Guisnet, E. Rombi, V. Solinas, Appl. Catal., A
215 (2001) 55–66.
4
. Conclusions
The catalyst activity, selectivity and stability for phenol meth-
[
[
23] A. Corma, C. Corell, V. Fornés, W. Kolodziejski, J. Pérez-Pariente, Zeolites 15
ylation greatly depend on the surface acid properties. Wide pore
zeolites containing strong surface Lewis and Brønsted acid sites
such as zeolites HBEA and HY are highly active for this reaction.
Zeolites HZSM5 and HMCM22 present also strong acidity, but the
(
1995) 576–582.
24] V.K. Diez, C.R. Apesteguia, J.I. Di Cosimo, Catal. Lett. 123 (2008) 213–219.
[25] A. Boréave, A. Aroux, C. Guimon, Microporous Mater. 11 (1997) 275–291.
[26] C.L. Padró, C.R. Apesteguía, Catal. Today 107–108 (2005) 258–265.

M.E. Sad et al. / Applied Catalysis A: General 475 (2014) 305–313
313
[
[
27] G. Elordi, M. Olazar, G. Lopez, P. Casta n˜ o, J. Bilbao, Appl. Catal., B 102 (2011)
24–231.
28] H.S. Cerqueira, G. Caeiro, L. Costa, F. Ramôa Ribeiro, J. Mol. Catal. A: Chem. 292
2008) 1–13.
[34] I. Graca, J.D. Comparot, S. Laforge, P. Magnoux, J.M. Lopes, M.F. Ribeiro, F. Ramoa
Ribeiro, Appl. Catal., A 353 (2009) 123–129.
[35] P. Casta n˜ o, G. Elordi, M. Olazar, A.T. Aguayo, B. Pawelec, J. Bilbao, Appl. Catal.,
2
(
B 104 (2011) 91–100.
[
[
[
[
29] H. Knözinger, K. Kochloeft, W. Meye, J. Catal. 28 (1973) 69–75.
30] H. Knözinger, D. Dautzenberg, J. Catal. 3 (1974) 142–144.
31] T. Tsoncheva, R. Dimitrova, Appl. Catal., A 225 (2002) 101–107.
32] S.H. Campbell, X.-Z. Jiang, R.F. Howe, Microporous Mesoporous Mater. 29
[36] R.T. Conley, Infrared Spectroscopy, second ed., Allyn and Bacon, Boston, MA,
1972.
[37] K.V.R. Chary, K. Ramesh, G. Vidyasagar, V.V. Rao, J. Mol. Catal. A: Chem. 198
(2003) 195–204.
(
1999) 91–108.
33] V. Vishwanathan, K.-W. Jun, J.-W. Kim, H.-S. Roh, Appl. Catal., A 276 (2004)
51–255.
[38] K. Tanabe, T. Nishizaki, in: G.C. Bond, P.B. Wells, F.C. Tompkins (Eds.),
Proceedings of the Sixth International Conference on Catalysis vol. 2, The Chem-
ical Society, London, 1977, p. 863.
[
2