www.elsevier.com/locate/micromeso

Microporous and Mesoporous Materials 79 (2005) 137–144

Aluminum chloride grafted mesoporous molecular sievesas alkylation catalysts

David Dube a, Sebastien Royer a, Do Trong On a, Francois Beland b, Serge Kaliaguine a,*

a Department of Chemical Engineering, Laval University, Pavillon A-Pouliot, Sainte Foy, Canada G1K 7P4b SiliCycle Inc., 1200 Ave. St-Jean-Baptiste, Suite 114, Quebec City, Canada G2E 5E8

Received 15 September 2004; received in revised form 30 October 2004; accepted 1 November 2004

Available online 8 December 2004

Abstract

The aim of this work is to study the influence of different mesoporous molecular sieves (MMSs) type and the effect of pore diam-

eter on MMSs silica grafted aluminum chloride catalysts and their activity in benzene alkylation by 1-dodecene. Four samples (two

hexagonal and two wormhole like structures with different pore diameters) were prepared by post-synthesis grafting reaction in

moisture free conditions (Schlenk technique). The physico-chemical properties were monitored by N2 adsorption, XRD, atomic

adsorption, 29Si and 27Al MAS NMR. Acidity data were obtained by TPD of ammonia, and FTIR of adsorbed pyridine. The alkyl-

ation of benzene is carried out in a Schlenk tube under dry conditions at room temperature. The products were analyzed by GC-MS.

Grafting aluminum chloride on MMSs increases the yield of monoalkylated products.

� 2004 Published by Elsevier Inc.

Keywords: Aluminum chloride; Mesoporous materials; Lewis acid; Solid acid catalyst; Alkylation reaction

1. Introduction

For decades, linear alkylbenzenes (LABs) have been

the precursors of sulfonated alkyl benzenes, which are

significant commercial detergents. The process to syn-

thesize them is usually carried out by Friedel-Craft

alkylation of linear olefins using hydrogen fluoride or

aluminum chloride as homogeneous catalysts. The useof these catalysts presents severe problems. For exam-

ple, aluminum chloride is difficult to separate after reac-

tion and produces a large amount of waste effluent. To

solve these problems, attempts have been made to chem-

ically support aluminum chloride on mesoporous molec-

ular sieves (MMSs) silica type [1]. Aluminum chloride

can be easily grafted onto the surface of such solids to

produce very active catalysts. Drago et al. grafted alumi-

1387-1811/$ - see front matter � 2004 Published by Elsevier Inc.

doi:10.1016/j.micromeso.2004.11.002

* Corresponding author. Tel.: +41 865 62708; fax: +41 865 63810.

E-mail address: kaliagui@gch.ulaval.ca (S. Kaliaguine).

num chloride on silica gel for applications in cracking,

isomerization reactions, dehydrochlorination and hyd-

rodechlorination [2–5]. Another study used the same

catalyst for work on the accelerating effect of ultrasonic

vibration on the reaction of benzene alkylation with

cyclohexene [6]. Clark et al. proposed an alternative

for the cationic polymerization of styrene [7,8]. They

used the Lewis acid catalyst supported on a porous silicasupport. The immobilized aluminum chloride on meso-

porous MCM-41 silica is a substitute catalyst for

liquid-phase isopropylation of naphthalene [9]. Re-

cently, several researchers have been using aluminum

chloride grafted or impregnated mesoporous support

for different applications [10,11]. Enhanced selectivity

in the preparation of linear alkylbenzenes using HMS

silica supported aluminum chloride was presented byClark et al. [12]. Another group investigated the pore

size engineering on MCM-41 over the monoalkylation

selectivity [13]. Aluminum chloride grafted on SBA-15

and MCM-41 silica supports were compared for their

138 D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144

catalytic activities for benzene alkylation [14]. This study

deals with the characterization of the aluminum chloride

species grafted on different MMSs and their catalytic

activity in the alkylation of benzene with 1-dodecene.

2. Experimental

2.1. Preparation of MMSs

The designation of the catalysts reflects three charac-

teristics of the material. Letters, H or WH, are attrib-

uted to the structure type of MMSs (H = hexagonal

and WH = wormhole like). The number after the typeof structure gives average pore diameter of the MMS ex-

pressed in Angstroms and the final term indicates the

molecule grafted on the MMS. For example, WH26-

AlCl3 is an aluminum chloride grafted on a wormhole

like MMS with a pore diameter of 26A.

2.1.1. Sample WH26

WH26 was obtained by following the synthesis proce-dure reported by Tanev and Pinnavaia [15]. Dodecyl-

amine was used as the template in a solution of

ethanol:water and tetraethyl orthosilicate was the silicon

source (I0/S0 = 4). Dodecylamine (1.8g) was added to a

solution of distilled water (90ml) and ethanol (34 ml).

Once the template was dissolved, tetraethyl orthosilicate

(8.2g) was added under vigorous stirring. The reaction

mixture was aged at ambient temperature for 20h. Theproduct was filtered and washed with three portions of

distilled water (300ml) and then calcined at 600 �C for

6h.

2.1.2. Sample WH50

WH50 was obtained following the synthesis proce-

dure reported by Stucky and coworkers [16]. Poly(alk-

ene oxide) block copolymer (Pluronic P-123) was usedas the template in ethanol and silicon tetrachloride

was the silicon source. Pluronic P-123 (2.5g) was added

in 25ml of ethanol. Once the template was dissolved, sil-

icon tetrachloride (4.2g) was added under vigorous stir-

ring for 0.5h. The resulting product was gelled in a Petri

dish at 60 �C for 3 days. WH50 was calcined at 600 �Cfor 6h.

2.1.3. H33 and H55

H33 and H55 were obtained following the synthesis

procedure reported by Trong On et al. [17] Poly(alkene

oxide) block copolymers, BRIJ-56 and P-123, were used

as template in distilled water for H33 and H55 respec-

tively. The silicon source used was a sodium silicate

solution (28.7% SiO2, 8.9% Na2O). The templates

(16g) were dissolved in 380g of distilled water at 40 �Cfor 2h. When the mixture was clear, 37g of sodium sil-

icate solution was added at room temperature with mag-

netic stirring for 1h. To this solution, 48g of sulfuric

acid 98% was added with vigorous stirring for 18h

and heated for 48h at 80 �C. H33 and H55 were filteredand washed with three portions of distilled water. The

H33 and H55 were calcined at 600 �C for 6h.

2.2. Grafting procedure (Schlenk technique)

The support (2.5g) was dried at 400 �C under vacuum(overnight) and kept under dry argon in a Schlenk tube.

The support was added to 100ml of benzene (sodium

dried) in the reaction tube under a dry argon atmo-

sphere. To this solution was added 1.5g of aluminum

chloride stored in a Schlenk tube. The mixture was re-fluxed for 3h under dry argon and then filtered and

washed three times in a Schlenk tube with dry benzene

(100ml). The grafted support was stored in a Schlenk

tube under a dry argon atmosphere.

2.3. Characterization

Powder X-ray diffraction (XRD) patterns were re-corded using a SIEMENS D5000 or a PHILIPS

diffractometer using CuKa radiation (k = 1.5406A).Diffraction patterns were recorded with a step scan of

0.02� for 2h between 1� and 10�.Nitrogen adsorption/desorption isotherms were mea-

sured at liquid nitrogen temperature using a QUANTA-

CHROME NOVA 2000 instrument. Before adsorption,

the samples were evacuated at 200 �C for 4h. Specificsurface area was calculated using the multi-point BET

method in the relative pressure range of 0.05–0.3. The

pore diameter was estimated from the peak position of

BJH pore size distribution. Sample compositions were

measured by atomic adsorption using a PERKIN EL-

MER model 1100B after dissolution of the catalyst in

a diluted mixture of HCl and HF at 60 �C.Adsorbed pyridine infrared spectra (IR-pyridine)

were recorded on a BIO-RAD FTS-60 FTIR spectrom-

eter. The wafers were prepared in a glove box under a

dry argon atmosphere. The self-supporting wafers

were evacuated in situ in an IR cell at 250 �C overnight.

Pyridine was permitted to desorb at 25 �C and 120 �C.The spectra were recorded after cooling at room tem-

perature.

The ammonia desorption experiment (TPD-NH3)was performed on a RXM 100 multicatalyst testing

and characterization system (Advanced Scientific De-

sign Inc.). The catalyst was prepared in a glove box

under dry argon atmosphere. For TPD-NH3, 15mg of

catalyst was treated at 25–300 �C (ramp = 10K/min)

and 1h isotherm at 300 �C under a He flow of 20ml/

min. The catalyst was cooled down to room temperature

under flowing helium. The catalyst was saturated withammonia at 120 �C under a flow of He (15ml/min) and

NH3 (5ml/min) for 5min. The set up was stabilized for

Table 1

Physico-chemical properties of the catalysts and their silica supports

Catalyst SBET (m2/g) Pore diameter (A) [Al] (mmol Al/g)

H33 1018 33

H33-AlCl3 654 30 2.00

H55 803 55

H55-AlCl3 548 53 2.62

WH26 874 26

WH26-AlCl3 675 23 2.30

WH50 560 50

WH50-AlCl3 344 37 1.90

0

2000

4000

6000

8000

10000

12000

14000

1 1.5 2 2.5 3

2 Theta

Inte

nsity

(a.u

.)

(a)

(b)

Fig. 1. X-ray diffraction pattern of: (a) H33 and (b) H33-AlCl3.

8000

D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144 139

3h at 120 �C under a flow of He (20ml/min). The cata-

lyst was then cooled down to room temperature. The

conditions of the desorption tests were: He (20ml/min)

temperature from 25 �C to 450 �C (ramp = 10K/min).

The catalyst was maintained for 5min at 450 �C after

the end of the ramp in order to achieve desorption. Athermal conductivity detector (TCD) was used for quan-

tification and masses of 16, 17, 28, 32, 36 were recorded

on a mass spectrometer to detect possible leaks of other

compounds. The amount of ammonia desorbed at dif-

ferent temperatures was calculated by peak integration.27Al and 29Si MAS NMR spectra were obtained at

room temperature on a BRUKER AVANCE 300

MHz. 27Al MAS NMR were recorded at a frequencyof 78.2MHz and were spun at 5KHz. 29Si were recorded

at 59.6MHz and were spun at 5KHz. The rotors were

filled in a glove box under a dry argon atmosphere.

The surface OH densities were obtained by reacting

the surface OH�s with a solution of trimethylaluminum(2mol/l) which produces methane at room temperature.

The surface OH density is calculated from the volume of

methane produced by this reaction. Samples were trea-ted at 300 �C under vacuum before the reaction.

2.4. Benzene alkylation by 1-dodecene

To a mixture of 0.5g of heterogeneous catalyst in dry

benzene (5ml or 25ml), 5ml of 1-dodecene was added.

Samples were taken after 15min and 60min of the final

addition of 1-dodecene. The products were analyzed byGC–MS equipped with a capillary DB5-MS column

(30m length, 0.25mm i.d.).

2.5. Recycle studies of the H33-AlCl3 catalyst

The initial alkylation test was carried out as previ-

ously described. Upon completion of the reaction, the

reaction mixture was filtered and washed with dry ben-zene. The reaction tube was recharged with dry benzene,

the used catalyst and fresh 1-dodecene. The procedure

was repeated several times. The products were analyzed

by GC–MS equipped with a capillary DB5-MS column

(30m length, 0.25mm i.d.).

0

1000

2000

3000

4000

5000

6000

7000

1 4

2 Theta

Inte

nsity

(a.u

.)

(a)

(b)

5 62 3

Fig. 2. X-ray diffraction pattern of: (a) WH26 and (b) WH26-AlCl3.

3. Results

3.1. Characterization

Four catalysts were synthesized and analyzed. Two of

them were based on grafting aluminum chloride onto

hexagonal mesoporous silica with different pore sizes,

33A and 55A. The other two were based on grafting

aluminum chloride on wormhole like mesoporous silicawith different pore size, 26A and 50A. The textural and

compositional properties of the catalysts are presented

in Table 1. The XRD data show that the mesoporous

structure is preserved after the addition of aluminum

chloride (Figs. 1 and 2). The specific area and the pore

diameter of each catalyst decrease. Those changes are

due to the surface modification by the grafting reaction

onto the surface of each mesoporous support.

The acidity of the catalysts was characterized using

FTIR spectra of adsorbed pyridine (pyridine-FTIR).

Fig. 3. Pyridine FTIR spectra (after desorption at 120�C) of: (a) H33,(b) H33-AlCl3, (c) H55-AlCl3, (d) WH26-AlCl3 and (e) WH50-AlCl3.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

2400

0 100 200 300 400Temperature (oC)

Inte

nsity

(a.u

.)

(a)(b)

(c)

(d)

(e)

Fig. 4. Ammonia TPD of: (a) H33-AlCl3 exposed to ambient air

moisture 5min, (b) H33-AlCl3, (c) H55-AlCl3, (d) WH26-AlCl3 and (e)

WH50-AlCl3.

Table 2

Amount of ammonia desorbed from the different grafted samples

Samples Amount of NH3 desorbed (lmol/g) Ratio NH3/Al

H33-AlCl3 2071 1.03

H33-AlCl3a 241 0.12

H55-AlCl3 2006 0.77

WH26-AlCl3 2287 0.99

WH50-AlCl3 1684 0.89

a Exposed 5min to ambient air moisture.

-80-302070120

ppm

inte

nsity

(a.u

.)

Fig. 5. 27Al MAS NMR spectra of H33-AlCl3.

140 D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144

Spectra of all catalysts show three bands (1456cm�1,

1496cm�1, 1545cm�1) attributed to pyridine adsorbed

on Bronsted and Lewis acids sites (Fig. 3). The band

at 1456cm�1 is attributed to Lewis acid sites, the band

at 1545cm�1 is attributed to Bronsted acid sites andthe band at 1496cm�1 is attributed to both Bronsted

and Lewis acid sites [18]. The mesoporous supports

did not display significant pyridine FTIR signals, indi-

cating that the MMSs have no acidic properties. For

every grafted aluminum chloride catalyst, the predomi-

nant acid species are Lewis acid sites with a minor

amount of Bronsted acid sites.

The density of acid sites was determined using ammo-nia thermo-desorption (ammonia TPD). The incorpora-

tion of aluminum chloride by a grafting reaction on

both MMSs creates two types of acid species. Fig. 4

shows the ammonia TPD of aluminum chloride grafted

on MMSs. Desorption curves present two important

peaks at 200 �C, 250 �C and a very small peak around

400 �C. Peaks at 200 �C and 250 �C are attributed to

Lewis acid type [19]. The smallest peak around 400 �Cis usually attributed to Bronsted acid type [20]. When

catalysts are exposed to ambient air moisture, they pres-

ent no important signal in ammonia TPD. Table 2

shows the amount of ammonia desorbed from the differ-

ent grafted samples. The ammonia desorbed drops

dramatically when the catalyst is exposed to moisture.

The ammonia desorbed went from 2071lmol NH3/g

to 241lmol NH3/g after air moisture exposition. Themolar ratio of ammonia desorbed per aluminum grafted

(NH3/Al) is 1.03 in this case. This indicates that each

chemisorbed ammonia occupies one acid site of the cat-

alyst. For both types of mesostructured supports the

large pore materials (H55-AlCl3 and WH50-AlCl3) have

lower NH3/Al ratio than the small pore materials (H33-

AlCl3 and WH26-AlCl3).

Figs. 5–8 show the 27Al MAS NMR spectra of alumi-

num chloride grafted MMSs and the deconvolution

curves. According to a study of Sato and Maciel, the

peak at 0ppm is attributed to six-coordinate Al species

[21]. This peak appears from air contamination whenthe catalysts were placed in an NMR cell. This cell

was manipulated under ambient air and it does not pro-

-80-302070120ppm

inte

nsity

(a.u

.)

Fig. 6. 27Al MAS NMR spectra of H55-AlCl3.

-80-302070120

ppm

inte

nsity

(a.u

.)

Fig. 7. 27AL MAS NMR spectra of WH26-AlCl3.

-80-302070120ppm

inte

nsity

(a.u

.)

Fig. 8. 27Al MAS NMR spectra of WH50-AlCl3.

D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144 141

tect entirely the catalyst from moisture. The peak at

36ppm is attributed to an aluminum chloride grafted

on a silanol group and having two neighbour silanol

groups. Those neighbours create the Bronsted acid sites.

The peak at 65ppm is assigned to an aluminum chloride

grafted on a silanol group with only one neighbour sila-

nol group. The peak at 75ppm is assigned to an alumi-

num chloride grafted on three silanol groups. Thestrongest peak at 85ppm is assigned to an aluminum

chloride grafted on two silanol groups. Fig. 9 summa-

rizes the structures of the various types of aluminum

chloride grafted on the MMSs.

Si Si Si Si

O O O O

AlAl

OOH

OH

OHOH

HO

HO

0ppmSi Si Si

OHO OH

Al

ClCl

36ppm

Fig. 9. Structures of aluminum chloride grafted on M

Fig. 10 shows the 29Si MAS NMR spectra of H33

and H33-AlCl3 with the deconvolution curve of each

peak. The MMS H33 presented three peaks at 92, 100

and 108ppm, which correspond to Q2, Q3 and Q4 silica

species. Each silica species is associated with a gradualincrease in silicate tetrahedron cross-linking. For the

H33 silica support, the ratios of these silica species are

recognized to be Q4:Q3:Q2 = 64:32.5:3.5. After the graft-

ing step, Q3 decreases, Q4 increases and the ratios be-

come Q4:Q3:Q2 = 75.4:20.8:3.8. These results indicate

that the aluminum chloride is grafted on the surface of

MMS H33 as the increase in Q4 corresponds to the dis-

appearance of some of the support OH�s.

3.2. Alkylation results

The catalytic properties of the catalysts were evalu-

ated in terms of the conversion of 1-dodecene and the

selectivity to monoalkylated products reported in Table

3. Both heterogeneous catalysts present better mono-

alkylation selectivity than aluminum chloride employedas a homogeneous catalyst. For a benzene:1-dodecene

molar ratio of 12.5:1, the reaction using H33-AlCl3 pre-

sented the highest monoalkylation selectivity (89.9% at

100% conversion). The percentages of monoalkylation

for the other catalysts are 85.0% for H55-AlCl3, 84.5%

for WH26-AlCl3, WH50-AlCl3 and 74.5% for homoge-

neous AlCl3. For a benzene:1-dodecene molar ratio

of 2.5:1, reactions using H33-AlCl3 and H55-AlCl3

Si Si

O OH

Al

ClCl

65ppm

Si Si Si

OO O

Al

75ppm

Si Si

O O

Al

Cl

85ppm

MSs with the 27Al MAS NMR chemical shifts.

-130 -120 -110 -100 -90 -80

ppm

inte

nsity

(a.u

.)

(a)Q4:Q3:Q2 = 64:32.5:3.5

(b)Q4:Q3:Q2 = 75.4:20.8:3.8

Fig. 10. 29Si MAS NMR spectra of: (a) H33 and (b) H33-AlCl3.

142 D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144

presented the highest monoalkylation selectivity (81.5%

and 82.6%). The percentage of monoalkylation for the

other catalysts are 69.6% for WH26-AlCl3, 64.3% for

WH50-AlCl3 and 52.8% for AlCl3. All the results indi-

cated better monoalkylation selectivity for aluminum

chloride grafted on hexagonal MMSs than for alumi-num chloride grafted on wormhole like MMSs for the

same 1-dodecene conversion. Distribution of monoalky-

lated products are presented in Table 4. For benzene/

1-dodecene molar ratio, 12.5:1 and 2.5:1, the most

important monoalkylated product is the 1-methyl1-

undecylbenzene. The heterogeneous catalysts are about

5% more selective than the homogeneous catalyst.

Fig. 11 shows the molar ratio of monoalkylate to ini-tial benzene versus benzene conversion plots. When the

rates of the two successive reactions are represented as

second order reaction rates,

benzene!k1 monoalkylate!k2 dialkylate ð1Þthe data reported in Fig. 11 should yield one curve cor-

responding to a constant k2/k1 ratio for each catalyst.

Even though the data in Table 3 yield only two data

points for each of these curves, estimates for k2/k1 can

be obtained by fitting these data to the equation [22]:

Table 3

Comparison of catalyst performance in the alkylation of benzene at room te

Catalyst Ratio Bz:1-Dod. Reaction time (min) C

H33-AlCl3 12.5:1 15 1

H55-AlCl3 12.5:1 15 1

WH26-AlCl3 12.5:1 15 1

WH50-AlCl3 12.5:1 15 1

AlCl3 12.5:1 15 1

H33-AlCl3 2.5:1 60

H55-AlCl3 2.5:1 60

WH26-AlCl3 2.5:1 60

WH50-AlCl3 2.5:1 60

AlCl3 2.5:1 60

a Conversion of 1-Dodecene.

k2k1

� 1

� �CM

CBz0

¼ CBz

CBz0

1� CBz

CBz0

� �k2k1�1

" #ð2Þ

where CM is the monoalkylate concentration, CBzthe benzene concentration and CBz0, initial benzene

concentration.

The estimates of k2/k1 are reported in Table 3. The

two hexagonal catalysts H33-AlCl3 and H55-AlCl3 pres-

ent the slowest dialkylation compared to monoalkyla-

tion whereas WH26-AlCl3 and WH50-AlCl3 yielded

higher values of k2/k1, still lower however than AlCl3as homogeneous catalyst.The H33-AlCl3 was submitted to recycling tests. Re-

sults of these experiments are presented in Table 4.

The catalyst was reused five times without any loss in

activity. During all experiments selectivity stayed essen-

tially stable around 82% of monoalkylation and 18%

dialkylation for the same 1-dodecene conversion.

4. Discussion

The results presented above suggest that the presence

of aluminum chloride on the surface does not affect the

mesoporous structure of the support. There is little

change in the XRD pattern recorded with and with

out the aluminum chloride grafted on both MMSs silica

types. In addition, there is only a diminution of the spe-cific area attributed to the presence on the aluminum

chloride on the surface of MMS.

The acid properties (pyridine-FTIR and TPD-NH3)

obtained for the four aluminum chloride grafted on

MMSs are coherent with the results presented in the lit-

erature for similar compounds. The adsorbed pyridine

infrared spectra showed the presence of Lewis and Bron-

sted acid types on aluminum chloride grafted MMSs.The spectra of pyridine-FTIR of the MMSs supports

did not present such signals. The predominant acid spe-

cies are Lewis acid sites. That is confirmed by the ammo-

nia TPD because desorption curves show for each

mperature

onversion (mol%)a Alkylation k2/k1

Mono (%) Di (%)

00 89.9 10.1

00 85.0 15.0

00 84.5 15.5

00 80.0 20.0

00 74.5 25.5

99 81.5 18.5 1.05

96 82.6 17.4 1.03

97 69.6 30.4 1.97

99 64.3 35.7 2.17

99 52.8 47.2 4.12

Table 5

Recycle studies on the H33-AlCl3 catalyst

Reaction Ratio

Bz:1-Dod.

Time

(min)

Alkylationa

Mono (%) Di (%)

1 2.5:1 60 82.0 18.0

2 2.5:1 60 83.1 16.9

3 2.5:1 60 81.2 18.8

4 2.5:1 60 82.7 16.3

5 2.5:1 60 83.0 17.0

a 1-Dodecene conversion is 100% for all tests.

Table 6

Surface OH density of the catalysts and supports

Catalyst Loading (mmol OH/g)

H33a 2.33

H33-AlCl3 0.62

WH26a 2.15

WH26-AlCl3 0.25

a Pre-treated at 400�C.

Table 4

Monoalkylation selectivity of each catalyst

Catalyst Ratio Bz/Dod. Monoalkylation (%)

Methylundecyl Ethyldecyl Propylnonyl Butyloctyl Pentylheptyl

H33-AlCl3 12.5:1 51.1 19.1 11.5 10.2 8.1

H55-AlCl3 12.5:1 54.0 18.0 11.4 8.9 7.7

WH26-AlCl3 12.5:1 53.5 17.3 11.5 9.8 7.1

WH50-AlCl3 12.5:1 53.9 18.1 11.3 9.8 6.9

AlCl3 12.5:1 46.9 16.8 13.0 12.5 10.8

H33-AlCl3 2.5:1 51.1 18.9 11.7 10.0 8.3

H55-AlCl3 2.5:1 53.7 18.3 11.1 9.3 7.6

WH26-AlCl3 2.5:1 53.2 17.6 11.6 9.7 7.1

WH50-AlCl3 2.5:1 53.5 18.5 11.3 9.6 7.1

AlCl3 2.5:1 47.1 16.6 13.2 12.3 10.8

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 5 10 15 20 25 30 35 40 45 50

Benzene conversion (%)

[mon

oalk

ylat

e]/[b

ezen

e 0]

(a)(b)

(c)(d)

(e) k2/k1=4.12

k2/k1=1.05

Fig. 11. Monoalkylate/benzene0 vs. benzene conversion plots using:

(a) H55-AlCl3, (b) H33-AlCl3, (c) WH26-AlCl3, (d) WH50-AlCl3 and

(e) AlCl3.

D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144 143

catalyst two peaks of Lewis acid site and only a minor

one for Bronsted acid sites. The peak around 200 �C is

attributed to aluminum chloride grafted on two silanol

groups onto the surface of MMS. The second peak at

250 �C is attributed to aluminum chloride grafted onone silanol group. This explanation is in good correla-

tion with the 27Al MAS NMR spectra. The most impor-

tant Lewis acid type present on the surface is the

aluminum chloride grafted on two silanol groups

(85ppm). The low Bronsted acidity of the catalyst is ex-

plained by the strictly moisture free conditions of the

tests. The presence of any moisture in the experimental

operation is catastrophic for the catalyst. A dramaticloss in activity was indeed observed upon contacting

the catalyst with ambient air, which yielded 90% reduc-

tion of the Lewis acid sites of the catalyst.

The activity in benzene alkylation of aluminum chlo-

ride grafted on MMSs is comparable to homogeneous

aluminum chloride but grafting results in better mono-

alkyl benzene selectivities. Catalysts supported onto

hexagonal mesoporous silica supports presented bettermonoalkylation selectivity than those supported onto

wormhole like mesoporous silica supports. This fact is

illustrated by the rate constant ratio k2/k1. Aluminum

chloride grafted onto hexagonal mesostructured support

(H33-AlCl3) presented a rate constant ratio k2/k1 of 1.05

(1.03 for H55-AlCl3), suggesting that the secondary

alkylation is relatively slower on the hexagonal support

than on the wormhole like support. This behaviour

(lower k2/k1) is believed to be associated with a morehighly polar MMS support which imposes a faster out-

ward diffusion of the monoalkyl product from the mes-

opore surface, thus limiting the secondary alkylation.

Table 5 presents measurements of the OH concentration

on the surface of catalysts. The results in Table 6 con-

firm that hexagonal MMS supported catalysts have a

higher surface density of OH groups than the ones sup-

ported on wormhole like MMS. These surfaces OH havea repulsive interaction with the monoalkylated benzene

which has a saturated alkyl chain. This interaction accel-

erates the outward diffusion of the monoalkyl product

144 D. Dube et al. / Microporous and Mesoporous Materials 79 (2005) 137–144

thus lowering its concentration in the mesopore environ-

ment of the Lewis acid active sites. Each catalyst can be

recycled as long as the moisture free conditions are

maintained. Catalysts have to be filtered and refilled

with fresh reagents in order to obtain a maximum

activity.

5. Conclusion

Aluminum chloride grafted on MMSs catalysts in

moisture free conditions show essentially pure Lewis

acid sites. Their catalytic activity in benzene alkylation

by 1-dodecene is comparable to that presented by alumi-num chloride under homogeneous catalysis conditions.

The monoalkylation selectivity is significantly enhanced

when aluminum chloride is grafted onto MMSs. Alumi-

num chloride supported onto hexagonal MMSs proved

to have the highest selectivity for the benzene alkylation.

The lower constant rate ratio k2/k1 is believed to be asso-

ciated to the highly polar MMS support which allows a

fast outward diffusion of the monoalkyl product, thuslimiting the secondary alkylation. AlCl3 grafted cata-

lysts have been recycled up to five times without any

change in catalytic activity and selectivity.

Acknowledgments

We thank NSERC for financial support. The authorsare grateful to M.G. Lemay for assistance in the exper-

imental part. We thank Dr. D. Desplantier-Giscard for

valuable discussions and M.S. Pelletier for some analyt-

ical support.

References

[1] D. Trong On, D. Desplantier-Giscard, C. Danumah, S. Kalia-

guine, Appl. Catal. A: Gen. 222 (2001) 299.

[2] R.S. Drago, S.C. Petrosius, P.H. Kaufman, J. Mol. Catal. 89

(1994) 317.

[3] R.S. Drago, S.C. Petrosius, C.W. Chronister, Inorg. Chem. 33

(1994) 367.

[4] E.E. Getty, R.S. Drago, Inorg. Chem. 29 (1990) 1186.

[5] E.E. Getty, S.C. Petrosius, R.S. Drago, J. Mol. Catal. 67 (1991)

127.

[6] S. Sato, F. Nozaki, S.J. Zhang, P. Cheng, Appl. Catal. A: Gen.

143 (1996) 271.

[7] J.K. Shorrock, J.H. Clark, K. Wilson, J. Chisem, Org. Process

Res. Dev. 5 (2001) 249.

[8] V. Sage, J.H. Clark, D.J. Macquarrie, J. Mol. Catal. A: Chem.

198 (2003) 349.

[9] X.S. Zhao, M.G.Q. Lu, C. Song, J. Mol. Catal. A: Chem 191

(2003) 67.

[10] M. Xu, A. Arnold, A. Buchholz, W. Wang, M. Hunger, J. Phys.

Chem. B 106 (2002) 12140.

[11] S. Jun, R. Ryoo, J. Catal. 195 (2000) 237.

[12] P.M. Price, J.H. Clark, K. Martin, D.J. Macquarrie, T.W.

Bastock, Org. Process Res. Dev. 2 (1998) 221–224.

[13] X. Hu, M.L. Foo, G.K. Chuah, S. Jaenicke, J. Catal. 195 (2000)

412.

[14] K.K. Kang, H.K. Rhee, Studies in Surface Sciences and Catalysis

146 (2003) 673.

[15] P.T. Tanev, T.J. Pinnavaia, Sciences 267 (1995) 865.

[16] P. Yang, D. Zhao, D.I. Margolese, B.F. Chmelka, G.D. Stucky,

Nature 396 (1998) 152.

[17] D. Trong On, S. Kaliaguine, P. Fournier, L. Fortier, World Patent

WO03037511A1 (2003).

[18] A. Ghanbari-Siahkali, A. Philippou, A. Garforth, C.S. Cundy,

M.W. Anderson, J. Dwyer, J. Chem. Mater. 11 (2001) 569.

[19] S.R. Zhai, W. Wei, D. Wu, Y.H. Sun, Catal. Lett. 89 (2003) 261.

[20] S.E. Dapurkar, P. Selvam, Appl. Catal. A: Gen. 254 (2003) 239.

[21] S. Sato, G.E. Maciel, J. Mol. Catal. A. Chem. 101 (1995) 153.

[22] S. Kaliaguine, Note book: GCH-10163 Calcul des reacteurs,

Universite Laval.