Mesostructured aluminosilicate alkylation catalysts for the production of aromatic amine antioxidants

Article abstract of DOI:10.1016/j.jcat.2004.04.009

The conversion of diphenylamine (DPA) and α-methylstyrene (AMS) to the antioxidants mono- and dicumenyldiphenylamine was carried out over mesostructured aluminosilicate catalysts with hexagonal (2% Al-MCM-41), wormhole (2% Al-HMS), and lamellar/vesicular (2% Al-MSU-G) framework structures. A commercial acid-treated clay catalyst, Engelhard F-20, was included for comparison perposes. The yields of the desired antioxidant, namely dicumenyldiphenylamine (DCDPA), increased in the order F-20 (～57%)<2% Al-MCM-41 (～85%)<2% Al-HMS, 2% Al-MSU-G (～90%) when the reaction was carried out under stoichiometric reaction conditions at 90°C. The DCDPA yields obtained with the mesostructured catalysts are the highest reported to date for this technologically important antioxidant. A heteropolyacid catalyst, H₃PW₁₂O₄₀·xH₂O (PW ₁₂) supported on mesostructured wormhole HMS and lamellar/vesicular MSU-G silica, also was examined as a catalyst for DCDPA production. The supported catalyst systems, however, afforded lower maximum yields of DCDPA (～73-80%) in comparison to the mesostructured aluminosilicate catalysts. The exceptionally high yields of alkylated products obtained with the mesoporous aluminosilicate catalysts in comparison to the F-20 clay and the supported PW₁₂ catalysts are attributable in part to intermediate acid strengths that minimize completing dimerization reactions of the AMS alkylating agent. Also, the pore structures of the mesostructured catalysts facilitate access to active sites on the framework walls and provide more efficient transport of reagents to framework reaction centers. Also, the regular mesoporosity of the aluminosilicate catalysts makes these structures less prone to pore plugging and to the masking of acidity through the adsorption of the high molecular weight reaction products.

Full text of DOI:10.1016/j.jcat.2004.04.009

Journal of Catalysis 225 (2004) 381–387
www.elsevier.com/locate/jcat
Mesostructured aluminosilicate alkylation catalysts for the production
of aromatic amine antioxidants
Yu Liu, Seong Su Kim, and Thomas J. Pinnavaia ∗
Department of Chemistry, Michigan State University, East Lansing, MI 48824-1322, USA
Received 12 January 2004; revised 7 February 2004; accepted 7 April 2004
Available online 25 May 2004
Abstract
The conversion of diphenylamine (DPA) and α-methylstyrene (AMS) to the antioxidants mono- and dicumenyldiphenylamine was carried
out over mesostructured aluminosilicate catalysts with hexagonal (2% Al-MCM-41), wormhole (2% Al-HMS), and lamellar/vesicular (2%
Al-MSU-G) framework structures. A commercial acid-treated clay catalyst, Engelhard F-20, was included for comparison perposes. The
yields of the desired antioxidant, namely dicumenyldiphenylamine (DCDPA), increased in the order F-20 (∼ 57%) < 2% Al-MCM-41
◦
(∼ 85%) < 2% Al-HMS, 2% Al-MSU-G (∼ 90%) when the reaction was carried out under stoichiometric reaction conditions at 90 C. The
DCDPA yields obtained with the mesostructured catalysts are the highest reported to date for this technologically important antioxidant.
A heteropolyacid catalyst, H PW
O
· xH O (PW ) supported on mesostructured wormhole HMS and lamellar/vesicular MSU-G silica,
3
12 40
2
12
also was examined as a catalyst for DCDPA production. The supported catalyst systems, however, afforded lower maximum yields of
DCDPA (∼73–80%) in comparison to the mesostructured aluminosilicate catalysts. The exceptionally high yields of alkylated products
obtained with the mesoporous aluminosilicate catalysts in comparison to the F-20 clay and the supported PW catalysts are attributable in
12
part to intermediate acid strengths that minimize completing dimerization reactions of the AMS alkylating agent. Also, the pore structures
of the mesostructured catalysts facilitate access to active sites on the framework walls and provide more efficient transport of reagents to
framework reaction centers. Also, the regular mesoporosity of the aluminosilicate catalysts makes these structures less prone to pore plugging
and to the masking of acidity through the adsorption of the high molecular weight reaction products.
2004 Elsevier Inc. All rights reserved.
Keywords: Mesoporous; MCM-41; HMS; MSU-G
1. Introduction
MCM-41 [5] and a Ga-MCM-41 derivative [6] were shown
to be far more active than zeolite Beta for the benzylation
of benzene. In addition, AlCl3/MCM-41 [7,8] was found to
provide better selectivity toward 2,6-disopropylnaphthalene
in the liquid phase isopropylation of naphthalene in compar-
ison to the zeolite mordenite.
Pore size alone, however, is not the only factor favor-
ing the reactivity of mesostructured metal oxide catalysts.
The framework pore connectivity and hierarchical structure
also are important in facilitating access to the intracrystal ac-
tive sites of a mesostructure. For example, aluminosilicates
with 3D wormhole framework motifs [9,10] are substantially
more active than unidimensional MCM-41 for the reduction
of NO by NH3 [11], the cracking of cumene [12,13], and
the alkylation [14] and the peroxidation [15] of 2,6-di-tert-
butylphenol. Similar differences in catalytic activity have
been observed between MCM-41 and lamellar mesostruc-
Although zeolite catalysts have been successfully used
for the production of fine organic chemicals [1–3] they have
limited utility for the transformations of large molecules in
the liquid state due to diffusion limitations caused by the
restricted pore sizes. Mesostructured aluminosilicates with
pores in the 2–50 nm range have been recognized as be-
ing potentially superior catalysts over zeolites for condensed
phase catalytic reactions, in part, because the large pore size
offers the possibility of minimizing diffusion limitations. In-
deed, Al-MCM-41 has been shown to be an effective catalyst
for the alkylation of a bulky phenol using cinnamyl alcohol
as the alkylating agent [4]. Also, a modified form of Al-
*
Corresponding author. Fax: +1-517-432-1225.
E-mail address: pinnavai@cem.msu.edu (T.J. Pinnavaia).
0021-9517/$ – see front matter
doi:10.1016/j.jcat.2004.04.009
2004 Elsevier Inc. All rights reserved.

382
Y. Liu et al. / Journal of Catalysis 225 (2004) 381–387
tures with a vesicular hierarchical structure that shortens the
framework pore length for more facile access to intracrys-
tal active sites [16]. Aside from the significance of frame-
work pore motif in the diffusion of large organic molecules
during the condensed-phase alkylation reaction [14], parti-
cle size and textural pore between particles are also impor-
tant. Chmelka and co-workers recently [17] reported that the
reaction rate of alkylation of toluene with benzyl alcohol
over small particle Al-SBA-15 was much higher than over
monolithic particles. As in the case of many other alkyla-
tion reactions where diffusion can limit the reaction rates
and selectivities [4,18,19] the alkylation of diphenylamine
(DPA) is carried out under condensed-phase reaction con-
ditions. Acid-treated and rare earth-modified clays have re-
ceived considerable attention as catalysts for this industrially
important reaction [20–22]. The F-series acid-treated clays
provided by Engelhard are especially active catalysts for the
alkylation of diphenylaminewith α-methylstyrene (AMS) as
the alkylating agent [22]. The desired alkylation products are
monocumyldiphenylamine (MCDPA) and dicumyldipheny-
lamine (DCDPA). These products, particularly the dialky-
lated derivative, are effective antioxidants in many formu-
lations, including hot melt adhesives, polyacetals, nylon 6,
polypropylene, polyethylene, ethylene–propylene copoly-
mers and tripolymers, ABS, synthetic lubricants, and poly-
ether polyols, among others. Although 100% conversion of
diphenylamine is achieved, the selectivity to DCDPA only
reaches 65% with a significant amount of MCDPA and unde-
sired side products being produced. Another drawback was
the substantial deactivation of the catalyst after only one
batch reaction cycle.
In the present work we have investigated the prop-
erties of mesostructured aluminosilicate catalysts for the
α-methylstyrene alkylation of diphenylamine. In addition
to hexagonal Al-MCM-41 with unidimensional framework
pores, we have included in the study Al-HMS and Al-MSU-G
mesostructures with wormhole and vesicular (lamellar)
framework structures, respectively. As noted above, these
latter mesostructures generally provide improved access to
framework acid sites in comparison to MCM-41, particu-
larly for large molecules under condensed-phase reaction
conditions. It was of interest to us, therefore, to determine
whether improved catalytic performance would also be re-
alized for the α-methylstyrene/diphenylamine reaction sys-
tem. To investigate the effect of acid strength on the yields
of DCDPA, we included the strongly acidic Engelhard F-20
clay and several supported versions of the hetero polyacid
H3PW12O40 · xH2O as acid catalysts.
was obtained from Engelhard Corporation (USA). Fumed
silica, tetraethylorthosilicate (TEOS), dodecylamine (DDA),
cetyltrimethylammonium bromide (CTAB), tetramethylam-
monium hydroxide (TMAOH), the 12-tungstophosphoric
acid H3PW12O40 · xH2O (PW12) and 2,6-di-tert-butylpyri-
dine were purchased from Aldrich.
2.2. 2% Al-MCM-41
A mixture of fumed silica, cetyltrimethylammonium bro-
mide, trimethylammonium hydroxide, and water was stirred
at room temperature for 1 h and then transferred into
an autoclave. The synthesis gel was heated at 150 ◦C for
24 h in an autoclave to obtain a hexagonal MCM-41 sil-
ica. The autoclave was cooled to room temperature and an
amount of Al(NO3)3 sufficient to provide a Si/Al molar ra-
tio of 49 was added under stirring. The resulting mixture of
molar composition 1.00SiO2:0.02Al(NO3)3:0.20CTAMBr:
0.26TMAOH:110H2O was further heated at 150 ◦C for an-
other 24 h. The final product was filtered, washed, dried at
room temperature, and calcined in air at 600 ◦C for 4 h to
remove the surfactant.
2.3. 2% Al-HMS
Tetraethylorthosilicate was added to a mixture of dode-
cylamine (DDA), water, ethanol (EtOH), and mesitylene
(MES) at room temperature. The mixture was stirred for
1 h and then Al(OBus)3 in sec-butanol was added under stir-
ring. After a reaction time of 24 h at room temperature, the
solid was filtered, washed, dried at room temperature, and
caclined in air at 620 ◦C for 4 h. The molar ratio of the above
synthesis mixture was 1.00SiO2:0.02Al:0.25DDA:1.12MES:
5.0ETOH:130H2O.
2.4. 2% Al-MSU-G
The pure silica form of MSU-G was prepared by reaction
of TEOS, neutral Gemini surfactant C12H25NH(CH2)2NH2,
EtOH, and water under hydrothermal conditions according
to our previously reported procedure [16,23]. Al(OBus)3
was added to the as-synthesized silica MSU-G in its mother
liquor to achieve an overall Si/Al ratio of 49:1. The resultant
mixture was heated again at 100 ◦C for 48 h under stirring.
The final solid was filtered, washed with water and ethanol,
air dried, and calcined at 650 ◦C for 4 h.
2.5. Supported H3PW12O40(PW12) catalysts
Calcined forms of mesostructured HMS and MSU-G sil-
icas were prepared according to the procedures described
above without the addition of Al(OBus)3 to the reaction mix-
ture. The pure silica mesostructures were dried at 150 ◦C
under vacuum conditions for 5–10 h and impregnated
with a methanol solution containing the desired amount of
H3PW12O40. The mixture was stirred at room temperature
2. Experimental
2.1. Materials
Diphenylamine and α-methylstyrene were obtained form
Aldrich and used as purchased. The acid-treated clay F-20

Y. Liu et al. / Journal of Catalysis 225 (2004) 381–387
383
for 24 h and then the methanol was removed under vacuum
at room temperature. The impregnated solid was then dried
at 130 ◦C under vacuum for 10 h.
of the isolated products were identical to those of authentic
MCDPA and DCDPA samples.
3. Results and discussion
2.6. Physical measurements
3.1. Structural and textural properties of the catalysts
Powder X-ray diffraction patterns were measured us-
ing Cu-Kα radiation (λ = 1.542 Å) and a Rigaku Rotaflex
diffractometer equipped with a rotating anode operated at
45 kV and 100 mA. Counts were accumulated every 0.02◦
(2θ) at a scan speed of 1◦2θ min−1. N2 adsorption and
desorption isotherms were obtained at −196 ◦C on a Mi-
cromeritics ASAP 2010 sorptometer using static adsorption
procedures. Samples were outgassed at 150 ◦C and 10−6
Torr for a minimum 12 h prior to analysis. BET surface
areas were calculated from a linear part of the BET plot
according to IUPAC recommendations. Pore-size distribu-
tions were calculated from the N2 adsorption branch using
the Horvath–Kawazoe model.
Fig. 1 provides the low angle X-ray powder diffrac-
tion patterns for 2% Al-MCM-41, 2% Al-HMS, and 2%
Al-MSU-G aluminosilicate mesostructures with hexagonal,
wormhole and lamellar/vesicular framework morphologies,
respectively. Included for comparison purposes are the pat-
terns for the supported PW12-HMS and PW12-MSU-G cat-
alysts, wherein the wormhole HMS and lamellar/vesicular
MSU-G silicas, respectively, have been intercalated at the
20 wt% level with the heteropoly acid H3PW12O40. The 2%
Al-MCM-41 mesophase exhibits four hkl diffraction lines
consistent with the expected hexagonal framework structure.
However, the wormhole and lamellar frameworks of HMS
and MSU-G, as well as the corresponding derivatives inter-
calated by the heteropoly acid H3PW12O40, each exhibit a
single diffraction line. TEM images, shown in Fig. 2, verify
the wormhole and lamellar/vesicular structure assignments
for the HMS and MSU-G mesophases, as well as the hexag-
onal framework structure for MCM-41.
TEM images were obtained on a JEOL 100CX micro-
scope equipped with a CeB6 gun operated at an acceleration
voltage of 120 kV. The specimen was loaded onto a holey
carbon film that was supported on a copper grid by dipping
the grid into a sample suspension in ethanol.
The acidities of 2% Al-HMS, 2% Al-MCM-41, and F-20
were measured by means of thermogravimetric analysis
(TGA) analysis of chemisorbed 2,6-di-tert-butylpyridine in
the temperature range of 150–600◦C [24]. The samples were
exposed to liquid 2,6-di-tert-butylpyridine at 80 ◦C for 4 h
and then kept at room temperature overnight so as to allow
the base to permeate the samples. TGA curves were obtained
using a CAHN121 TGA analyzer. The samples were purged
with N2 at 150 ◦C for 1 h to removal the physically adsorbed
2,6-di-tert-buty◦lpyridine then heated to 600 ◦C at a heating
rate of 10 min/ C. The molecule sizes of MCDPA and DC-
DAP were calculated by Spartan software. The amount of
organic compounds trapped in the used catalysts was mea-
sured by TGA in the temperature range of 150–800◦C in an
air flow. Prior to the TGA measurement, the used catalysts
were washed thoroughly with methanol.
The textural properties of the aluminosilicate mesophases,
together with those for the commercial acid-treated F-20
clay and PW12-intercalated HMS catalysts, are provided in
Table 1. The surface areas, as determined by fitting the BET
equation to nitrogen adsorption isotherms in the partial pres-
sure region below a partial pressure of 0.30, decreased in the
order 2% Al-MCM-41 > 2% Al-HMS > 2% Al-MSU-G >
2.7. Catalytic studies
The catalytic reactions were performed in a 25-ml three-
necked flask containing the desired amount of dipheny-
lamine and α-methylstyrene. The reactions were carried out
at 90 ◦C for 6–24 h. Prior to reaction, the catalysts were dried
under vacuum at 100 ◦C for 8 h. The reaction products were
analyzed by GC on a HP 5890 instrument equipped with a
FID detector. A 15-m SPB-1 capillary column was used for
the analysis of diphenylamine, α-methylstyrene, and mono-
and dicumyldiphenylamine. The monocumyldiphenylamine
and dicumyldiphenylaminewere separated from the reaction
mixtures by fractional distillation and their structures were
verified by GC-MS, 1H NMR, and 13C NMR. The properties
Fig. 1. Low angle XRD patterns of alkylation catalysts.

384
Y. Liu et al. / Journal of Catalysis 225 (2004) 381–387
Fig. 3. Thermally programmed desorption of 2,6-di-tert-butylpyridine over
F-20 clay and mesostructured 2% Al-MCM-41 and 2% Al-HMS alumi-
nosilicate catalysts.
sufficiently large to accommodate both the DPA starting
reagent and the DCDPA reaction product with molecular di-
mensions of 0.91 × 0.49 × 0.52 and 1.66 × 0.66 × 0.72 nm,
respectively, as estimated using SPARTAN software. The
acid-treated F-20 clay catalyst lacks framework mesoporos-
ity. Intercalating mesostructured wormhole framework of
HMS silica with up to 20 wt% H3PW12O40 did not sub-
stantially affect the pore size or the surface area of the host
structure.
In order to judge the relative acidity of the mesostructured
catalysts in comparison to the acid-treated F-20 clay cata-
lyst, we examined the temperature-programmed desorption
of chemisorbed 2,6-di-tert-butylpyridine from each catalyst.
The desorption curves for F-20 clay, 2% Al-MCM-41, and
2% Al-HMS mesostructures are shown in Fig. 3. The total
amount of chemisorbed 2,6-di-tert-butylpyridine was some-
what greater for the F-20 clay (0.21 mmol/g) than for the
mesostructures (0.13–0.17 mmol/g). Although each catalyst
showed bimodal acid strengths, as reflected by the presence
of low and high temperature desorption peaks, the high tem-
perature desorption peak occurred at 538 ◦C for the clay
catalyst, whereas the corresponding peak for the mesostruc-
tured catalyst was centered near 500 ◦C. Thus, the clay cat-
alyst exhibited not only more acid sites in comparison to
the mesostructured catalysts, but also stronger acid sites.
Also, access to the acid sites of each catalyst by the 2,6-di-
tert-butylpyridine molecule, with the approximate molecular
dimensions of 0.75 ×0.69 × 0.43 nm, was not limited by
pore-size considerations.
Fig. 2. TEM images showing the hexagonal framework pore structure of
2% Al-MCM-41, the intraparticle textural pores (arrows) and the worm-
hole framework structure (insert) of 2% Al-HMS, and the lamellar/vesicular
structure of 2% Al-MSU-G.
F-20. As will be shown below, the activities and selectivities
of these catalysts for diphenylamine alkylation did not paral-
lel the BET surface areas. Note that the mesostructured cata-
lysts have framework pore sizes are in the range 3.4–3.8 nm,

Y. Liu et al. / Journal of Catalysis 225 (2004) 381–387
385
Table 1
a
Textural properties of alkylation catalysts
2
b
Catalyst
Framework structure
BET surface; area (m /g)
Pore vol. (cc/g)
HK pore size (nm)
Acidity (mmol DBPD/g)
2% Al-MCM-41
2% Al-HMS
Hexagonal
Wormhole
Wormhole
Wormhole
Lamellar/vesicular
None
907
812
890
876
444
380
1.0
3.6
3.8
4.3
4.2
3.4
–
0.13
0.17
–
1.3
10% PW -HMS
1.5
12
20% PW -HMS
12
2% Al-MSU-G
1.3
–
0.57
0.38
–
F-20 clay
a
0.21
◦
Samples were out-gassed at 150 C for 24 h prior to the determination of textural properties by nitrogen adsorption.
b
The reported pore volumes were obtained from the nitrogen uptake at P/P = 0.98.
0
3.2. Catalytic properties of mesostructured
aluminosilicates
reaction time from 6 to 24 h, indicating that competing re-
actions deplete the AMS alkylating agent and limit the yield
of the desired dialkylated antioxidant. The unsaturated AMS
alkylating agent is known to form dimeric species in the
presence of strong acid catalysts [25] according to the re-
action Scheme 2.
The alkylation of diphenylamine to monocumyldipheny-
lamone and dicumyldiphenylamine with α-methylstyrene as
the alkylating reagent is represented in Scheme 1.
We attribute analogous dimerization reactions to the de-
pletion of the AMS alkylating agent with F-20 clay as the
alkylation catalyst. This acid-washed quasicrystalline alu-
minosilicate is known to be sufficient in acid strength to
promote dimerization of AMS [26]. Moreover, the acidity
of F-20 clay has been reported to be even stronger than sul-
fated zirconia and proton-exchangedresins for the alkylation
of 4-methoxyphenol with MTBE [27]. On the other hand,
mesostructured aluminosilicates with amorphous framework
walls are recognized as being considerably weaker acid cat-
alysts [28,29]. Independent verification of the relative acid
strengths of F-20 clay in comparison to 2% Al-HMS and
2% Al-MSU-G is provided by the thermogravimetric des-
orption curves for 2,6-di-tert-butylpyridine in Fig. 3. The
peak desorption rate occurs at a substantially higher tem-
perature for the clay catalyst (538 ◦C) in comparison to the
mesostructured catalysts (500 ◦C). Thus, the lower acidity
for the mesostructures suggests that they should be less
prone to form high surface concentrations of onium ions
through alkyl group protonation and, therefore, less likely
to promote AMS dimerization reactions.
As is shown by the catalytic results presented in Table 2
for the reaction at a stoichiometric DPA:AMS molar ratio of
1:2, all three mesostructured aluminosilicates provide more
than 90% conversion of DPA after a reaction time of 6 h at
90 ◦C. An even higher DPA conversion (∼ 99%) is achieved
with the acidic F-20 clay under equivalent reaction condi-
tions. However, the mesostructured catalysts afford primar-
ily the mono- and dialkylated products at this point in the
reaction, whereas the F-20 clay catalyst provides substantial
fractions of other reaction products. These latter features of
the catalysts can be gleaned from the product distributions
obtained by extending the reaction time to 24 h. All four
catalysts provide 100% conversions of DPA after 24 h, but
only the mesostructured catalysts transformed monoalky-
lated amine to dialkylated amine under these reaction con-
ditions, resulting in DCDPA yields of ∼ 90%. However, the
DCDPA yield obtained with the F-20 clay catalyst is only
marginally improved from 52.8 to 56.7% by extending the
Because the yields of the desired dialkylated antioxi-
dant are limited by the consumption of the alkylating agent
through the dimerization reactions shown in Scheme 2, we
repeated the alkylation reaction at a DPA:AMS ratio of 1:3
(see Table 2). As expected in the presence of excess alky-
lating agent, the conversion of DPA was complete after a
reaction time of 6 h, and the yields of DCDPA increased at
the expense of MCDPA in comparison to the reactions car-
ried out at a 1:2 DPA:AMS molar ratio.
Scheme 1. The scheme of alkylation of DPA with AMS.
Scheme 2.

386
Y. Liu et al. / Journal of Catalysis 225 (2004) 381–387
Table 2
◦
Aluminosilicate catalysts for the conversion of diphenylamine (DPA) at 90 C to monocumyldiphenylamine (MCDPA) and dicumyldiphenylamine (DCDPA)
with α-methylstyrene (AMS) as the alkylating agent
a
Catalyst
Rxn time (h)
DAP:AMS ratio
DPA conv. (%)
Distribution of alkylated products (%)
Yield of DCDPA (%)
MCDPA
DCDPA
2% Al-MCM-41
2% Al-HMS
6.0
24.0
1:2
1:2
93.7
100
27.5
14.9
72.5
85.1
67.9
85.1
6.0
24.0
1:2
1:2
95.8
100
25.6
8.9
74.4
90.1
71.3
90.1
2% Al-MSU-G
6.0
24.0
1:2
1:2
96.7
100
23.8
8.7
76.2
91.3
73.7
91.3
F-20
F-20
6.0
24.0
1:2
1:2
98.7
100
46.5
43.3
53.5
56.7
52.8
56.7
2% Al-MCM-41
2% Al-HMS
6.0
6.0
6.0
6.0
1:3
1:3
1:3
1:3
100
100
100
100
20.5
3.5
79.5
96.5
97.2
64.5
79.5
96.5
97.2
64.5
2% Al-MSU-G
2.8
F-20
a
35.2
The amount of catalyst used was 1.0 g per 0.050 mol AMS (DPA:AMS = 1:2) or 0.075 mol AMS (DPA:AMS = 1:3).
Table 3
◦
Mesostructure-supported phosphotungstic acid (PW ) for the catalytic conversions of diphenylamine (DPA) at 90 C to monocumyldiphenylamine (MCDPA)
and dicumyldiphenylamine (DCDPA) with α-methylstyrene (AMS) as the alkylating agent
12
a
Catalyst
Rxn time (h) DAP:AMS ratio DPA conv. (%) AMS conv. (%) Distribution of alkylated products (%)
MCDPA DCDPA
Yield of DCDPA (%)
10% PW -HMS
6.0
6.0
1:2
1:2
93.5
93.8
100
100
22.1
18.9
77.9
81.1
72.8
76.0
12
20% PW -HMS
12
10% PW -MSU-G 6.0
1:2
1:2
94.2
96.5
100
100
18.4
17.6
81.6
82.4
76.9
79.5
12
20% PW -MSU-G 6.0
12
a
The amount of catalyst used was 1.0 g per 0.050 mol AMS.
In order to investigate the longevity of the mesostruc-
tured aluminosilicate catalysts for the conversion of DPA to
DCDPA under stoichiometric reaction conditions, the cat-
alysts were recovered after the first reaction cycle, washed
with ethanol, and subjected to a second 24-h reaction cycle
at 90 ◦C. Within experimental uncertainty, the mesostruc-
tured catalysts afforded the same DPA conversions (100%)
and high yields (ꢀ 85%) of dialkylated antioxidant that was
achieved in the first cycle. However, the DPA conversion and
the PCDPA product yield obtained with the F-20 clay cata-
lyst dropped from 100 and ∼ 57%, respectively, in the first
cycle, to 74 and 31% in the second cycle. The loss of alkyla-
tion reactivity for analogous acid-washed clay catalysts has
been noted previously and attributed to micropore blocking
and the loss of acid sites by adsorbed reaction products [22].
These adsorbed components led to a decrease in the surface
area and to the masking of acid sites on the surface of the
catalyst. Indeed, we find that the F-20 clay looses ∼ 23% of
its normalized nitrogen BET surface area after the first reac-
tion cycle due to the adsorption of 20% by weight of the high
molecular weight reaction products in the micropores. How-
ever, the normalized surface areas and the mesopore sizes
of the mesostructured aluminosilicate catalysts remained un-
changed after the first reaction cycle.
3.3. Catalytic properties of mesostructure-supported
H3PW12O40
Mesostructured silicas have been shown to be effec-
tive supports for the immobilization of the heteropoly acid
H3PW12O40 (PW12) for acid-catalyzed phenol alkylations
[30,31] and iso-butane/butene alkylation [32]. Similar activ-
ity may be anticipated for the alkylation of DPA. Accord-
ingly, we have examined the conversion of DPA and AMS
over H3PW12O40 supported on wormhole HMS and lamel-
lar/vesicular MSU-G silicas.
Table 3 provides the yields of the desired DCDPA antioxi-
dant obtained at H3PW12O40 loadings of 10 and 20% (w/w),
a DPA:AMS molar ratio of 1:2 and a reaction time of 6 h at
90 ◦C. Under these reaction conditions the AMS alkylating
reagent is completely depleted from the reaction mixture ow-
ing to its reaction with DPA to form mono- and dialkylated
reaction products and, notably, to the competitive dimeriza-
tion reaction according to reaction (Scheme 2) above. Due
to the depletion of AMS through dimerization, the yields of
the desired DCDPA antioxidant are maximized after a reac-
tion time of 6.0 h or less to values of 73–76 and 77–80% for
PW12 supported on HMS and MSU-G silicas, respectively.
As expected, increasing the reaction time over these cata-
lysts to 24 h resulted in no improvement in DCDPA yields.

Y. Liu et al. / Journal of Catalysis 225 (2004) 381–387
387
Although the rates of alkylation are slower over 2% Al-HMS
and 2% Al-MSU-G, resulting in DCDPA yields of 71.3 and
73.7% after 6.0 h reaction times, the AMS dimerization rates
for these catalysts also are lower. Thus, increasing the re-
action time over these latter catalysts to 24 h boosts the
yields of DCDPA to values of 90.1 and 91.3% (cf. Table 2).
Thus, the mesostructure-supportedH3PW12O40 catalysts are
substantially less selective alkylation catalysts than the cor-
responding aluminated mesostructures. It appears that the
high acid strength of the supported H3PW12O40 catalysts,
which compromises the alkylation of both DPA and MCDPA
through AMS dimerization, causes the catalytic properties to
be similar to those of the strong acid F-20 clay catalyst.
parison to the commercial F-20 clay catalyst by being less
prone to the masking of acidity through the adsorption of
reaction products.
Acknowledgment
The support of this research by the National Science
Foundation through CRG Grant CHE-0211029 is gratefully
acknowledged.
References
[1] A. Corma, Chem. Rev. 95 (1995) 559.
[2] A. Corma, Chem. Rev. 97 (1997) 2373.
4. Conclusions
[3] A. Corma, H. Garcia, Catal. Today 38 (1997) 257.
[4] E. Armengol, M.L. Cano, A. Corma, H. Garcia, M.T. Navarro, Chem.
Commun. (1995) 519.
[5] S. Jun, S.H. Joo, R. Ryoo, M. Kruk, M. Jaroniec, Z. Liu, T. Ohsuna,
O. Terasaki, J. Am. Chem. Soc. 122 (2000) 10712.
[6] K. Okumura, K. Nishigaki, M. Niwa, Micropor. Mesopor. Mater. 44–
45 (2001) 5096.
[7] X.S. Zhao, G.Q. Lu, C. Song, Chem. Commun. (2001) 2306.
[8] X.S. Zhao, G.Q. Lu, C. Song, J. Mol. Catal. A: Chem. 191 (2003) 67.
[9] P.T. Tanev, T.J. Pinnavaia, Science 267 (1995) 865.
[10] S.A. Bagshaw, E. Prouzet, T.J. Pinnavaia, Science 269 (1995) 1242.
[11] W. Zhang, T.R. Pauly, T.J. Pinnavaia, Chem. Mater. 9 (1997) 2491.
[12] K.R. Kloetstra, J.C. Jansen, Chem. Commun. (1997) 2281.
[13] R. Mokaya, W. Jones, J. Catal. 172 (1997) 211.
[14] T.R. Pauly, Y. Liu, T.J. Pinnavaia, S.J.L. Billinge, T.P. Rieker, J. Am.
Chem. Soc. 121 (1999) 8835.
[15] P.T. Tanev, M. Chibwe, T.J. Pinnavaia, Nature 368 (1994) 321.
[16] S.S. Kim, Y. Liu, T.J. Pinnavaia, Micropor. Mesopor. Mater. 44 (2001)
489.
[17] J.J. Chiu, D.J. Pine, S.T. Bishop, B.F. Chmelka, J. Catal. 221 (2004)
400.
[18] E.G. Derouane, C.J. Dillon, D. Bethell, S.B. Derouane-Abd Hamid,
J. Catal. 187 (1999) 209.
[19] E.G. Derouane, G. Crehan, C.J. Dillon, D. Bethell, H. He, S.B.
Derouane-Abd Hamid, J. Catal. 194 (2000) 410.
All of the solid acid catalyst examined in this study are ef-
fective in converting diphenylamine to monocumyldipheny-
lamine in an initial alkylation step using α-methylstyrene
as the alkylating agent (cf. Tables 2 and 3). However, the
hexagonal 2% Al-MCM-41, wormhole 2% Al-HMS, and
lamellar/vesicular 2% Al-MSU-G mesostructures are sub-
stantially more selective in comparison to the commercial
F-20 clay and supported H3PW12O40 catalysts for the con-
version of MCDPA to the desired dicumyldiphenylamine
antioxidant in a second alkylation step.
The superior selectivity of the aluminosilicate mesostruc-
tures for the second alkylation process is attributable in
part to an acid strength that allows for the protonation of
the α-methylstyrene and coadsorption of MCDPA, while
at the same time minimizing the surface concentration of
α-methylstyrene for conversion to undesired dimers. At an
initial DPA:AMS reaction stoichiometry of 1:2, ∼ 27% of
the AMS alkylating agent is converted to dimeric products
over the commercial F-20 clay catalyst, and ∼ 12–15% of
the alkylating agent is lost to dimeric products over the
mesostructure-supported H3PW12O40 catalysts. Thus, these
competing dimerization reactions limit the DCDPA yields
to only 57% for the F-20 clay and to 73–80% in the case
of the supported H3PW12O40 catalysts, depending on the
12-tungstophosphoric acid loading and the framework struc-
ture of the support. In comparison, the hexagonal, wormhole
and lamellar/vesicular aluminosilicate mesostructures con-
vert only 7.5, 5.0, and 4.4% of AMS to dimeric products,
respectively, under the same reaction conditions. By mini-
mizing the competing dimerization reaction of the alkylating
agent, the hexagonal, wormhole and lamellar/vesicular alu-
minosilicate mesostructures provide DCDPA yields of 85,
90, and 91%, respectively. These latter values represent the
highest yields reported to date for this technologically im-
portant antioxidant under stoichiometric reaction conditions.
Finally, the regularly mesoporous aluminosilicates sili-
cate catalysts show good longevity and recyclability in com-
[20] J.T. Lai, D.S. Filla, EP 810 200, 1997.
[21] J.T. Lai, D.S. Filla, U.S. 5 672 752, 1998.
[22] S.R. Chitnis, M.M. Sharma, J. Catal. 160 (1996) 84.
[23] S.S. Kim, W.Z. Zhang, T.J. Pinnavaia, Science 282 (1998) 1302.
[24] A. Corma, V. Fornes, L. Forni, F. Marquez, J. Martinez-Triguero, D.
Moscotti, J. Catal. 179 (1998) 451.
[25] A. Heidekum, M. Harmer, W.F. Holderich, Catal. Lett. 47 (1997) 243.
[26] B. Chaudhuri, M.M. Sharma, Ind. Eng., Chem. Res. 28 (1989) 1757.
[27] G.D. Yadav, M.S.M. Mujeebur Rahuman, Appl. Catal. A 253 (2003)
113.
[28] A. Corma, V. Fornes, M.T. Navarro, J. Perez-Pariente, J. Catal. 148
(1994) 569.
[29] A. Corma, M.S. Grande, V. Gonzalez-Alfaro, A.V. Orchilles, J. Ca-
tal. 159 (1996) 375.
[30] I.V. Kozhevnikov, K.R. Kloetstra, A. Sinnema, H.W. Zandbergen, H.
van Bekkum, J. Mol. Catal. A: Chem. 114 (1996) 287.
[31] I.V. Kozhevnikov, A. Sinnema, R.J.J. Jansen, K. Pamin, H. van
Bekkum, Catal. Lett. 30 (1995) 241.
[32] W. Chu, Z. Zhao, W. Sun, X. Ye, Y. Wu, Catal. Lett. 55 (1998) 57.