Friedel-Crafts alkylation of sodium salicylate with 4-tert butylbenzyl chloride performed in aqueous dispersions of mesoporous oxides

Article abstract of DOI:10.1016/j.molcata.2012.09.020

Reactant incompatibility is a common problem in organic chemistry. In this study we investigate the use of concentrated aqueous dispersions of mesoporous oxides to overcome incompatibility in the Friedel-Crafts reaction between sodium salicylate and 4-tert-butylbenzyl chloride. The mesoporous material was first impregnated with the water-soluble nucleophile, sodium salicylate, and the loaded particles were then dispersed in the apolar electrophile, 4-tert-butylbenzyl chloride. A range of different mesoporous oxides and one clay mineral, montmorillonite, were investigated as catalyst for the reaction. These were all characterised with small angle X-ray scattering (SAXS) and with nitrogen adsorption (BET and BJH methods). Their Lewis and Bronstedt acidities were determined by ammonia adsorption experiments using diffuse reflection infrared Fourier transform (DRIFT) spectroscopy as detection method. The reaction proceeded well and gave high yields provided proper stirring was maintained. Alumina, an aluminosilicate and montmorillonate were the most efficient catalysts. These were also the materials that showed the strongest Lewis acidity. In general, there was good correlation between Lewis acidity and efficiency of the material as catalyst for the Friedel-Crafts alkylation. Attempts to reuse the catalyst were not entirely successful. Deactivation occurred after the first run. ESCA indicated that the reduction in performance was due to adsorption of carbonaceous residues on the catalyst.

Full text of DOI:10.1016/j.molcata.2012.09.020

Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Friedel–Crafts alkylation of sodium salicylate with 4-tert butylbenzyl chloride
performed in aqueous dispersions of mesoporous oxides
Zebastian Bohströma,∗, Hanna Härelindb, Krister Holmberga
a Department of Chemical and Biological Engineering, Applied Surface Chemistry, Chalmers University of Technology, SE-412 96 Göteborg, Sweden
b Department of Chemical and Biological Engineering, Applied Surface Chemistry and Competence Centre for Catalysis, Chalmers University of Technology, SE-412 96 Göteborg,
Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Reactant incompatibility is a common problem in organic chemistry. In this study we investigate the
use of concentrated aqueous dispersions of mesoporous oxides to overcome incompatibility in the
Friedel–Crafts reaction between sodium salicylate and 4-tert-butylbenzyl chloride. The mesoporous
material was first impregnated with the water-soluble nucleophile, sodium salicylate, and the “loaded”
particles were then dispersed in the apolar electrophile, 4-tert-butylbenzyl chloride. A range of different
mesoporous oxides and one clay mineral, montmorillonite, were investigated as catalyst for the reaction.
These were all characterised with small angle X-ray scattering (SAXS) and with nitrogen adsorption (BET
and BJH methods). Their Lewis and Brønstedt acidities were determined by ammonia adsorption exper-
iments using diffuse reflection infrared Fourier transform (DRIFT) spectroscopy as detection method.
The reaction proceeded well and gave high yields provided proper stirring was maintained. Alumina, an
aluminosilicate and montmorillonate were the most efficient catalysts. These were also the materials
that showed the strongest Lewis acidity. In general, there was good correlation between Lewis acidity
and efficiency of the material as catalyst for the Friedel–Crafts alkylation. Attempts to reuse the catalyst
were not entirely successful. Deactivation occurred after the first run. ESCA indicated that the reduction
in performance was due to adsorption of carbonaceous residues on the catalyst.
Received 4 January 2012
Accepted 18 September 2012
Available online 26 September 2012
Keywords:
Reactant incompatibility
Friedel–Crafts alkylation
Sodium salicylate
4-tert-Butylbenzyl chloride
Mesoporous oxides
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
water-soluble reactants may lead to problems with poor compat-
ibility between reagents. Performing a Friedel–Crafts alkylation
Friedel–Crafts alkylation is one of the most useful types of reac-
tions in synthetic organic chemistry and a common method for
reaction with incompatible reactants in a two-phase system would
require the use of a phase transfer catalyst (PTC). The PTC, usually
a quaternary ammonium compound or a crown ether, is a rela-
tively toxic species and needs to be completely removed from the
product after completed reaction [6]. As an alternative to the PTC
approach, a solvent capable of dissolving both hydrophobic and
hydrophilic reactants may be used. Examples of such solvents are
the polar aprotic liquids, such as dimethylsulfoxide, acetonitrile
and tetrahydrofuran. These solvents are relatively expensive and
have non-negligible toxicity [7–9]. It has previously been reported
that a suspension of mesoporous oxide particles in an apolar
medium is an environmentally attractive approach to the problem
of reactant incompatibility [10–12]. The hydrophilic mesoporous
particles are filled with an aqueous solution of the polar reactant
and the surrounding apolar phase contains the apolar reactant. If
the apolar reactant is a liquid, it may constitute the apolar phase,
i.e., the reaction is then performed without any organic solvent. The
environmental and safety advantages of such reaction systems are
obvious.
C
C bond formation to aromatic rings [1]. Typically, an alkyl halide
is reacted with an arene using a strong Lewis acid, such as AlCl3,
ZnCl2 or BF3 [1]. The alkyl halide may be replaced by an alkene
or some other species capable of generating a carbocation when
exposed to strong acid [1]. The carbocation formed acts as the
electrophile in an electrophilic aromatic substitution. Also strong
Brønstedt acids, such as H2SO4 and HF, can be used as catalyst [2].
Liquid phase Friedel–Crafts alkylations performed in large indus-
trial scale, employing Lewis or Brønstedt mineral acids, give rise to
large volumes of corrosive waste streams [3,4].
Increased public awareness and more rigid environmental leg-
islation have elevated the drive towards sustainable development
and cleaner, less hazardous synthetic routes [5]. Environmen-
tally benign synthesis routes sometimes suffer from increased
cost and/or reduced process effectiveness and the use of
Solid catalysts are attractive for industrial processes because
they are normally relatively easy to separate from the product and
∗
Corresponding author. Tel.: +46 031 772 2812.
E-mail address: zebastian.bostrom@chalmers.se (Z. Bohström).
1381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2012.09.020

172
Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
they can often be reused. A number of solid, acidic catalysts, such as
zeolites [13], clays [14] and metal oxides [15], have been used for
the Friedel–Crafts alkylation reaction performed in conventional
organic solvent systems [16]. Mesoporous oxides, such as meso-
porous SiO2/Al2O3, are particularly attractive since they combine a
large and well-defined internal surface area with pore dimensions
large enough not to be easily clogged by the reaction products. Clog-
ging of the pores is a well-known problem with zeolites [17,18].
Clays are also useful but they have lower surface areas than meso-
porous oxides and zeolites and they therefore expose fewer active
sites per weight unit of catalyst.
2.3. Friedel–Crafts alkylation in mesoporous slurry
Mesoporous material (0.60 g), aqueous sodium salicylate solu-
tion (1.50 g, 4.68 mmol), and 4-tert-butylbenzyl chloride (0.90 g,
4.93 mmol) were added in that order into a round bottomed flask
fitted with a magnetic stirrer. The slurry was heated in an oil bath
under a condenser at 90 ◦C. Aliquots were taken at various intervals
during 24 h and analysed with 1H NMR and 13C NMR for the reaction
products. Sodium 3-(4-tert-butylbenzyl)-2-hydroxybenzoate was
characterized by the following NMR pattern: 1H NMR (400 MHz,
CDCl3) ı (ppm): 1.35 (s, 3H), 3.96 (s, 2H), 6.96–7.04 (d/t, H),
7.15 (d, H), 7.34 (d, H), 7.40 (d, H), 7.75–7.85 (d, H). 13C NMR
(400 MHz, CDCl3) ı (ppm): 31.3, 34.2, 35.1, 120.9, 123.3, 125.5,
128.8, 130.0, 132.2, 135.3, 138.4, 148.8, 162.5, 171.0. Sodium 5-
(4-tert-butylbenzyl)-2-hydroxybenzoate was characterized by the
following NMR pattern: 1H NMR (400 MHz, CDCl3) ı (ppm): 1.35 (s,
3H), 3.96 (s, 2H), 6.96–7.04 (d/t, H), 7.15 (d, H), 7.34 (d, H), 7.40 (d,
H), 7.75–7.85 (s, H). 13C NMR (400 MHz, CDCl3) ı (ppm): 31.3, 34.2,
41.3, 119.7, 123.3, 125.5, 128.8, 132.2, 133.8, 134.2, 148.8, 162.7,
171.0.
We here report the use of a concentrated suspension of
acidic mesoporous oxide particles as reaction system for
a
Friedel–Crafts reaction, alkylation of sodium salicylate with 4-tert-
butylbenzyl chloride. A range of different mesoporous oxides is
explored. Sodium salicylate, which is water-soluble, resides in the
water-filled pores of the particles and 4-tert-butylbenzyl chloride
constitutes the continuous phase. The catalytic sites are integrated
into the pore walls. The suspension is kept under stirring and the
reaction occurs at the interface between the two liquid phases, i.e.,
at the pore openings.
2.4. Recycling of mesoporous oxides
2. Experimental
Mesoporous material, obtained after reaction at 90 ◦C and a
stirring rate of 800 rpm, was recovered by vacuum filtration and
washed with CH2Cl2. The material was air dried at room tempera-
ture and was then transferred to a reaction flask and used for a new
24 h reaction under the same experimental conditions. The meso-
porous material was stored at room temperature in air atmosphere
between the experiments. To validate the washing procedure the
mesoporous material was impregnated with the reactants, 4-tert-
butylbenzyl chloride and sodium salicylate, and 1H NMR was used
to confirm that no product resided in the mesoporous material after
the washing procedures.
2.1. Materials and reagents
Pluronic P123 (BASF), ethanol (Kemetyl, 99.5%), aluminium
chloride (Aldrich, 98.0%), silicon tetrachloride (Aldrich, 99%), zinc
chloride (Sigma–Aldrich 98%) aluminium isopropoxide (Aldrich,
98%), hexadecyltrimethylammonium bromide (CTAB; Sigma, 98%)
and sodium hydroxide (Aldrich, 97% A.C.S.), were used as received
for preparation of catalyst samples. 4-tert-Butylbenzyl chloride
(99%, Aldrich) and sodium salicylate (99.5%, Aldrich) were the
reactants used to perform the Friedel–Crafts alkylation reac-
tion, montmorillonite (K30, Aldrich) was used as catalyst, and
chloroform-d (ARMAR Chemicals, 99.8 atom % D) was used for the
13C- and 1H NMR analyses.
2.5. Analysis techniques
1H NMR spectra were recorded at 400 MHz using a JEOL,
model Eclipse FT-NMR Oxford instrument. The consumption of
4-tert-butylbenzyl chloride was determined by monitoring the
decrease in the signal from the CH2Cl group in 4-tert-butylbenzyl
chloride, which appears at ı (ppm) = 4.6–4.7. The yield of the alkyl-
ation product, two isomers of sodium 4-tert-butylbenzylsalicylate,
was estimated by monitoring the increase in the signal of the
Ar CH2 Ar group, which appears at ı (ppm) = 3.9–4.0. The yield
was calculated from the ratio between the appearing and the disap-
pearing peaks. The 1H NMR experiments were made at 25 ◦C using
CDCl3 as solvent.
2.2. Preparation of mesoporous materials
2.2.1. Mesoporous silica, alumina, and mixed alumina/silica
The materials were prepared following the method described
by Yang et al. [19]. The template, EO-PO-EO triblock copolymer
P123 (10.0 g), was dissolved in ethanol (100 g) under stirring at
room temperature. After 30 min the chloride precursor was step-
wise added and the stirring was continued for 30 min. The solution
was transferred to an open petri dish and gelled by standing for an
appropriate time in the oven. The thin film obtained was directly
calcined; see Table 1 for conditions, amounts and reagents.
13C NMR spectra were recorded using the same instrument. The
products; sodium 3-(4-tert-butylbenzyl)-2-hydroxybenzoate and
sodium 5-(4-tert-butylbenzyl)-2-hydroxybenzoate, could be iden-
tified by the position of the signal from the Ar CH2 Ar group,
which appears at ı (ppm) = 35.1 and 41.3, respectively. The 13C NMR
experiments were made at 25 ◦C using CDCl3 as solvent and run for
14 h. 13C NMR was only used for analysing the product composition
for reactions with the three best performing catalytic materials;
Al2O3, Al2O3/SiO2-5 and montmorillonite.
Gas chromatography (GC) was performed on a Varian 3400
GC fitted with a SUPELCO Chromosorb WHP 80-100 Mesh, SS
1.8 m × 1/8 mm × 2.0 mm diatomite column, and a flame ionization
detector (FID). The injected sample in CHCl3 was run using a 30 min
program with a temperature ramp ranging from 40 to 300 ◦C. The
program is capable of eluting C5–C30 aliphatic hydrocarbons, as
proved by the calibration run.
2.2.2. Mesoporous alumina/silica
The material was synthesised essentially following the prepara-
tion method described by Wang et al. [20]. Aluminium isopropoxide
was added to a water solution (90 g) containing sodium hydroxide
(7.8 g) under stirring at room temperature. Additional water was
added (910 g) and the solution was stirred for 30 min. The tem-
plate was added and the temperature was fixed at 25 ◦C. When
the template had completely dissolved, tetraethyl orthosilicate was
added and the resulting gel was stirred for 2 h. Hydrochloric acid
(37 wt%) was drop-wise added until the pH had been adjusted to
12. After 4 h under stirring at pH 12 the pH was lowered to 11 and
the stirring was continued for another 12 h. The solid was vacuum
filtered, washed with water and calcined. See Table 1 for conditions,
amounts and reagents.
Electron spectroscopy for chemical analysis (ESCA or XPS)
was performed on a Quantum 2000 scanning ESCA microprobe

Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
173
Table 1
Reagents, amounts and conditions used for synthesis of the mesoporous oxides.
Material
Precursor
Type
Template
Type
Calcination
Ramp. (h)
Amount (g)
Amount (g)
Daysa
Time (h)
Temp. (◦C)
TiO2
WO3
SiO2
Al2O3
(A)-Al2O3/SiO2-1
(A)-Al2O3/SiO2-2
(A)-Al2O3/SiO2-3
(A)-Al2O3/SiO2-4
(A)-Al2O3/SiO2-5
(B)-Al2O3/SiO2-1
(B)-Al2O3/SiO2-2
TiCl4
WCl6
SiCl4
AlCl3
AlCl3/SiCl4
AlCl3/SiCl4
AlCl3/SiCl4
AlCl3/SiCl4
AlCl3/SiCl4
Si(OEt)4/Al(OiPr)3
Si(OEt)4/Al(OiPr)3
18.9
39.8
17.0
13.3
12.7/1.7
10.0/4.2
6.7/8.5
3.3/12.7
1.3/15.3
10.8/6.25
12.8/4.25
P123
P123
P123
P123
P123
P123
P123
P123
P123
CTAB
CTAB
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
32.7
32.7
2
5
7
5
3
3
3
3
3
–
–
1
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
400
400
400
400
500
500
500
500
500
520
520
4
16
16
a
Days of ageing at 40 ◦C.
from Physical Electronics. An A1K␣ (1486.6 eV) X-ray source was
used and the beam size was 100 ␮m. The analysed area was
500 ␮m × 500 ␮m and the take off angle was 45◦ with respect to
the sample surface. The information depth was 4–5 nm.
Determination of the specific surface area was performed
on an ASAP 2010 instrument, using nitrogen adsorption and
the BET (Brunauer–Emmett–Teller) method [21]. The pore size
distribution was calculated from the isotherms using the BJH
(Barett–Joyner–Halenda) procedure [22]. All samples were dried
at 225 ◦C in a vacuum oven for approximately 3 h before the mea-
surement.
Small angle X-ray scattering (SAXS) was performed with a
Kratky compact small angle system on a HECUS Mbraun, Graz
instrument. All runs were performed under vacuum at 50 kV and
40 mA. The runs were performed with monochromatic Cu K␣1 radi-
ation using a Ni filter. The samples were prepared by grinding the
mesoporous oxides and then placing the particles in a paste holder
with thin mica windows.
Samples for the transmission electron microscopy (TEM), run on
a JEOL 1200 EX II instrument at 120 kV, were prepared by placing
a drop of an ethanol dispersion of the mesoporous material onto a
copper Holey grid.
In situ FTIR (Fourier transform infrared) spectroscopy measure-
ments were carried out using a BioRad FTS 6000 spectrometer
equipped with a Harrick Praying Mantis DRIFT (diffuse reflec-
tion infrared Fourier transform) reaction cell [23,24]. The sample
was put in the DRIFT cell and the gases, Ar, NH3 and O2, were
introduced via mass flow controllers (Bronkhorst Hi-Tech) to the
cell. The samples were initially pre-treated in O2 (8%) at 500 ◦C
for 30 min and then evacuated in Ar for 15 min (keeping the total
flow rate constant at 200 ml/min). Adsorption of NH3 (1000 ppm)
was performed at 25 ◦C during 30 min, followed by evacuation in
Ar (6 scans/min, 1 cm−1 resolution). Background spectra were col-
lected in Ar (6 scans/min, 1 cm−1 resolution).
3. Results and discussion
3.1. The mesoporous materials
The block copolymer Pluronic P123 was used as template for
the mesoporous materials. The physical chemical properties, such
as phase behaviour and gelling, of this amphiphilic polymer have
been extensively studied [19,25]. TEOS (tetraethyl orthosilicate)
was used as silica source and calcination was employed as template
removal method. The specific surface area and the pore character-
istics of the mesoporous materials were examined with nitrogen
sorption, see Table 2.
All the mesoporous materials displayed nitrogen adsorption
isotherms of type IV, typical for mesoporous materials, see Fig. 1
[26]. The type of hysteresis differed between the prepared meso-
porous materials. SiO2 displayed a H2 hysteresis loop, characteristic
for porous materials [26]. Al2O3, TiO2, (A)-Al2O3/SiO2-1-2, (B)-
Al2O3/SiO2-2 and montmorillonite displayed adsorption hysteresis
H3, characteristic of particle aggregates with plate-like and slit
shaped pores [26]. (A)-Al2O3/SiO2-3-5 had hysteresis loops of type
H1. The WO3 material also had a hysteresis loop of type H1; how-
ever, this material did not display a narrow pore size distribution.
As can be seen from Table 2, the materials span a broad range of
specific surface areas, with SiO2 at the high end and WO3 and (B)-
Al2O3/SiO2-1 at the low end. The surface area of WO3, 31 m2 g−1, is
Table 2
Specific surface area and pore characteristics of the mesoporous materials obtained from the BET and BJH methods.
a
b
c
Material
Specific surface area (m2 g−1
)
Pore size (nm)b
Pore size (nm)c
Pore volume (cm3 g−1
)
Pore volume (cm3 g−1
)
Mesoporous oxides
TiO2
WO3
SiO2
Al2O3
(A)-Al2O3/SiO2-1
(A)-Al2O3/SiO2-2
(A)-Al2O3/SiO2-3
(A)-Al2O3/SiO2-4
(A)-Al2O3/SiO2-5
(B)-Al2O3/SiO2-1
(B)-Al2O3/SiO2-2
108
31
9.12
11.5
2.85
10.0
9.94
8.87
6.56
6.40
4.00
9.12
7.03
8.21
10.1
2.88
11.9
8.53
7.90
5.76
5.84
3.73
8.42
6.71
0.30
0.10
0.54
0.92
0.79
0.77
0.48
0.67
0.52
0.15
0.20
0.30
0.10
0.56
0.76
0.80
0.79
0.50
0.70
0.64
0.16
0.22
928
346
258
319
295
440
712
81
167
Other
Montmorillonite
253
6.06
5.30
0.35
0.37
a
Calculated from nitrogen sorption using the BET method.
Calculated from the adsorption branch of the nitrogen sorption isotherm using the BJH method.
Calculated from the desorption branch of the nitrogen sorption isotherm using the BJH method.
b
c

174
Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
Fig. 1. Nitrogen sorption isotherms from the adsorption branch: A: Al2O3; B: SiO2; C: (A)-Al2O3/SiO2-3; D: TiO2; E: WO3; F: (B)-Al2O3/SiO2-1; G: montmorillonite; and H:
(A)-Al2O3/SiO2-5.
remarkably small. Repeated material preparations and gas adsorp-
tion experiments gave consistent results. As can also be seen from
Table 2, the mixed oxide materials labelled (A)- had larger surface
areas compared to those labelled (B)-. This shows the importance
of the choice of precursor; the former materials were made from
metal chlorides and the latter from an alkoxide (see Table 1).
The mesoporous materials were also characterized by transmis-
sion electron microscopy (TEM) and selected micrographs can be
found in the supplementary data section. The materials were ana-
lysed by small angle X-ray scattering (SAXS). The SAXS analysis
indicated that none of the prepared mesoporous materials pos-
sessed long range order.
In situ DRIFT experiments with ammonia adsorption were per-
formed in order to characterize the surface acidity of the samples
(Fig. 2). Several peaks appeared when the samples were exposed to
ammonia and the peaks differed depending on the acidic charac-
teristics of the samples. For clarity the peaks are first discussed for
one of the samples, (A)-Al2O3/SiO2-5, and then the other samples
are discussed in relation to the results for this sample.
When ammonia was adsorbed on the (A)-Al2O3/SiO2-5 sample,
multiple peaks at 3390, 3299, 3237, 1625, and 1289 cm−1 evolved.
According to Wang et al. and Tsyganenko et al., these peaks can
be ascribed to NH3 adsorbed on Lewis acid sites [27–30]. A peak
+
Fig. 2. DRIFT spectra showing surface NH3 and NH4 species after adsorption of
ammonia during 30 min.

Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
175
Fig. 3. Friedel–Crafts alkylation of sodium salicylate with 4-tert-butylbenzyl chloride resulting in a mixture of the two monoalkylated isomers: sodium 3-(4-tert-butylbenzyl)-
2-hydroxybenzoate (left) and 5-(4-tert-butylbenzyl)-2-hydroxybenzoate (right).
at 3187 cm−1 likely arises from Lewis-NH3 species perturbed by
H-bonding to oxygen in the alumina structure or to a surface
OH group [27–30]. Another group of peaks at 1702, 1502 and
sodium salicylate, and the “loaded” particles were then dispersed
in the apolar electrophile, 4-tert-butylbenzyl chloride. Thus, no sol-
vent was employed. The slurry was kept under stirring and aliquots
were taken at specific time intervals in order to assess the progress
of the reaction.
The gas chromatography analysis was primarily performed to
identify any dialkylated product. However, only peaks derived from
the solvent, starting material and monoalkylated products were
obtained. Thus, one can conclude that under the conditions used
dialkylation did not occur.
The conversion of starting material to product was monitored
by 1H NMR for all the reactions. For the reactions over the three best
performing catalytic materials; Al2O3, (A)-Al2O3/SiO2-5 and mont-
morillonite, 13C NMR was also used, primarily to assess the relative
amounts of isomers formed. While Al2O3 gave only substitution in
the ortho position of the phenolic hydroxyl group, montmorillonite
and (A)-Al2O3/SiO2-5 gave the para and the ortho isomers in rel-
ative ratios of 9:1 and 1:1, respectively. The meta isomer was not
detected in any of the reactions. The difference in regioselectivity
may be correlated to differences in pore size of the catalysts. The
material with the smallest pore size, (A)-Al2O3/SiO2-5, gave the
highest amount of para isomer while Al2O3, with the largest pore
size, gave only the ortho isomer. Inspection of molecular models
shows that the ortho isomer is more bulky than the para isomer.
The reaction was performed with a small excess of alkyl-
ating agent and the reaction profiles obtained with a range of
mesoporous oxide catalysts, as well as with montmorillonite, are
shown in Fig. 4. As can be seen from this figure and also from
Table 3, where conversions after 24 h reaction are given, Al2O3 and
1466 cm−1 can be attributed to formation of NH4 species on
+
Brønstedt acid sites [27–30]. Furthermore, the negative peaks
at 3747 and 3705 cm−1 are indicative of blocking of surface
OH species by adsorbed NH3 [31]. The (A)-Al2O3/SiO2-5, (B)-
Al2O3/SiO2-2, montmorillonite, Al2O3, SiO2, (A)-Al2O3/SiO2-2 and
(A)-Al2O3/SiO2-3 samples showed peaks in the Lewis acid regions,
however with varying intensities. The other samples showed minor
or negligible peaks in these regions. Intense peaks connected to
Lewis acid sites imply strong Lewis acidic character of the sample
and a qualitative analysis is possible. Hence, the Lewis acidity was
found to decrease in the following order: (A)-Al2O3/SiO2-5 ≈ (B)-
Al2O3/SiO2-2 ≈ montmorillonite ≥ Al2O3 > SiO2 > (A)-Al2O3/SiO2-
2 ≈ (A)-Al2O3/SiO2-3 > (A)-Al2O3/SiO2-4 > TiO2, (A)-Al2O3/SiO2-1.
The peaks connected to Brønstedt acid sites (1702, 1502
and 1466 cm−1) were significant for the (A)-Al2O3/SiO2-5, (B)-
Al2O3/SiO2-2, Al2O3/SiO2-2, (A)-Al2O3/SiO2-3, Al2O3/SiO2-4 and
the montmorillonite samples but absent for the remaining sam-
ples. The Brønstedt acidity decreased in the following order:
(A)-Al2O3/SiO2-5 ≈ (B)-Al2O3/SiO2-2 > (A)-Al2O3/SiO2-2 ≈ (A)-
Al2O3/SiO2-3 ≥ (A)-Al2O3/SiO2-4 ≈ montmorillonite > Al2O3, SiO2,
TiO2, (A)-Al2O3/SiO2-1.
3.2. The Friedel–Crafts alkylation
The reaction studied was alkylation of sodium salicylate with
4-tert-butylbenzyl chloride, see Fig. 3. The reaction is an example
of a Friedel–Crafts reaction in which the reactants differ widely
in terms of polarity. The nucleophile, sodium salicylate, is water
soluble while the electrophile, 4-tert-butylbenzyl chloride, has very
low solubility in water.
The Friedel–Crafts alkylation is a single step C C coupling reac-
tion [32]. The substrate used in this study, sodium salicylate,
contains one strongly activating para and ortho directing group,
OH, and one moderately deactivating meta directing group, COO−;
hence, the nucleophilicity should be high and one can expect that
the product is composed of the two isomers shown in Fig. 3 [32].
Dialkylation, i.e., a product with the 4-tert-butylbenzyl group in
both ortho and para position of the phenolic hydroxyl group, is also
conceivable.
The Friedel–Crafts reaction generally requires a strong acid
catalyst and a Lewis acid is usually employed. In this work the
catalyst was a mesoporous oxide; either silica, alumina or mixed
silica/alumina. The catalytic ability of the oxides can be expected
to be related primarily to their Lewis acidity but for the aluminosil-
icates the Brønstedt acid sites may also contribute. The negative
charge carried by the constituent aluminium ions, which are in
4-coordination with oxygen, is balanced by cations such as H+,
giving rise to Brønstedt acidity. The reaction was performed in a
concentrated suspension of the silica, alumina or alumina/silica
particles. Montmorillonite, a mesoporous clay material, was also
included in the study its characteristics with respect to specific
surface area, pore size, etc., is given in Table 2. The mesoporous
material was first impregnated with the water-soluble nucleophile,
Fig. 4. Reaction profiles, expressed as conversion vs. time, for the Friedel–Crafts
alkylation of sodium salicylate with 4-tert-butylbenzyl chloride performed at 90 ◦
at 800 rpm with different solid catalysts. ( ) Al2O3, (-ꢀ-) (A)-Al2O3/SiO2-5, (
C
)
montmorillonite, (-ꢀ-) SiO2 with 0.06 g of AlCl3, ( ) (B)-Al2O3/SiO2-2, ( ) SiO2 with
0.03 g of AlCl3, (-ꢁ-) SiO2 with 0.03 g of ZnCl2, ( ) SiO2 and ( ) (A)-Al2O3/SiO2-3.

176
Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
Table 3
Conversion after 24 h for the Friedel–Crafts alkylation of sodium salicylate with 4-tert-butylbenzyl chloride performed with different solid catalysts.
Entry
Mesoporous catalyst
Reaction temperature (◦C)
Stirring (rpm)
Conversion (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
TiO2
WO3
SiO2
Al2O3
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
60
45
60
45
60
45
90
90
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
800
400
0
35
28
36
87
29
54
50
45
89
64
86
88
80
79
31
37
0
(A)-Al2O3/SiO2-1
(A)-Al2O3/SiO2-2
(A)-Al2O3/SiO2-3
(A)-Al2O3/SiO2-4
(A)-Al2O3/SiO2-5
(B)-Al2O3/SiO2-1
(B)-Al2O3/SiO2-2
Montmorillonite
SiO2 + AlCl3 (2:1)a
SiO2 + AlCl3 (1:1)a
SiO2 + ZnCl2 (1:1)a
(B)-Al2O3/SiO2-2
(B)-Al2O3/SiO2-2
(A)-Al2O3/SiO2-5
(A)-Al2O3/SiO2-5
Al2O3
42
1
45
4
60
34
Al2O3
Al2O3
Al2O3
a
Stoichiometric (1:1) and twice the stoichiometric (2:1) amount of acid was added (stoichiometric amounts calculated on 4-tert-butylbenzyl chloride).
b
Conversion was calculated as the molar ratio of product to starting material after 24 h with 1H NMR.
(A)-Al2O3/SiO2-5 were the most efficient catalysts. Then followed
scaled up three times to be able to quantitatively isolate the prod-
uct. The yield obtained was 61%, i.e., considerably lower than the
conversion figure of 88% given in Table 3. The discrepancy between
the conversion and the yield of isolated product is probably partly
due to formation of undefined carbonaceous material that adsorbs
on the solid catalyst. This will be discussed in Section 3.3.
montmorillonite, SiO2 with added AlCl3 and (B)-Al2O3/SiO2-2. One
may note that at least for SiO2 + AlCl3, addition of more than the
stoichiometric amount of the Lewis acid only has a marginal effect.
Then follows a group of materials with approximately equal effi-
ciency: (A)-Al2O3/SiO2-3, SiO2 with added ZnCl2 and SiO2 without
added Lewis acid. These catalysts only gave a 25–40% conversion
of sodium salicylate after 24 h of reaction. It is noteworthy that the
best mesoporous oxides were more efficient as catalysts than SiO2
with twice the stoichiometric amount of AlCl3 added. One should
note that the conversion figures listed in Table 3 are not isolated
yields but molar ratios of reactants to product as determined from
the 1H NMR spectra. In order to obtain real yield figures one of the
reactions, the one performed with montmorillonite as catalyst, was
As mentioned above, the Lewis acidity, as measured by ammonia
adsorption, decreased in the following order: (A)-Al2O3/SiO2-
5 ≈ (B)-Al2O3/SiO2-2 ≈ montmorillonite ≥ Al2O3 > SiO2 > (A)-
Al2O3/SiO2-2 ≈ (A)-Al2O3/SiO2-3 > (A)-Al2O3/SiO2-4 > TiO2,
(A)-Al2O3/SiO2-1. It can clearly be seen that the order of cat-
alyst efficiency has a good fit with the order of Lewis acidity.
The two best catalysts, Al2O3 and (A)-Al2O3/SiO2-5, are also the
materials with the most pronounced Lewis acid character.
The order of Brønstedt acidity deviated slightly from that of
Lewis acidity. Both Al2O3 and SiO2 are further down on the
Brønstedt acidity list than on the Lewis acidity list. In general, the
correlation between catalyst efficiency and catalyst acidity seemed
to be stronger for Lewis acidity than for Brønstedt acidity. A good
correlation between Lewis acid strength and efficiency as a catalyst
for Friedel–Crafts alkylation has been reported before [1,32–36].
Another parameter that might play a role for the catalyst effi-
ciency is the specific surface area. The Lewis acid sites are situated
on the walls of the mesoporous materials and a larger surface is
likely to expose more catalytically active sites, which may be ben-
eficial for the reaction kinetics. We found no correlation, however,
between specific surface area and reaction rate. Al2O3, which was
the most catalytically active material, had only a moderately large
surface area (346 m2 g−1) and the material with the largest surface
area, SiO2 (with a value of 928 m2 g−1), was not a very active cata-
lyst. Montmorillonite, with a specific surface area of 253 m2 g−1 is
relatively far down on that list but number 3 in terms of catalytic
activity. A logical interpretation of these results is that the surface
areas of all the investigated porous materials are above the level at
which the specific surface area becomes a rate limiting factor. This is
contrary to the results from a previous study carried out in our labo-
ratory, where we investigated a Heck type carbon–carbon coupling
reaction between two apolar reactants and with the water-filled
pores of mesoporous oxides serving as host for a homogeneous
catalyst [37]. For that reaction, the yield was almost proportional
Fig. 5. Reaction profiles, expressed as conversion vs. time, for the Friedel–Crafts
alkylation of sodium salicylate with 4-tert-butylbenzyl chloride at 90 ◦C and at dif-
ferent stirring rates using Al2O3 as catalyst at (-ꢂ-) 800 rpm, (-᭹-) 400 rpm, and
(-ꢀ-) no agitation.

Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
177
Fig. 6. Reaction profiles, expressed as conversion vs. time, for the Friedel–Crafts alkylation of sodium salicylate with 4-tert-butylbenzyl chloride at different temperatures
for: Al2O3, (A)-Al2O3/SiO2-5; and (B)-Al2O3/SiO2-2. Plots indicated by the symbols (-ꢂ-), (-᭹-), and (-ꢀ-) denote reaction at 90 ◦C, 60 ◦C, and 45 ◦C, respectively.
to the added amount of mesoporous material, i.e., available pore
area.
a new batch reaction. The results from three consecutive runs are
shown in Fig. 7.
The effect of agitation on the reaction rate was studied for
one mesoporous oxide, Al2O3. The reaction profiles for stirring at
800 rpm, 400 rpm and no stirring are shown in Fig. 5. As can be
seen, stirring is an important parameter, as can be expected for
this multi-phase reaction system. One may note, however, that the
reactions proceeded, although slowly, also without stirring.
Table 3 also includes conversion figures for reactions at tem-
peratures lower than 90 ◦C for three of the most efficient catalysts;
Al2O3, (A)-Al2O3/SiO2-5 and (A)-Al2O3/SiO2-2. In Fig. 6 the conver-
sions obtained with these three catalysts have been plotted at three
different temperatures; 90 ◦C, 60 ◦C and 45 ◦C. Again Al2O3 stands
out as a very efficient catalyst.
As can be seen, all three catalysts exhibited an approximate 50%
drop in conversion between the first and the second run. Al2O3 and
(A)-Al2O3/SiO2-5, the two materials with the highest efficiency as
catalysts, then gave approximately the same conversion in the third
run while (B)-Al2O3/SiO2-2 continued down.
The Al2O3 material was investigated by TEM and by ESCA after
the consecutive experiments. The microscopy study showed no
significant differences in pore structure after the first and the sec-
ond run. The TEM images can be found in the supplementary data
section.
ESCA was performed on the Al2O3 material after the first run
and the result was compared with that of fresh material, i.e., Al2O3
that had not been used as catalyst. There was a striking difference
in carbon content between the used and the unused material, 48%
vs. 11%. (Pure Al2O3 contains no carbon, of course, but a certain per-
centage of C is almost always obtained with ESCA because of the
spontaneous adsorption of hydrocarbons from the environment.
11% carbon is not an unreasonable value.) The very large increase
in carbon content is most likely due to organic residues from the
3.3. Recycling of the catalyst
Attempts to recycle the mesoporous oxides were made with
three of the most efficient catalysts; Al2O3, (A)-Al2O3/SiO2-5 and
(B)-Al2O3/SiO2-2. After completed reaction the inorganic material
was removed by filtration, washed with CH2Cl2 and then used in

178
Z. Bohström et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 171–178
Acknowledgments
Mrs. Anne Wendel is thanked for the ESCA analysis. The Swedish
Research Council is acknowledged for economic support of the
project.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.molcata.2012.09.020.
References
[1] G.A. Olah, Friedel–Crafts and Related Reactions. Vol. II. Alkylation and Related
Reactions. Part 1, William Clowes and Sons, London, 1964.
[2] N.E. Poh, H. Nur, M.N.M. Muhid, H. Hamdan, Catal. Today 114 (2006) 257–262.
[3] N. Candua, S. Wuttkeb, E. Kemnitzb, S.M. Comana, V.I. Parvulescua, Appl. Catal.
A 391 (2011) 169–174.
[4] R. Luque, V. Budarin, J.H. Clark, P. Shuttleworth, R.J. White, Catal. Commun. 12
(2011) 1471–1476.
[5] J.H. Clark, D.J. Macquarrie, Environ. Chem. Soc. Rev. 25 (1996) 303–309.
[6] M.R. Scott, J.E. Hutchison, J. Chem. Ed. 12 (2000) 1627–1628.
[7] P. Wexler, Encyclopedia of Toxicology, second ed., Elsevier Academic Press, San
Diego, 2005.
[8] N.G. Anderson, Practical Process Research & Development, Academic Press, San
Diego, 2000.
Fig. 7. Recycling experiments performed with the catalysts Al2O3 (black), (A)-
Al2O3/SiO2-5 (grey), and (B)-Al2O3/SiO2-2 (white).
[9] R. Gani, C. Jimeénez-Gonàlez, D.J.C. Constable, Comput. Chem. Eng. 29 (2005)
1661–1676.
[10] P. Handa, K. Holmberg, M. Sauthier, Y. Castanet, A. Mortreux, Micropor. Meso-
por. Mater. 116 (2008) 424–431.
[11] Z. Bohström, I. Rico-Lattes, K. Holmberg, Green Chem. 12 (2010) 1861–1869.
[12] Z. Bohström, J. Porous Mater. 19 (2012) 921–933.
[13] V.R. Choudhary, S.K. Jana, N.S. Patil, S.K. Bhargava, Micropor. Mesopor. Mater.
57 (2003) 21–35.
[14] K. Wilson, D.J. Adams, G. Rothenberg, J.H. Clark, J. Mol. Catal. A 159 (2000)
309–314.
[15] V.R. Choudhary, K. Mantri, Catal. Lett. 81 (2002) 163–168.
[16] G. Sartori, R. Maggi, Chem. Rev. 106 (2006) 1077–1104.
[17] M. Guidotti, C. Canaff, J-M. Coustard, P. Magnoux, M. Guisnet, J. Catal. 230 (2005)
375–383.
Friedel–Crafts reaction that were not removed by the washing
procedure. It is likely that this carbonaceous material is the cause
of the strong decrease in catalyst efficiency between the first and
the second run. Even if the pores in the Al2O3 material are large
(10–12 nm, see Table 2) and therefore not easily clogged (which
would have been seen in the TEM images) the residual organics
may adsorb on the pore walls and reduce the amount of available
catalytically active sites. No attempts have been made to remove
the organic contaminates by calcination.
[18] E.G. Derouane, C.J. Dillon, D. Bethell, S.B. Derouane-Abd Hamid, J. Catal. 187
(1999) 209–218.
[19] P. Yang, D. Zhao, D.I. Margolese, B.F. Chmelka, G.D. Stucky, Nature 396 (1998)
152–155.
4. Conclusion
[20] Y. Wang, N. Lang, A. Tuel, Micropor. Mesopor. Mater. 93 (2006) 46–54.
[21] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.
[22] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380.
[23] A. Hinz, M. Skoglundh, E. Fridell, A. Andersson, J. Catal. 201 (2001) 247–257.
[24] P. Alexandridis, Coll. Surf. A 95 (1995) 1–46.
[25] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T.
Siemieniewska, Pure Appl. Chem. 57 (1985) 603–619.
[26] Z.-M. Wang, M. Yamaguchi, I. Goto, M. Kumagai, Phys. Chem. Chem. Phys. 2
(2000) 3007–3015.
[27] A.A. Tsyganenko, D.V. Pozdnyakov, V.N. Filimonov, J. Mol. Struct. 29 (1975)
299–318.
[28] A.A. Tsyganenko, V.N. Filimonov, J. Mol. Struct. 19 (1973) 579–589.
[29] A.A. Tsyganenko, P.P. Mardilovich, J. Chem. Soc., Faraday Trans. 92 (1996)
4843–4852.
[30] G. Busca, Phys. Chem. Chem. Phys. 1 (1999) 723–736.
[31] M. Wallin, C-J. Karlsson, A. Palmqvist, M. Skoglundh, Topics Catal. 30–31 (2004)
107–114.
[32] J. March, Avanced Organic Chemistry: Reactions, Mechanisms, and Structure,
third ed., John Wiley & Sons Inc., New York, 1985.
[33] G.A. Olah, S. Kobayashi, M. Tashiro, J. Am. Chem. Soc. 94 (1972) 7448–7461.
[34] A.G. Smith, J.S. Johnson, Org. Lett. 12 (2010) 1784–1787.
[35] M.J. Gracia, E. Losada, R. Luque, J.M. Campelo, D. Luna, J.M. Marinas, A.A. Romero,
Appl. Catal. A 349 (2008) 148–155.
[36] K.-I. Shimizu, K. Niimi, A. Satsuma, Appl. Catal. A 349 (2008) 1–5.
[37] P. Handa, M. Stjerndahl, K. Holmberg, Micropor. Mesopor. Mater. 100 (2007)
146–153.
We have demonstrated that the Friedel–Crafts reaction
between two incompatible reactants, sodium salicylate and 4-
tert-butylbenzyl chloride, can be performed in a concentrated
dispersion of a mesoporous material, provided the material is a
strong Lewis acid. No external Lewis acid is then needed. Out of a
range of tested materials the highest reaction rates were obtained
with Al2O3 and (A)-Al2O3/SiO2-5, which were also the materials
that had the highest Lewis acidity. In general there was a strong
correlation between Lewis acidity, as determined by ammonia
adsorption, and efficiency as catalyst in the Friedel–Crafts alkyl-
ation. The surface area of the materials seemed not to be important,
however. Agitation of the reaction system was needed in order to
obtain proper reaction rate. Attempts to reuse the catalyst were
made but with limited success. The activity dropped considerably
after the first run. ESCA analysis revealed large amounts of carbona-
ceous material on the catalyst. Most likely these organic residues
have adsorbed on the pore walls, thus reducing the amount of avail-
able catalytically active sites.