Friedel-Crafts catalysis using sol-gel derived supported reagents: Metal diketonate modified mesoporous aluminosilicates

Article abstract of DOI:10.1039/a707137j

Aluminosilicates have been prepared from aluminium tri-sec-butoxide and tetraethyl orthosilicate in butanol incorporating transition metal pentane-2,4-dionate complexes, which on calcination give active Friedel-Crafts alkylation catalysts having textural

Full text of DOI:10.1039/a707137j

View Article Online / Journal Homepage / Table of Contents for this issue
Friedel–Crafts catalysis using sol–gel derived supported reagents
Metal diketonate modiÐed mesoporous aluminosilicates
Jack M. Miller,*¤ David Wails, J. Stephen Hartman and Jennifer L. Belelie
Department of Chemistry, Brock University, St. Catharines, Ontario, Canada L 2S 3A1
Aluminosilicates have been prepared from aluminium tri-sec-butoxide and tetraethyl orthosilicate in butanol incorporating tran-
sition metal pentane-2,4-dionate complexes, which on calcination give active FriedelÈCrafts alkylation catalysts having textural
mesoporosity. Addition of b-diketonates to the aluminosilicate solÈgel results in complexation to aluminium which, on calcination,
generates a mesoporous aluminosilicate with increased surface area and pore volume. This provides a suitable framework upon
which zinc chloride may be supported at higher loadings than for either the unmodiÐed material or K10, the support for the
commercial catalyst Clayzic, to give materials with signiÐcantly higher catalytic activity in a FriedelÈCrafts alkylation reaction.
FriedelÈCrafts catalysis traditionally uses Lewis acid catalysts
such as aluminium trichloride or Bronsted acids such as
hydrolysis rates, and in a zirconium alkoxide system prevents
aggregation of nanosized particles.16 We here describe the
modiÐcation of our previously reported aluminosilicate
system13 with b-diketones and metal b-diketonates to gener-
ate improved mesoporous aluminosilicate based FriedelÈ
Crafts alkylation catalysts.
Ž
hydroÑuoric acid, which have environmental drawbacks in
addition to often poor selectivity and lack of reusability.1 In
recent years, alternative catalysts such as Clayzic2 (zinc chlo-
ride supported on montmorillonite K10) have been introduced
as useful alternatives. Since Clayzic was Ðrst reported in
1
989,3 it has been extensively studied.4h11 The activity of the
Experimental
Instrumentation
catalyst can be further enhanced by acid treatment of the
support, which destroys much of the original framework
structure generating mesoporous materials. The pore size of
the mesopores is critical, with the optimum diameter being
NMR experiments were carried out on a Bruker DPX-300
instrument using a standard bore magic angle spinning (MAS)
probe with 4 mm zirconia rotors. 29Si cross-polarization
magic-angle spinning (CP MAS) and 27Al MAS spectra were
obtained under the conditions reported in our earlier study.13
8
È12 nm for maximum catalytic activity of the resulting sup-
ported reagent.10,11 Other supports such as mesoporous silica
are also more e†ective than K10, and the pore size character-
istics appear to be the most signiÐcant factor, rather than the
actual chemical nature of the support.10,11 These studies
however have been limited to commercially, available
1
3C CP MAS spectra were obtained at 75.48 MHz using
cross-polarization with proton decoupling during acquisition
and referenced to external tetramethylsilane (TMS). Samples
were spun at 4 kHz in air. A 10 ms cross-polarization time
was used with a 4 s delay between pulses, and up to 2048
free-induction decays (FIDs) were typically accumulated over
a spectral window of 28 000 Hz.
materials of limited pore size distributions. AlCl supported
on 7 nm diameter mesoporous silica has recently been shown
3
to be highly e†ective as a FriedelÈCrafts alkylation catalyst.12
We have taken the approach of solÈgel synthesis of the
support and have recently described the synthesis of FriedelÈ
Crafts catalysts based on a solÈgel derived aluminosilicate.13
SolÈgel methodologies o†er an attractive approach to the
synthesis of materials for a number of applications including
ceramics, catalysts and optical devices.14 The low temperature
solÈgel route o†ers a number of advantages over the tradi-
tional high temperature melt processes. Both the chemical
composition of the materials and physical properties such as
surface area and pore size distributions can often be precisely
controlled. The solÈgel process involves the controlled hydro-
lysis of suitable precursor molecules, often metal alkoxides in
a (typically alcoholic) solvent, often with acid or base cataly-
sis. During hydrolysis, partially hydrolyzed metal alkoxides
polymerize to give a sol (a colloidal suspension of solid par-
ticles in the liquid phase). Further polycondensation results in
gelation. The Ðnal stages in the process involve ageing the gel,
during which further condensation reactions may occur, and
drying of the gel to give the Ðnal product. Pentane-2,4-dione
Surface area and pore volume determinations were carried
out at 77 K using a Coulter SA3100 instrument with an
automated gas volumetric method employing nitrogen as the
adsorbate. Surface areas were determined using the BET
adsorption method17 and pore size distributions using the
Barrett, Joyner and Halenda (BJH) method on the desorption
branch of the isotherm.18 For materials having type II
hysterisis loops, BJH calculations were made on the adsorp-
tion branch of the isotherm which is generally believed to give
a more accurate representation of the catalyst pore distribu-
tion in this system. Samples were outgassed under high
vacuum at 200 ¡C for 1 h immediately prior to analysis.
Scanning electron microscopy (SEM) analysis was carried
out on gold coated samples with a Hitachi SEM S2400 with
an electron gun energy of 8 keV.
Simultaneous di†erential scanning calorimetry (DSC) and
thermogravimetry (TG) was carried out using a Thermal Sci-
ences STA625. Approximately 10È20 mg of sample were
heated from room temperature to 600 ¡C at 10 ¡C min~1
under Ñowing nitrogen.
Mass spectrometry (MS) was carried out on a Kratos
Concept IS mass spectrometer. Samples were heated in silica
tubes to approximately 650 ¡C at a heating rate of 2 ¡C s~1.
Evolved materials were analyzed under electron impact condi-
tions.
(
Hacac) has been used previously to alter the properties of
solÈgel derived materials. Addition of Hacac to a tetramethyl-
orthosilicate based system prevented formation of crystalline
precipitates, leading to increased transparency of the resulting
gels.15 Complexation of Hacac to metal alkoxides slows
¤
E-mail:jmiller=sandcastle.cosc.brocku.ca.
J. Chem. Soc., Faraday T rans., 1998, 94(6), 789È795
789

View Article Online
Table 1 Comparison of physical properties and catalytic activity of various materials
optimum
loading
surface area
pore volume
catalysta
/m2 g~1
/cm3 g~1
/mmol g~1
conversion(%)b
Zn(acac)2
treated
pentanedione treated
untreated
392.89
462.42
425.16
226.29
1.00
1.45
1.19
0.30
È
7.5
5.0
1.0
28
68
54
7
K10
a Materials are mesoporous aluminosilicates: optimal loadings are shown in Table 2. b Conversion(%) is deÐned in the Experimental section.
Reagents
increases the surface area and pore volume of the support
compared to the untreated material, resulting in an increase in
the optimum zinc chloride loading from 5.0 to 7.5 mmol g~1
with a corresponding increase in catalytic activity. The
optimum loading of zinc chloride is typically that which
results in a monolayer surface coverage and Ðlled meso-
pores.10,11 Further addition of zinc chloride results in a
system in which the zinc chloride behaves as the crystalline
material, which has an inherently low activity in FriedelÈ
Crafts catalysis.10,11 Increasing the surface area and pore
volume therefore increases the maximum loading achievable.
Benzene and ethanol were obtained from Caledon Labor-
atories. Metal acetylacetonate complexes excluding Zn(acac)
2
(
Aldrich) were obtained from Gelest Inc., Tullytown, PA. All
other chemicals were obtained from Aldrich and used as
received.
Catalyst preparation
Aluminum tri-sec-butoxide (ATSB) (20.0 g, 0.08 mol) was
mixed with butanol (200 ml) with stirring. Tetraethyl orthosili-
cate (TEOS) (17.4 ml, 0.08 mol) was then added. The mixture
was stirred vigorously and heated to 70 ¡C for approximately
Incorporation of Zn(acac) reduces the surface area and pore
2
volume slightly, however it is likely that some pores are Ðlled
5
min until a clear solution was obtained. For studies in which
an experimental variable was to be altered [e.g. amount of
Zn(acac) ], this solution was divided to give Ðve identical
samples.2
For materials modiÐed with metal complexes
[
typically 1.5 g Zn(acac) , 5.6 mmol or molar equivalent], the
2
complex was added to the hot solution and stirred to dissolve.
For ketone and diketone additives, (typically 1.0 ml Hacac, 9.8
mmol or molar equivalent) the solution was allowed to cool
to room temperature before addition. The solution was stirred
for 15 min, water added (11 ml), stirred for an additional 15
min and allowed to stand overnight. Solvent was then
removed by heating in an oven at 150 ¡C. Materials were then
activated at 500 ¡C in air to remove residual organics and
decompose metal complexes. Elemental analysis of the
Zn(acac) modiÐed material gave 20.89% Al, 22.79% Si, corre-
2
sponding to an Al : Si ratio of 0.98 : 1, compared with the
0
.93 : 1.00 ratio obtained for the unmodiÐed material.13
Fig. 1 27Al MAS NMR spectra of catalysts. a: Untreated; b:
For supports which were loaded after preparation, loading
Zn(acac) treated; c: Hacac treated. An asterisk denotes a spinning
2
side band.
was carried out as previously described from a methanolic
solution of zinc chloride.13 Catalyst testing was carried out
using our previously reported method, using between 0.25 and
0.75 g of catalyst, depending on the series of materials under
investigation.13 (i.e. the alkylation of benzene with benzyl
chloride at room temperature for 15 min. The percentage con-
version is deÐned as the total conversion of benzyl chloride to
diphenylmethane and the isomers of dibenzylbenzene).
Results and Discussion
Aluminosilicates have been prepared by a number of solÈgel
based routes.13,19h21 ATSB and TEOS were heated in a 1 : 1
molar ratio in butanol to give a clear solution, as previously
described.13 The system was modiÐed by the addition of either
a metal diketonate, ketone or diketone before addition of
water for hydrolysis. Drying and activation at 500 ¡C resulted
in complete removal of residual organics as evidenced by 13C
CP MAS NMR and di†use reÑectance IR (DRIFT) spectros-
copy. The materials, like the unmodiÐed analogues were
amorphous by powder X-ray crystallography. Table 1 details
the e†ect of incorporation of Hacac or Zn(acac) on the
2
aluminosilicate support, and compares the properties to the
Fig. 2 29Si CP MAS NMR spectra of catalysts. a: Untreated; b:
untreated material and Clayzic. Incorporation of Hacac
Zn(acac) treated; c: Hacac treated.
2
790
J. Chem. Soc., Faraday T rans., 1998, V ol. 94

View Article Online
with zinc oxide upon activation at 500 ¡C. This supported zinc
oxide also shows high catalytic activity in the model FriedelÈ
Crafts reaction. A number of zinc salts including halides,
nitrates, phosphates and acetates have been supported on K10
and studied as potential FriedelÈCrafts alkylation catalysts,
although supported zinc oxide has not previously been investi-
gated.3 Thermal decomposition of soluble metal complexes
giving the metal oxide provides a viable route to the synthesis
of supported metal oxides.22
NMR studies
27Al MAS NMR spectra and 29Si CP MAS NMR spectra of
the materials are shown in Fig. 1 and 2. The 27Al spectra
show three environments,23 corresponding to tetrahedral
framework (ca. 55 ppm), octahedral (ca. 6 ppm) and a third
intermediate site (ca. 25 ppm) which is usually interpreted as a
Ðve-coordinate site, although recent studies suggest that this
site may be better attributed to a distorted tetrahedral site.24
The unmodiÐed material shows the greatest level of octa-
hedral, non-framework sites. Hacac is known to complex to a
number of metal alkoxides and substitution reactions, i.e.,
Fig. 3 13C CP MAS NMR spectrum of Hacac modiÐed material. a:
Hacac peaks; b: butanol peaks; c: Al(acac) peaks.
3
elimination of Al(acac) .27 No evidence of transmetalation in
3
the Zn(acac) modiÐed system [leading to formation of
2
Al(acac) ] was observed by 13C CP MAS NMR or MS.
3
M(OR) ] xHacac ] M(OR) (acac) ] xROH
3
3~x
x
SEM/Nitrogen adsorption studies
are well documented. Complexation reduces hydrolysis rates
and has been used to prevent premature precipitation of one
component in a multicomponent solÈgel system.25
Since metal alkoxides hydrolyze more rapidly than silicon
alkoxides, acac complexation to aluminium reduces the
hydrolysis rate increasing the number of AlÈOÈSi species and
therefore increasing the framework aluminium content. The
Addition of Hacac or Zn(acac) to the aluminosilicate system
2
has a signiÐcant e†ect on the textural properties of the
materials, as shown by SEM and nitrogen adsorptionÈ
desorption studies. The SEM of the unmodiÐed material [Fig.
5(a)] has a highly open, amorphous texture, which alters sig-
niÐcantly upon addition of Hacac or Zn(acac) . In the
2
Zn(acac) material [Fig. 5(b)], agglomerates of ZnO are visible
2
Zn(acac) modiÐed material shows a larger four-coordinate Al
2
peak than the other materials, consistent with the spectra
on the surface of the particle. In the Hacac modiÐed material,
obtained for ZnCl modiÐed materials.13 This suggests a pos-
these agglomerates [presumably Al O from
decomposition] are even more prominent [Fig. 5(c)].
Nitrogen adsorptionÈdesorption isotherms for the three
Al(acac)3
2
2
3
sible role for Zn in the formation of four- over Ðve-coordinate
environments. It is also possible that the anion has some role
since chloride is known to catalyze alkyl silicate
polymerization26 and Hacac is known to complex to and alter
silicate based polymerization species.14
The 29Si spectra show broad, unresolved peaks consistent
with a range of Si(OSi) (OH)
and Si(OAl) (OSi)
sites.23
n
4~n
n
4~n
The untreated material has the centre of the band at [95
ppm, compared to [91 ppm for the Zn(acac) modiÐed and
2
[
93 ppm for the Hacac modiÐed materials. A shift to higher
Ðeld is consistent with a decrease in the number of framework
Al sites, and supports the interpretation of the 27Al spectra
previously described.
The 13C CP MAS NMR spectra of the Hacac modiÐed
material dried at 150 ¡C shows a number of peaks in addition
to residual butanol and Hacac (Fig. 3) which correlate to
those from an authentic sample of Al(acac) .
3
Thermal studies
Thermal desorption mass spectrometry shows the presence of
two intense ions at m/z \ 225 and 324, corresponding to
Al(acac) ` and Al(acac) ` respectively. The decomposition of
2
3
the complex is also observed by thermal analysis (Fig. 4). The
unmodiÐed sample shows a broad endotherm up to 200 ¡C
corresponding to loss of residual solvent. Both the Hacac
Zn(acac) modiÐed materials show an additional broad endo-
2
therm in the 175È275 ¡C region which are not present in the
blank sample, corresponding to metal acetylacetonate decom-
position. The overall weight loss for these two samples is sig-
niÐcantly higher than that of the untreated material.
Carboxylate alumoxanes, [Al(O) (OH) (O CR) ] , have been
doped with metal acetylacetonate complexes, and trans-
metalation observed, leading to a metal doped alumoxane and
Fig. 4 Thermal analysis of materials. a: Untreated; b: Hacac
x
y
2
z n
treated; c: Zn(acac) treated. The line descending continuously from
2
left to right is the weight loss curve, and the other is the energy plot.
J. Chem. Soc., Faraday T rans., 1998, V ol. 94
791

View Article Online
system, the isotherm switches to type IV, indicative of a meso-
porous structure. The nature of the hysteresis associated with
the adsorptionÈdesorption branches of the isotherm suggests
that the mesopores are interparticulate voids rather than
framework pores.28 Pore size distributions calculated using
the BJH method of analysis are shown in Fig. 7. The
Zn(acac) modiÐed material shows a sharp pore size distribu-
2
tion centred around 8.42 nm, and the Hacac modiÐed material
a broader distribution centred around 13.93 nm.
In a zirconia based system modiÐed with Hacac, dynamic
quasi-elastic light scattering was used to show the modiÐ-
cation of the textural properties of the materials. Hacac addi-
tion was found to prevent aggregation of nanosized zirconia
particles. The e†ect of Hacac on nitrogen adsorption proper-
ties in a solÈgel system has not previously been reported,
although the mesoporous nature of a silica system has been
modiÐed using Pt(acac) .22
2
Varying sol-gel preparation conditions
Ageing. The e†ect of varying ageing treatments is illustrated
in Table 2 (entries 1È5) for the system modiÐed with Zn(acac)
2
(1.00 g per sample). After preparation, samples were aged
under varying thermal treatments before Ðnal treatment at
00 ¡C. Allowing the sample to dry completely at room tem-
5
Fig. 5 SEM analysis of catalysts. a: Untreated; b: Zn(acac) treated;
2
c: Hacac treated
materials are shown in Fig. 6. The type II adsorptionÈ
desorption isotherm of the unmodiÐed material indicates that
the material is either non-porous or macroporous.28 The
SEM analysis previously described suggests the latter. Upon
Fig. 7 BJH pore distribution plots for materials. (ÈÈÈ) Untreated;
(
È È È È) Zn(acac) treated; (É É É É É É) Hacac treated.
2
Table 2 E†ect of varying preparation conditions on physical
properties and catalytic activity of various materials
addition of Hacac or Zn(acac) into the aluminosilicate
2
surface
area
/m2 g~1
pore
volume
/cm3 g~1
preparation
condition
conversion
(%)
Zn(acac) modiÐed system
2
ageing (overnight/subsequently)
1
2
3
4
5
RT/RT
50 ¡C/50 ¡C
RT/50 ¡C
150 ¡C/150 ¡C
RT/150 ¡C
426.70
424.75
449.59
446.70
428.36
0.36
0.66
0.43
0.94
0.74
5
21
18
41
29
amount of Zn(acac)
2
6
7
8
9
1.0 g
425.16
386.20
341.71
227.87
1.19
0.99
0.90
0.69
36
25
26
48
1.5 g
2.0 g
5.0 g
amount of Hacac (ZnCl loaded at 7.5 mmol g~1)
2
1
1
1
1
0
1
2
3
0.5 ml
1.0 ml
1.5 ml
2.0 ml
578.10
462.42
392.30
335.04
1.37
1.45
1.79
1.82
24
46
73
74
Fig. 6 Nitrogen adsorptionÈdesorption isotherms for materials.
(
ÈÈÈ) Untreated; (É É É É É É) Zn(acac) treated; (È È È È) Hacac treated.
2
792
J. Chem. Soc., Faraday T rans., 1998, V ol. 94

View Article Online
Fig. 11 E†ect of amount of Hacac added on pore size distributions.
(
(
È]È) 0.5 ml Hacac; (È)È) 1.0 ml Hacac; (ÈLÈ) 1.5 ml Hacac;
È|È) 2.0 ml Hacac.
Fig. 8 Nitrogen adsorptionÈdesorption isotherms for room tem-
perature dried material (entry 1 in Table 2). (ÈKÈ) adsorption;
(
---)---) desorption.
diameters (Fig. 9), resulting in lower-levels of benzyl chloride
conversion. The most active catalyst with the largest pore
diameter was dried at 150 ¡C immediately after preparation. It
is likely that heating, particularly for a wet gel where precur-
sors are still partly in solution, increases the interaction
between acac and the solÈgel resulting in the greatest e†ect on
the system.
perature before activation gives a material with poor activity,
the nitrogen adsorption isotherm for which exhibits a type II
hysteresis loop (Fig. 8). This has been associated with “ink-
bottleÏ type pores, having narrow necks and wide bodies,
where the mechanism of condensation and evaporation is dif-
ferent, although this is now believed to be an over simpliÐed
model.28 Allowing the materials to age at lower temperatures
results in materials with lower pore volumes, and lower pore
Amount of additive. The e†ect of varying the amount of
Zn(acac) is illustrated in entries 6È9 of Table 2, and the e†ect
of varying Hacac in entries 10È13. For both systems, surface
2
areas initially increase compared to the untreated material,
then steadily decrease. For the Zn(acac) system, pore
2
volumes decrease, presumably as a result of Ðlling with ZnO,
although the pore diameter increases slightly and the distribu-
tion broadens somewhat (Fig. 10). No obvious trend in cata-
lytic performance is observed, possibly as a result of reduced
pore volume countering the e†ects of increasing the amount of
Zn2` in the system. In the Hacac system, pore volumes
increase with the amount of additive, although the relative
increase is much less at higher Hacac levels. Conversion of
benzyl chloride in the model reaction appears to correlate well
with total pore volume, rather than the pore diameter (Fig.
11). This contrasts to previous studies which suggested that
the pore diameter (8È10 nm) was the most signiÐcant factor,11
although the most active catalysts also had the greatest total
pore volumes. With increasing amounts of additive, the pore
diameter increases and broadens signiÐcantly such that at the
higher levels, no deÐnite maximum in pore diameter is
evident.
Fig. 9 E†ect of ageing on pore size distributions. (ÈKÈ) room tem-
perature; (È)È) 50 ¡C; (ÈLÈ) RT overnight, then 50 ¡C; (È|È)
Other metal–acac complexes. A number of metalÈacac com-
plexes were also studied for both their e†ect on the textural
properties of the materials and their catalytic ability (Table 3,
Fig. 12). Each was incorporated into the aluminosilicate
1
50 ¡C; (È]È) RT overnight, then 150 ¡C.
framework at the same molar ratio as Zn(acac) . With the
2
exception of Fe, V and Mn, all gave broad mesoporous dis-
tributions up to 25 nm. Generally, surface areas and pore
volumes are comparable to the Zn(acac) modiÐed alumino-
2
silicate, although in some cases pore volumes are signiÐcantly
greater. Interestingly, a number of the resulting supported
metal oxides show reasonable activity as FriedelÈCrafts alkyl-
ation catalysts. K10 supported copper and magnesium chlo-
rides have been used as FriedelÈCrafts alkylation catalysts9
and the corresponding oxides are also shown to be active in
the present study.
Other b-diketones and ketones
A number of other organic modiÐers in addition to Hacac
were investigated (Table 4). Catalytic activity after loading
with zinc chloride (5.0 mmol g~1) correlates well with the total
Fig. 10 E†ect of Zn(acac) loading on pore size distribution.
2
(
ÈKÈ) 1.0 g Zn(acac) ; (È)È) 1.5 g Zn(acac) ; (ÈLÈ) 2.0 g
2
2
Zn(acac) ; (È|È) 5.0 g Zn(acac) .
2
2
J. Chem. Soc., Faraday T rans., 1998, V ol. 94
793

View Article Online
Table 3 E†ect of varying metal acetylacetonate complex on physical
pore volume of the support, which in many cases is signiÐ-
cantly greater than that of Hacac, although surface areas are
generally on the same order. Only the b-diketones, which are
capable of complexing to transition metals through ketoÈenol
tautomerization, gave materials with type IV isotherms, the
corresponding BJH pore size distributions for which are
shown in Fig. 13. Other additives including ketones and non-
complexing diketones gave type II isotherms similar to those
of the unmodiÐed aluminosilicate. MS analysis of the
materials dried at 150 ¡C showed only residual solvent for the
materials exhibiting type II isotherms. Organic additives have
been used to alter the pore volumes and pore size distribution
of solÈgel derived materials either in conjunction with tetra-
alkylammonium tempates29 or as drying control chemical
additives (DCCAs).14 DCCAs such as dimethylformamide
which are less volatile than the solvent replace the solvent in
pores during the Ðnal stages of drying, and the di†erent
surface tension in the Ðnal drying stages controls pore dis-
tributions.30 It is possible that the ketones and non-
complexing diketones are acting as DCCAs, although
generally up to half of the solvent is replaced by the DCCA.14
In the present work only a small amount of additive
properties and catalytic activity of various materials
surface
area
/m2 g~1
pore
volume
/cm3 g~1
pore
diameter
/nma
metal
complex
conversion
(%)
1
2
3
4
5
6
7
8
9
0
Zn(acac)2
392.89
396.32
431.66
414.86
432.79
553.14
362.65
480.72
440.01
448.69
1.00
1.43
2.07
1.06
0.93
1.09
1.74
1.09
0.98
1.26
8.42
È
È
10.75
8.94
7.64
È
8.94
6.34
11.95
99
2
0
26
55
55
3
63
44
33
Fe(acac)3
V(acac)3
Cr(acac)
3
Mg(acac)
2
Ni(acac)2
Mn(acac)
2
Cu(acac)2
Co(acac)2
1
Al(acac)3
a Where pore diameters are not listed they covered a very wide range.
(approximately 5% by volume) produces a signiÐcant e†ect in
the resulting solÈgel. For the materials exhibiting type IV iso-
therms, MS indicated the presence of aluminium diketonate
(
L) complexes for 3-methylpentane-2,4-dione (AlL ` m/z
2
\
253,
AlL ` m/z \ 366), 2,2,6,6-tetramethylheptane-3,5-
3
dione (AlL ` m/z \ 393, AlL ` m/z \ 576) and dibenzoyl-
2
3
methane (AlL ` m/z \ 473, AlL ` m/z \ 696). For the
2
3
Fig. 13 E†ect of b-diketones on pore size distributions. (È]È)
Hacac; (È)È) 1,1,1,5,5,5-hexaÑuoropentane-2,4-dione; (È]È) 3-
Fig. 12 E†ect of metal acetylacetonate complexes on pore size dis-
tributions
methylpentane-2,4-dione;
(ÈÈÈ)2,2,6,6-tetramethyl-heptane-3,5-
dione; (È|È)dibenzoyl methane.
Table 4 E†ect of varying organic additive on physical properties and catalytic activity of various materials
surface
area
/m2 g~1
pore
volume
/cm3 g~1
pore
diameter
/nma
additive
conversion(%)
1
2
3
4
5
6
7
8
9
0
1
2
Hacac
526.63
340.29
521.90
480.42
457.91
476.63
561.53
429.83
471.77
443.28
438.33
442.72
1.09
0.83
1.76
2.22
2.26
0.97
1.25
1.54
2.39
2.21
1.61
1.67
8.58
7.64
26
11
80
75
78
15
42
89
83
87
65
79
1,1,1,5,5,5-hexaÑuoropentane-2,4-dione
3-methylpentane-2,4-dione
pentane-2,3-dione
hexane-2,5-dione
2,2,6,6-tetramethylheptane-3,5-dione
dibenzoylmethane
propanone
pentan-2-one
heptan-4-one
acetophenone
È
È
7.64
7.11
È
È
È
1
1
1
È
È
2-hydroxy-4,6-dimethylacetophenone
a Where pore diameters are not listed they covered a very wide range. ZnCl loadings were 7.5 mmol g~1.
2
794
J. Chem. Soc., Faraday T rans., 1998, V ol. 94

View Article Online
6
D. R. Brown, H. G. M. Edwards, D. W. Farwell and J. Massam,
J. Chem. Soc., Faraday T rans., 1996, 92, 1027.
1
,1,1,5,5,5-hexaÑuoropentane-2,4-dione modiÐed material, an
unidentiÐable intense peak (m/z \ 393) dominated the mass
spectrum. It is possible that drying at 150 ¡C has already
begun to decompose the aluminiumÈdione complex, since the
highly electronegative triÑuoromethyl group would be
expected to weaken the CwOwAl bond in the complex.
7
8
J. Massam and D. R. Brown, Catal. L ett., 1995, 35, 335.
A. Cornelius, P. Laszlo and S. Wang, T etrahedron L ett., 1993, 34,
3
849.
9
0
1
2
3
S. J. Barlow, J. H. Clark, M. R. Darby, A. P. Kybett, P. Landon
and K. Martin, J. Chem. Res., 1991, (S) 74.
C. N. Rhodes and D. R. Brown, J. Chem. Soc., Faraday T rans.,
1
C. N. Rhodes and D. R. Brown, J. Chem. Soc., Faraday T rans.,
1
J. H. Clark, K. Martin, A. J. Teasdale and S. J. Barlow, J. Chem.
Soc., Chem. Commun., 1995, 2037.
1
1
1
1
992, 88, 2269.
Conclusions
993, 89, 1387.
Metal acetylacetonate complexes can be heated within a
solÈgel derived aluminosilicate to generate supported metal
oxides with signiÐcantly altered textural porosity which have
catalytic activity in FriedelÈCrafts alkylation. b-Diketones
may also be incorporated into the aluminosilicate gel where
complexation to solÈgel derived aluminium species again
results in a metal diketonate complex, which generates a
mesoporous aluminosilicate framework with a higher pore
volume than the unmodiÐed material. This material provides
a suitable support onto which zinc chloride may be loaded at
higher levels than Clayzic (K10 supported zinc chloride) or the
unmodiÐed material resulting in signiÐcantly greater catalytic
activity.
J. M. Miller, D. Wails, J. S. Hartman and J. L. Belelie, J. Chem.
Soc., Faraday T rans., 1997, 93, 2439.
14 C. J. Brinker and G. W. Scherer, SolÈGel Science: T he Physics
and Chemistry of SolÈGel Processing, Academic Press, NY, 1990.
15
16
17
Z. Zhang, Y. Tanigami and R. Terai, J. Non-Cryst. Solids, 1995,
91, 304.
J. Livage, F. Babonneau, M. Chatry and L. Coury, Ceram. Int.,
1
1
997, 23, 13.
S. Brunauer, P. H. Emmett and E. Teller, J. Am. Chem. Soc.,
1938, 60, 309.
18 E. P. Barrett, L. G. Joyner and P. H. Halenda, J. Am. Chem. Soc.,
1
951, 73, 373.
A. D. Irwin, J. S. Holmgren and J. Jonas, J. Mater. Sci., 1988, 23,
908.
1
9
0
2
2
G. Sheng, L. Chu, W. A. Zeltnew and M. A. Anderson, J. Non-
Cryst. Solids, 1992, 147/148, 548.
We thank The Environmental Science and Technology Alli-
ance of Canada (ESTAC), Natural Sciences and Engineering
Research Council (NSERC) and Ontario University Research
Incentive Fund for Ðnancial support, Mr. T. R. B. Jones for
MS analysis, Mr. P. Elliott (University of York, UK) for TG
analysis and Dr. S. J. Tavener (University of York, UK) for
SEM analysis.
21 G. A. Pozarnsky and A. V. McCormick, J. Non-Cryst. Solids,
1995, 190, 212.
2
3
4
2
2
2
R. D. Gonzalez, T. Lopez and R. Gomez, Catal. T oday, 1997, 35,
2
93.
G. Engelhardt and D. Michel, High Resolution Solid State NMR
of Silicates and Zeolites, Wiley, Chichester, 1987.
K. J. D. Mackenzie, R. H. Meinhold, J. E. Patterson, H. Schnei-
der, M. Scmucker and D. Voll, J. Eur. Ceram. Soc., 1996, 16,
1
299.
References
2
2
5
6
J. Livage and C. Sanchez, J. Non-Cryst. Solids, 1992, 11, 145.
E. J. A. Pope and J. D. Mackenzie, J. Non-Cryst. Solids, 1986, 87,
1
2
3
4
5
J. H. Clark, A. P. Kybett and D. J. Macquarrie, Supported
Reagents Preparation, Analysis and Applications, VCH, NY, 1992.
Clayzic is a trademark of Contract Catalysts Ltd., Merseyside,
UK.
J. H. Clark, A. P. Kybett, D. J. Macquarrie, S. J. Barlow and P.
Landon, J. Chem. Soc., Chem. Commun., 1989, 1353.
J. H. Clark, S. R. Cullen, S. J. Barlow and T. W. Bastock, J.
Chem. Soc., Perkin T rans. 2, 1994, 1117.
1
85.
2
7
8
A. Kareiva, C. J. Harlan, D. B. MacQueen, R. L. Cook and A. R.
Barron, Chem. Mater., 1996, 8, 2331.
S. J. Gregg and K. S. Sing, Absorption, Surface Area and Porosity,
2
2
nd edn., Academic Press, London, 1982.
2
3
9
0
J. S. Beck, US Pat., 5 057 296, 1991.
T. Adachi and S. Sakka, J. Non-Cryst. Solids, 1988, 99, 118.
S. J. Barlow, T. W. Bastock, J. H. Clark and S. R. Cullen, T etra-
hedron L ett., 1993, 34, 3339.
Paper 7/07137J; Received 2nd October, 1997
J. Chem. Soc., Faraday T rans., 1998, V ol. 94
795