Activation and reactivity of epoxides on solid acid catalysts

Article abstract of DOI:10.1016/j.jcat.2007.10.002

The aminolysis of epoxides over novel solid catalysts (Broensted-acidic SBA-15 functionalized with propylsulfonic acid and Lewis-acidic Ti-MCM-41) is reported. The acidic properties of these catalysts were determined by FTIR spectroscopy and temperature-programmed desorption of pyridine and NH₃, respectively. The mesoporous solid acids of the present study are reusable and exhibit significantly higher catalytic activities than known catalysts for opening of the oxirane ring with nitrogen (aromatic and aliphatic amines)-containing and oxygen (alcohols)-containing nucleophiles. A range of β-amino alcohols with high regioselectivity and stereoselectivity were synthesized. Adsorption studies as well as the sigmoid shape of the conversion-versus-time plots show that the epoxide and amine compete for adsorption on the acidic sites ({single bond}SO₃H or Ti⁴⁺) on the catalyst surface. Epoxide adsorption and activation on acid sites are the more critical processes. Catalytic activity decreases with increasing basicity of the amines and/or the alcohol, as well as the dielectric constant of the solvent.

Full text of DOI:10.1016/j.jcat.2007.10.002

Journal of Catalysis 252 (2007) 148–160
www.elsevier.com/locate/jcat
Activation and reactivity of epoxides on solid acid catalysts
L. Saikia, J.K. Satyarthi, D. Srinivas ∗, P. Ratnasamy ∗
National Chemical Laboratory, Pune 411 008, India
Received 24 July 2007; revised 28 September 2007; accepted 2 October 2007
Abstract
The aminolysis of epoxides over novel solid catalysts (Brönsted-acidic SBA-15 functionalized with propylsulfonic acid and Lewis-acidic Ti-
MCM-41) is reported. The acidic properties of these catalysts were determined by FTIR spectroscopy and temperature-programmed desorption
of pyridine and NH , respectively. The mesoporous solid acids of the present study are reusable and exhibit significantly higher catalytic activities
3
than known catalysts for opening of the oxirane ring with nitrogen (aromatic and aliphatic amines)-containing and oxygen (alcohols)-containing
nucleophiles. A range of β-amino alcohols with high regioselectivity and stereoselectivity were synthesized. Adsorption studies as well as the
4+
sigmoid shape of the conversion-versus-time plots show that the epoxide and amine compete for adsorption on the acidic sites (–SO H or Ti
)
3
on the catalyst surface. Epoxide adsorption and activation on acid sites are the more critical processes. Catalytic activity decreases with increasing
basicity of the amines and/or the alcohol, as well as the dielectric constant of the solvent.
© 2007 Elsevier Inc. All rights reserved.
Keywords: SBA-15 functionalized with propylsulfonic acid; Ti-MCM-41; Solid acids; Aminolysis and alcoholysis of epoxides; β-Amino alcohols; Ring opening
of epoxides
1. Introduction
effective for weakly nucleophilic amines and yields low regios-
electivity [6,7]. Various catalyst systems, including sulfamic
acid, amberlyst resin, metal triflates, transition metal halides,
ionic liquids, zeolites, and Lewis acids, in a supercritical car-
bon dioxide medium have been investigated [8–15]. All of
these catalysts necessitate long reaction times, elevated tem-
peratures, high pressures, stoichiometric amounts of catalyst,
and use of expensive reagents or catalysts. Recently, the ring
opening of epoxides in water in the absence of any catalyst
was reported, but this required a long reaction time, and the
product yield was low [16]. Even though heterogeneous cat-
alysts have been used in the ring-opening addition of such
nucleophiles as amines and alcohols to epoxides, their activ-
ity and selectivity also have been low. For example, Garcia et
al. [17] reported that in the reaction between styrene epoxide
and n-propanol, zeolite Y gave conversions only in the range of
5–10%, and many secondary products were formed. Robinson
et al. [18] reported that mesoporous aluminosilicates catalyzed
the alcoholysis of epoxides to give β-alkoxyalcohols in high
yields at room temperature in short reaction times (1–4.5 h).
Recently, we reported preliminary results on the aminolysis of
Epoxides are versatile and important intermediates in phar-
maceutical and agrochemical industries. The three-membered
heterocyclic ring is strained and offers an uncommon combi-
nation of reactivity, synthetic flexibility, and atom economy. It
is susceptible to attack by a range of nucleophiles, including
nitrogen (e.g., ammonia, amines, azides), oxygen (e.g., water,
alcohols, phenols, acids), and sulfur (thiol)-containing com-
pounds, leading to bifunctional molecules of great industrial
value. For example, β-amino alcohols (Scheme 1) are used in
the synthesis of β-blockers, insecticidal agents, and oxazolines
and as chiral ligands in asymmetric synthesis [1–5]. The clas-
sical synthesis of β-amino alcohol involves the ring opening
of epoxides with amines. These reactions are conventionally
carried out with a large excess of the amines at elevated tem-
peratures and in the presence of solvents. This method is less
*
Corresponding authors. Fax: +91 20 2590 2633.
E-mail addresses: d.srinivas@ncl.res.in (D. Srinivas),
p.ratnasamy@ncl.res.in (P. Ratnasamy).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.10.002

L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
149
Scheme 1. Reactions of epoxides with nucleophiles.
epoxides over Lewis-acidic titanosilicates [19]. In the present
2. Experimental
report, we present more detailed information on this system
and compare its activity and selectivity with that of a solid
Brönsted acid (sulfonic acid-functionalized SBA-15) in this re-
action.
2.1. Material preparation
SBA-15 was prepared [36] using tetraethyl orthosilicate
(TEOS, Aldrich) as the silica source, poly(ethylene glycol)-
block-poly(propylene glycol)-block-poly(ethylene glycol) (Plu-
ronic P123, EO20PO70EO20, average molecular weight = 5800,
Aldrich) as the template, and HCl as the pH controlling agent.
3-Mercaptopropyltrimethoxysilane (MPTMS) was procured
from Aldrich. All of the solvents and chemicals were of A.R.
grade and purchased from Merck India Ltd.
Sulfonic acid-functionalized mesoporous silica [20,21] has
been used as a catalyst for the esterification of carboxylic acid
with alcohols [22–24], the synthesis of polycaprolactone [25],
and oxidation reactions [26,27]. Titanosilicate molecular sieves
(TS-1 and Ti-MCM-41) have long been known to have re-
markable selective oxidation activity of organic molecules at
mild conditions using H2O2 as an oxidant [28,29]. We have
reported on their novel application as solid catalysts for the
synthesis of polycarbonate and polyurethane precursors us-
ing CO2 instead of toxic phosgene [30–33] and for transes-
terification reactions of carbonates and carboxylic acid esters
[30,34]. Garro et al. [35] were the first to report their use as a
solid Lewis acid in Mukaiyama-type aldol condensation reac-
tions.
The aims of the present study are (1) to compare the activ-
ity and especially the selectivity of the solid Lewis (titanosili-
cates) and Brönsted (sulfonated SBA-15) acids for the aminol-
ysis of epoxides, and (2) to delineate the dominant structural,
adsorption, and other kinetic parameters that control the rates
and selectivity in this reaction. In contrast to the alcoholy-
sis of epoxides, not much information is available on the use
of solid Lewis or Brönsted acids for the aminolysis of epox-
ides. A range of β-amino and alkoxy alcohols were synthesized
using these solid catalysts in very high yields from symmetri-
cal and unsymmetrical epoxides (with aromatic and aliphatic
amines/alcohols) at ambient temperatures and under solvent-
free conditions.
2.1.1. Preparation of SBA-15 functionalized with propylthiol
Propylthiol functionalization was accomplished by con-
densing the surface silanol groups of SBA-15 with MPTMS
[37–39]. In a typical preparation, calcined SBA-15 (4 g) was
first activated under vacuum at 423 K for 4 h, then dispersed
in dry toluene (100 mL). Then 5.42 g of MPTMS was added
to it, in small amounts over 10 min. The contents of the flask
were refluxed under nitrogen for 24 h. Soxhlet extraction, first
with dichloromethane for 12 h and then with acetone for 12 h,
yielded the thiol-functionalized SBA-15 (SBA-15-pr-SH). This
was dried at 353 K. The thiol loading estimated by sulfur analy-
sis was 1.5 mmol/g.
2.1.2. Preparation of SBA-15 functionalized with
propylsulfonic acid
To 4 g of SBA-15-pr-SH (activated at 373 K) was added
65 mL of 30% aq. H2O2 [40–42]. The mixture was then stirred
for 8 h at 298 K. The solid was filtered, washed with deion-
ized water and dried, first at 298 K for 24 h and then at 353 K

150
L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
Table 1
Chemical composition and structural properties of functionalized SBA-15 materials
System
Elemental analysis (wt%)
Organic
functional
group
XRD
N
adsorption
Wall
thickness
(nm)
2
C
H
S
d
100
Unit cell
parameter
(nm)
Pore
diameter
(nm)
Specific
Total pore
volume
(cm /g)
(nm)
surface area
a
(mmol/g)
2
3
(m /g)
BET
BJH
SBA-15
SBA-15-pr-SH
SBA-15-pr-SO H
0.4
6.3
4.8
0.9
1.8
1.8
0
4.7
1.8
0
9.5
9.9
9.7
11.0
11.5
11.3
6.5
6.5
6.2
692
534
533
686
554
546
1.13
0.87
0.83
4.5
5.0
5.1
1.5 (1.7)
0.6 (0.7)
3
a
Determined from the S-content. Values in parentheses are those estimated from the thermo-gravimetric analysis data.
for 12 h, to obtain propylsulfonic acid-functionalized SBA-15
(SBA-15-pr-SO3H). Based on the elemental sulfur analysis, the
content of the -pr-SO3H functional group in this material was
0.6 mmol/g silica (Table 1). The acidity/ion exchange capacity
of the catalyst was estimated as 0.73 mEq/g silica. This esti-
mation by the titrimetric method agrees well with that obtained
from the sulfur analysis (Table 1).
pellets) were recorded on a Shimadzu 8201 PC spectropho-
tometer in the 400–4000 cm−1 region.
2.3. Acidity measurements
IR spectra of pyridine adsorbed on titanosilicates were
recorded on a Shimadzu SSU 8000 DRIFT-IR spectrometer
equipped with a liquid nitrogen-cooled MCT detector. Sam-
ples were activated at 698 K for 2 h under nitrogen flow, then
cooled to 323 K, and pyridine (30 µL) was adsorbed. The sam-
ple temperature was raised to a desired value and held at that
temperature for 30 min, after which the spectrum was recorded.
Temperature-programmed desorption (TPD) of ammonia
was performed on a Micromeritics Autochem 2910 instrument.
Titanosilicates (500 mg) were first activated at 773 K for 2 h in
He flow (20 mL/min), then cooled to 353 K, after which 10%
NH3 in He was adsorbed for 30 min. The samples were flushed
with He (30 mL/min) for 1 h at 373 K, and the desorption
was monitored by raising the temperature from 373 to 723 K
at a ramp rate of 10 K/min. In the case of SBA-15-pr-SO3H
(50 mg), the sample was first activated at 423 K for 1 h in He
flow (5 mL/min), then cooled to 298 K, after which 10% NH3
in He was adsorbed for 1 h. The sample was flushed with He
(10 mL/min) for 30 min at 373 K, and the desorption of NH3
was measured by raising the temperature from 373 to 923 K at
a ramp rate of 10 K/min and holding it at 923 K for 20 min.
Because sulfonic acid groups were thermally unstable beyond
475 K, only the desorption of NH3 at 373–473 K was consid-
ered for the present work.
2.1.3. Preparation of titanosilicate molecular sieves
TS-1 (Si/Ti = 33; SBET = 485 m2/g) was synthesized as
described previously [43]. In a typical synthesis of Ti-MCM-
41 (input ratio of Si/Ti = 30), 2.67 g of sodium hydroxide
was dissolved in 147 g of distilled water. To this, 5.94 g of
cetyltrimethylammonium bromide (CTMABr, S.D. Fine Chem-
icals Ltd., India) was added. Then 4 g of fumed silica (Aldrich)
was added slowly under stirring over 45 min. The stirring was
continued for another 1 h, with the pH of the gel maintained at
9–10 using dilute H2SO4 solution. To this was added 0.657 g
of titanium isopropoxide (Aldrich) in 5–10 mL of iso-propanol
over 15 min. The resulting gel was stirred for another 5 h at
298 K, then transferred into a Teflon-lined stainless steel au-
toclave and heated to 373 K for 48 h. The solid product was
filtered, washed with distilled water, dried at 353 K, and finally
calcined at 813 K for 6 h. The output Si/Ti ratio was found to
be 35.
2.2. Characterization techniques
X-ray diffraction (XRD) was performed with a Philips
X’Pert Pro diffractometer using CuKα radiation and a propor-
tional counter as a detector. Morphological characteristics of
the samples were determined by scanning electron microscopy
(SEM) using a Leica 440 microscope and transmission electron
microscopy (TEM) using a JEOL 1200EX microscope operat-
ing at 100 kV. Elemental analyses (C, H, and S) of the samples
were performed using a Carlo-Erba 1106 analyzer. The spe-
cific surface areas (BET) of the samples were determined using
a Quanta Chrome NOVA 1200 instrument. The data points of
p/p0 in the range of ca. 0.05–0.3 were used in the calcula-
tions. The micropore volume was estimated from the t-plot,
and the pore diameter was estimated using the BJH model.
Thermogravimetric analysis was done on a Seiko DTA-TG 320
instrument under air (50 mL/min), at a ramp rate of 10 K/min,
in the temperature range of 308–1078 K. FTIR spectra (KBr
2.4. Ion-exchange capacity of SBA-15-pr-SO3H
About 50 mg of SBA-15-pr-SO3H was dispersed in 10 mL
of 0.01 M NaCl solution and allowed to equilibrate for about
30 min. Then the solid was separated out, and the liquid por-
tion was titrated with 0.01 M NaOH solution. The ion-exchange
capacity of the catalyst was determined from the titration
value [44].
In some studies, SBA-15-pr-SO
3H was initially adsorbed
with different amounts (0.25–1.5 mmol) of styrene oxide or ani-
line. Ion-exchange capacities of these modified catalysts were
determined in the same manner as described above. Thus, the
influence of substrate adsorption on the ion exchange capacity
of SBA-15-pr-SO3H was determined.

L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
151
2.5. Reaction procedure
of a preformed, vacuum-dried SBA-15 with MPTMS [37–39].
In the second step, this was oxidized with aqueous H2O2
to the sulfonic acid functionality (SBA-15-pr-SO3H) [40–42].
MPTMS has three potential sites [–Si(OCH3)3] for binding
with the silica surface. Although in principle, all three sites
can react with surface silanol groups to form Si–O–Si bonds,
it is also possible that only one or two of the alkoxy groups
will react with surface silanol groups. Possibly, the groups
grafted via mono and, to some extent, bidentate modes can
leach out initially during reactions. In addition, two of the
SH-propyltrimethoxysilane molecules also may react with each
other, forming direct Si–O–Si bonds between the organic func-
tional molecules. However, at the low concentrations of func-
tional groups used in the present study, such disulfide (S–S)
bond formation was not detected by spectral analysis.
The SBA-15 materials showed an XRD pattern (Fig. 1a)
corresponding to a two-dimensional hexagonal p6mm symme-
try [37–42]. Well-resolved (100), (110), and (200) reflections,
consistent with a long-range mesopore ordering for SBA-15,
appeared at 2θ values of 0.919, 1.570, and 1.809◦, respectively.
On organo-functionalization, these peaks shifted to higher 2θ
values: 1.019, 1.755, and 2.035◦ for SBA-15-pr-SH and 1.034,
1.779, and 2.059◦ for SBA-15-pr-SO3H (Fig. 1a). Shift in the
peak positions to higher 2θ values also have been reported by
other researchers when transition metal ions and metal com-
plexes were introduced into mesoporous architectures [37–42].
The d-spacing (d100) and unit cell parameters (Table 1) agree
well with previously reported values [37–42].
Organic functionalization did not alter the long-range meso-
porous arrangement (Fig. 2). In general, the catalysts exhibited
type IV nitrogen adsorption/desorption isotherms with H1 hys-
teresis. On organic functionalization, a marked decrease in BET
surface area (from 692 to 533 m2/g) and total pore volume
(from 1.13 to 0.83 cm3/g) was detected (Table 1). The pore
diameters were in the range of 6.2–6.5 nm, corresponding to
mesopores [37–42]. The value of the C constant of the BET
equation decreased from 1.05×102 (for SBA-15) to 0.78×102
for SBA-15-pr-SH and 0.83 × 102 for SBA-15-pr-SO3H, due
to the elimination of acidic surface OH on silylation. In such
cases, the BET theory does not apply. However, the surface
area determined by BJH using the cumulative area of the pores
also showed a similar decrease as a consequence of organic
functionalization (Table 1). Thus, this large decrease in sur-
face area on functionalization with a few molecules of modifier
(-pr-SH and -pr-SO3H) is rather surprising. Apparently some
reconstruction of the solid occurred that we were not able to
detect.
2.5.1. Reaction of epoxides and amines: Synthesis of β-amino
alcohols
In a typical reaction, known quantities of catalyst and epox-
ide and an equimolar quantity of amine were added to a double-
necked round-bottomed flask (50 mL) placed in a temperature-
controlled oil bath and fitted with a water-cooled condenser.
The reaction was conducted at a specified temperature and for
a specified period. The progress of the reaction was monitored
by obtaining an aliquot of the sample, diluting it with a known
quantity of dichloromethane, separating the catalyst by cen-
trifugation, and subjecting the diluted liquid to gas chromato-
graphic analysis (Varian 3400; CP-SIL8CB column; 30 m long
and 0.53 mm i.d.). The products were identified by GC–MS
(Varian CP-3800; 30 m long, 0.25 mm i.d., with a 0.25-µm-
thick CP-Sil8CB capillary column). They were also isolated by
column chromatography (eluent: petroleum ether–ethyl acetate
mixture) and characterized by 1H NMR studies. The character-
istics of some isolated products are as follows:
2-Phenylamino-2-phenyl ethanol: GC–MS (m/e): 214.8,
1
213.8, and 182.0. H NMR: 3.7 (1H, dd), 3.9 (1H, dd), 4.4
(1H, dd), 6.5 (2H, d), 6.7 (1H, t), 7.1 (2H, t), 7.3 (5H, m).
2-(4-Chlorophenylamino)-2-phenyl ethanol: GC–MS (m/e):
249.1, 248.2, and 246.3. 1H NMR: 3.7 (1H, dd), 3.9 (1H, dd),
4.4 (1H, dd), 6.5 (2H, d), 7.05 (2H, d), 7.4 (5H, m).
2-(3-Methylphenylamino)-2-phenyl ethanol: GC–MS (m/e):
228.0, 227.4, 197.3, 196.4, and 118.1. 1H NMR: 2.3 (3H, s), 3.7
(1H, dd), 3.9 (1H, dd), 4.5 (1H, dd), 6.5 (3H, m), 7.0 (1H, t),
7.3 (5H, m).
2-(2-Methylphenylamino)-2-phenyl ethanol: GC–MS (m/e):
228.0, 227.4, 197.3, 196.4, and 118.1. 1H NMR: 2.3 (3H, s), 3.8
(1H, dd), 3.9 (1H, dd), 4.5 (1H, dd), 6.37 (1H, d), 6.63 (1H, t),
6.95(1H, t), 7.07 (1H, d), 7.35 (5H, m).
Kinetic studies were done by classic methods, taking differ-
ent molar ratios of reactants and conducting the experiments at
different temperatures.
2.5.2. Reaction of epoxides and alcohols: Synthesis of
β-alkoxy alcohols
Equimolar amounts of styrene oxide and alcohol and known
quantities of the catalyst were reacted in a 50-mL double-
necked, round-bottomed flask placed in a temperature-control-
led oil bath and fitted with a water-cooled condenser. The
progress of the reaction and analysis of the products were de-
termined as described in Section 2.5.1.
3. Results and discussion
SBA-15 shows characteristic IR peaks at around 1040–
1260, 820, and 500 cm−1 due to Si–O–Si stretching vibrations
and a broad, asymmetric feature at 2900–3800 cm−1 due to
the O–H of silanols and water (Fig. 1b) [37–39]. Function-
alization with -pr-SH showed additional features at 2928 and
2852 cm−1 due to C–H stretching vibrations and a weak peak
at 2575 cm−1, confirming the presence of SH functionality in
the solid [37–39]. In the case of SBA-15-pr-SO3H, the band
at 2575 cm−1 was completely absent, indicating that the –SH
groups originally present in SBA-15-pr-SH were all oxidized
3.1. Composition and structure-texture characterization of
catalysts
3.1.1. SBA-15 functionalized with propylsulfonic acid
The elemental compositions of the functionalized-SBA-15
materials are given in Table 1. Propylsulfonic acid-functional-
ized SBA-15 was prepared by the postgrafting technique. In
the first step, SBA-15-pr-SH was prepared by condensation

152
L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
Fig. 1. (a) X-ray diffractograms and (b) FTIR spectra of “neat” and organo-functionalized SBA-15. Characteristic IR peaks are indicated by arrows.
Fig. 2. Scanning electron micrographs (top) and transmission electron micrographs (bottom) of (a) SBA-15, (b) SBA-15-pr-SH and (c) SBA-15-pr-SO H.
3
to –SO3H groups. The sulfonic acid group showed bands at 650
and 1060 cm−1 in SBA-15-pr-SO3H [40–42].
Thermogravimetric analysis of SBA-15 showed three stages
of weight loss: stage I (313–475 K, exothermic), correspond-
ing to the loss of physically held water; stage II (475–815 K,
exothermic), corresponding to the loss of water within the mi-
cropore walls; and stage III (815–1273 K, endothermic), due
to silanol condensation [26,45]. SBA-15-pr-SH and SBA-15-
pr-SO3H showed four stages of weight loss: stages I, IIA, IIB,
and III. Stages IIA (475–661 K) and IIB (661–815 K) are due to

L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
153
Fig. 3. DRIFT spectra of adsorbed pyridine on (a) TS-1 and (b) Ti-MCM-41. NH -TPD: (c) functionalized SBA-15 and (d) titanosilicate catalysts.
3
desorption of water from the micropore walls, as well as decom-
position of the organic functional groups [26,45]. The amount
of organic functional groups grafted was estimated from the
weight losses; this value agreed well with that estimated from
the elemental analyses data (Table 1).
1485 cm−1 due to pyridine-coordinated to weak Lewis acid
sites (Figs. 3a and 3b). Strong Lewis acid sites (peaks at
1623 and 1455 cm−1) and Brönsted sites (peaks at 1639 and
1546 cm−1) were absent. The IR bands were relatively more
intense in Ti-MCM-41 than in TS-1, consistent with the eas-
ier accessibility of the Ti sites to pyridine in the former than in
the latter. The IR peaks due to pyridine disappeared completely
above 398 K on TS-1 and above 523 K on Ti-MCM-41, sug-
gesting that the Lewis acid sites are stronger on Ti-MCM-41
than on TS-1. Easy accessibility of Ti sites in Ti-MCM-41 with
an open tetrahedral Ti(OH)(OSi)3 structure leads to a higher
concentration of Ti-pyridine complexes compared with Ti in
TS-1.
3.1.2. Titanosilicates
The crystallinity and phase purity of TS-1 and Ti-MCM-
41 samples were confirmed based on their XRD patterns [31].
FTIR and diffuse-reflectance UV–visible spectroscopy tech-
niques confirmed the framework substitution of Ti in titanosil-
icates [28]. The textural properties of Ti-MCM-41 are as fol-
lows: SBET = 1055 m2/g; pore volume = 0.77 cm3, and aver-
age pore diameter = 29.4 Å.
3.2.2. NH3-TPD and ion-exchange capacity
3.2. Acidic properties
The organic functional groups on SBA-15-pr-SO3H are ther-
mally unstable beyond 475 K (Section 3.1.1). Thus, acidity esti-
mation from NH3-TPD may not be an accurate method for sul-
fonic acid-functionalized SBA-15 samples. Neat SBA-15 did
not desorb NH3 in the temperature range 373–473 K (Fig. 3c).
SBA-15-pr-SO3H showed a desorption peak with a maximum
at 470 K; 0.26 mmol/g of catalyst was desorbed.
3.2.1. DRIFT spectroscopy
The acidic properties of the samples were determined by
DRIFT spectroscopy of adsorbed pyridine and NH3-TPD
measurements. The samples showed peaks at around 1595
and 1445 cm−1 due to H-bonded pyridine and at 1580 and

154
L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
Table 2
Aminolysis of epoxides over SBA-15-pr-SO H and Ti-MCM-41
3
a
b
Run
No.
Epoxide
Amine
SBA-15-pr-SO H
Ti-MCM-41
3
c
c
Epoxide
conversion
(mol%)
Product selectivity (%)
TOF
Epoxide
conversion
(mol%)
Product selectivity (%)
TOF
−1
−1
(h
)
(h
)
A
B
A
B
1
2
3
4
5
6
6a
6b
7
Epichlorohydrin
Epichlorohydrin
Epichlorohydrin
Epichlorohydrin
Propene oxide
Styrene oxide
Aniline
92.6
58.0
87.5
36.3
59.7
38.1
100
87.0
87.0
92.6
100
89.8
0
165
103
156
65
76.6
83.6
80.8
92.2
86.3
80.5
100
85.0
98.0
90.8
81.0
93.8
0
15.0
2.0
9.2
19.0
6.2
161
175
170
194
181
169
m-Toluidine
n-Butyl amine
Dibutyl amine
Aniline
13.0
13.0
7.4
0
107
68
(24)
Aniline
10.2
(7.4)
d
d
d
d
(54.2)
(4.6)
(92.6)
e
e
e
(100)
94.4
100
88.2
100
85.0
100
(0)
Styrene oxide
Styrene oxide
Styrene oxide
Styrene oxide
Styrene oxide
Styrene oxide
Styrene oxide
Styrene oxide
Cyclohexene oxide
p-Chloroaniline
p-Anisidine
m-Toluidine
o-Toluidine
Cyclohexylamine
n-Butyl amine
Piperidine
34.0
15.0
32.3
19.1
10.0
7.2
89.1
44.6
23.8
5.6
0
11.8
0
15.0
0
0.6
5.2
0
61
27
58
34
18
13
159
80
70.1
50.0
72.9
68.8
12.6
3.6
54.4
51.3
32.7
94.2
96.2
94.5
95.8
92.4
100
95.1
93.7
5.8
3.8
5.5
4.2
7.6
0
4.9
6.3
0
147
105
153
144
26
8
9
10
11
12
13
14
15
7
99.4
94.8
100
114
108
68
Morpholine
Aniline
43
100
a
Reaction conditions: SBA-15-pr-SO H, 25 mg (-pr-SO H loading was 0.56 mmol/g); epoxide, 10 mmol; amine, 10 mmol; temperature, 298 K; reaction time,
3
3
4 h.
b
Reaction conditions: Ti-MCM-41 (Si/Ti = 35), 50 mg; epoxide, 20 mmol; epoxide:amine (molar ratio), 1:1; temperature, 308 K, reaction time, 4 h.
c
d
e
Turnover frequency (TOF) = moles of epoxide converted per mol of –SO H or Ti in the catalyst per hour.
3
Reaction conducted with 100 mg of the catalyst.
Reaction conducted in the absence of a catalyst.
Based on the titrimetric method, we found that the ion-
MCM-41 catalysts at ambient temperatures and under solvent-
free conditions (Table 2). Both steric hindrance and nucle-
ophilicity of the amine influenced the reactivity.
exchange capacity of this sample was 0.7 mEq/g catalyst,
which is in good agreement with the expectation from thermal
and elemental analyses of –SO3H groups. This suggests that
in SBA-15-pr-SO3H, almost all of the protons of sulfonic acid
can be exchanged with sodium ions. Titrimetric studies also
were conducted for thiol-functionalized SBA-15 (SBA-15-pr-
SH) and “bare” SBA-15. The ion-exchange capacities of these
materials were 0.2 and 0.04 mEq/g catalyst, respectively. Note
that the ion-exchange capacity of SBA-15-pr-SH was much
lower than the actual thiol content of the catalyst (Table 1);
thus, only a fraction of the surface SH protons were replaced
by Na+. Because the thiol group is a weaker acid than the sul-
fonic acid, the dissociation of the S–H protons and consequent
ion exchange with Na+ are lower for SBA-15-pr-SH than for
SBA-15-pr-SO3H.
NH3-TPD studies on titanosilicates revealed a desorption
peak at 448 K (Fig. 3d). This desorption was more intense
and asymmetric in Ti-MCM-41, indicating, in agreement with
the DRIFT studies, the differences in the greater strength of
the Lewis acid sites. This differences in accessibility and acid
strength of Ti sites in TS-1 and Ti-MCM-41 influence their cat-
alytic activity.
The reaction of styrene oxide with aniline yielded two types
of regio-isomers: A and B (Scheme 1). Selectivity for the A-
isomer was always higher (Table 2). The intrinsic catalytic
activity (turnover frequency [TOF]) of SBA-15-pr-SO3H was
higher than all of the hitherto known catalysts for this reac-
tion [8–16]. For example, for the reaction between styrene ox-
ide and aniline, TOF values over sulfamic acid, amberlyst-15,
Sc(OTf)3, and NaY catalysts were 32, 2, 7, and 0.8 h−1, respec-
tively [8–16]. The corresponding value for SBA-15-pr-SO3H
was found to be 68 h−1 (Table 2).
Very high conversion (92.6 mol%) and selectivity of product-
A (100 mol%) were obtained when epichlorohydrin instead
of styrene oxide was reacted with aniline (Table 2, run 1).
Despite the fact that epichlorohydrin has many reactive posi-
tions and can in principle lead to some other products, we did
not observe (by gas chromatography) products other than the
amino alcohols A and B. Moreover, the conversions estimated
based on the disappearance of either of the two reactants (epox-
ide/amine) were about the same. The reactivity (% conversion)
decreased in the following order: epichlorohydrin > propene
oxide > styrene oxide > cyclohexene oxide. Cyclohexene ox-
ide is a symmetrical epoxide. Reaction with aniline yielded cis-
and trans-diastereomeric products (Scheme 1). Only the trans-
3.3. Catalytic activity
1
diastereo isomer was formed (product characterization by H
3.3.1. Aminolysis of epoxides
β-Amino alcohols were synthesized by the ring opening of
epoxides with amines over SBA-15-pr-SO3H, TS-1 and Ti-
NMR study) with 100% selectivity over these catalysts (Ta-
ble 2, run 15).

L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
155
Table 3
ble 3, runs 4 and 5). On Ti-MCM-41, aminolysis with aro-
matic amines yielded higher conversion than aminolysis with
aliphatic n-butyl amine. The product yields were also signif-
icantly higher on terminal epoxides than on symmetrical cy-
cloalkene oxides. In general, the catalytic activity (TOF) of Ti-
MCM-41 catalysts was higher than that of SBA-15-pr-SO3H.
The use of solvents suppressed catalytic activity (Table 4).
Solvent molecules compete with the substrate molecules for ad-
sorption on the surface (refer to the discussion on adsorption
studies presented later). The conversion was higher in nonpolar
and weakly polar solvents (e.g., dichloromethane, toluene) than
in polar solvents (e.g., acetonitrile, ethyl acetate) (Table 4).
Temperature had a notable effect on catalytic activity but
only a marginal effect on selectivity (Table 5). Fig. 4 shows
the influence of reaction time and temperature on styrene oxide
conversion and product selectivity over Ti-MCM-41. The con-
version increased with reaction time, but the product selectivity
was unaffected. At near-ambient conditions, complete conver-
sion of styrene oxide and very high selectivity of the A-type
regioisomer (97.6 mol%) were obtained (Fig. 4i).
Influence of Ti structure: reaction of styrene with aniline over different cata-
a
lysts
Catalyst
Styrene oxide
conversion (%)
Product selectivity (%) TOF
−1 b
(h
)
A
B
Nil
TS-1
6.6
11.5
80.5
89.7
90.8
93.8
95.1
94.2
86.0
86.3
10.3
9.2
6.2 169
4.9 162
5.8 162
–
24
Ti-MCM-41
Ti-MCM-41, 1st recycle 77.4
Ti-MCM-41, 2nd recycle 77.0
TiO (2 mg)
9.1
48.0
14.0
13.7
18
4
2
TiO (50 mg)
2
a
Reaction conditions: catalyst (TS-1/Ti-MCM-41 (Si/Ti = 35)), 50 mg;
styrene oxide, 20 mmol; styrene oxide:aniline (molar ratio), 1:1; reaction time,
4 h; reaction temperature, 308 K.
b
Turnover frequency (TOF) = moles of epoxide converted per mol of Ti in
the catalyst per hour.
A range of β-amino alcohols was produced with high regios-
electivity from styrene oxide and different amines. The reactiv-
ity (epoxide conversion and TOF) depended on the nature and
type of the amine molecule, however. The product yield was
low when cyclohexyl amine instead of aniline was used. High
conversions were obtained with aliphatic amines like n-butyl
amine (Table 2, run 3). With the secondary amine piperidine,
conversion of 89.1 mol% and product isomer-A selectivity of
99.4 mol% were obtained (Table 2, run 13). However, with
morpholine, which contains an additional oxygen atom in the
hexa-cyclic saturated ring, the conversion was only 44.6 mol%
(run 14).
3.3.2. Alcoholysis of epoxides
SBA-15-pr-SO3H demonstrated catalytic activity for the re-
action of styrene oxide with different alcohols at ambient tem-
perature and solvent-free conditions (Table 6). Selectivity of
product A (Scheme 1) was 100%. The intrinsic catalytic ac-
tivity of SBA-15-pr-SO3H was significantly higher compared
with the hitherto known catalysts [17]. Both products A and B
were detected over Ti-MCM-41, with A being the major one.
We also conducted the reaction of styrene oxide with aniline
at 298 K in the presence of thiol-functionalized SBA (SBA-
15-pr-SH) and “bare” SBA-15. Conversions of styrene oxide
in those experiments were only 29 and 24.6 mol%, in contrast
to 38.1 mol% in the presence of SBA-15-pr-SO3H (Table 2,
run 6). These variations in catalytic activity follow changes
in the ion-exchange capacities/acid strengths of these SBA-15
materials. Whereas isomers A and B (with product selectivity
of 89.8 and 10.2 mol%, respectively) formed over SBA-15-
pr-SO3H, only the isomer-B formed over SBA-15-pr-SH and
SBA-15.
Over Ti-MCM-41 catalysts, a styrene oxide conversion of
80 mol% and isomer-A selectivity of 93.8 mol% were obtained
at 308 K (Table 2, run 3) [19]. At similar reaction conditions,
TS-1 showed only 11.5 mol% conversion and 90.8 mol% of
A-isomer product selectivity (Table 2, run 2). Due to diffusion
limitations, the reaction over TS-1 occurs mainly at the outer
surface of particles [19].
3.4. Kinetic studies
The kinetic rate constants and apparent activation energies
for the aminolysis of styrene oxide with aniline over SBA-
15-pr-SO3H and Ti-MCM-41 catalysts were determined using
different molar ratios of the reactants and at different tempera-
tures. The reaction was first order with respect to both styrene
oxide and aniline. The apparent activation energies (Ea) were
calculated using the Arrhenius equation, k = Aexp(−Ea/RT ).
These values for both solid acid catalysts are listed in Table 7.
The rate constants were higher and the activation energy was
lower over Ti-MCM-41 than over SBA-15-pr-SO3H. Brönsted-
acidic SBA-15-pr-SO3H possibly protonates the amines instead
of activating the epoxide molecules. Thus, the Lewis acid (Ti-
MCM-41) is likely more active.
3.5. Mechanistic considerations
We also carried out experiments with anatase-TiO2. The Ti
content in the sample was kept the same as that used in the ex-
periments with Ti-MCM-41. The conversion was much lower
over anatase-TiO2 (9.1 mol%) than on Ti-MCM-41 (Table 3;
compare runs 6 and 3). These findings indicate that dispersion
and surface accessibility of the Ti4+ ions are important in de-
termining the catalytic activity of titanosilicate materials.
To check the recyclability of the catalyst, the recovered cat-
alyst was washed first with dichloromethane and then with
methanol, then dried at 373 K for 2 h before reuse (Ta-
Aminolysis of epoxides is a bimolecular reaction. An addi-
tional factor, not present in homogeneous catalysis, can influ-
ence the rate and mechanism of the reaction in catalysis over
solid catalysts: the competitive access of the epoxide and amine
molecules to the acidic, active sites on the surface. The different
basicities of the amine and epoxide can influence their relative
surface coverages and, consequently, the reaction rates.
Adsorption of one substrate (epoxide or amine), as well as
competitive adsorption of both, were investigated. In the ad-

156
L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
Table 4
a
Influence of solvent on the adsorption characteristics and reaction of styrene oxide with aniline over SBA-15-pr-SO H and Ti-MCM-41
3
Solvent
SBA-15-pr-SO H
Ti-MCM-41
3
(dielectric
constant)
One substrate adsorption
(mmol/g catalyst)
Epoxide
conversion
(mol%)
Product selectivity
(mol%)
TOF
Epoxide
conversion
(mol%)
Product selectivity
(mol%)
TOF
b
−1 c
−1 c
(h
)
(h
)
Styrene oxide
Aniline
A
B
A
B
No solvent
0.20
0.20
0.21
0.21
38.1
33.7
89.8
62.0
10.2
38.0
68
30
99.5
77.6
71.2
81.0
25.0
27.7
97.6
99.2
98.8
98.8
98.6
97.6
2.4
0.8
1.2
1.2
1.4
2.4
52
8
7
9
3
Dichloromethane (9.1)
Carbon tetrachloride (2.2)
Toluene (2.4)
Acetonitrile (37.5)
Ethyl acetate (6.0)
0.17
0.06
0.04
0.20
0.13
0.11
26.4
4.1
3.9
100
100
100
0
0
0
24
4
3
3
a
Reaction conditions: SBA-15-pr-SO H (-pr-SO H loading is 0.56 mmol/g); catalyst, 25 mg; styrene oxide, 5 mmol (with solvent) or 10 mmol (without solvent);
3
3
reaction temperature, 298 K. Ti-MCM-41 (Si/Ti = 35); catalyst, 50 mg; styrene oxide, 1 mmol (with solvent) or 5 mmol (without solvent); reaction temperature,
303 K. Solvent, 0 or 5 mL; styrene oxide:aniline (molar ratio), 1:1; reaction time, 4 h.
b
Activated (393 K for 4 h) SBA-15-pr-SO H (50 mg) was dispersed, for 1 h, in 0.5 mmol of styrene oxide or aniline substrate dissolved in 5 mL of solvent. The
3
solid catalyst was, then, separated and the concentration of the substrate in the liquid portion was determined by gas chromatography. The amount adsorbed on the
catalyst surface was determined by difference.
c
Turnover frequency (TOF) = moles of epoxide converted per mol of –SO H or Ti in the catalyst per hour.
3
Table 5
Influence of temperature on the reaction of styrene oxide with aniline over SBA-15-pr-SO H and Ti-MCM-41
a
3
a
b
SBA-15-pr-SO H
Ti-MCM-41
3
Reaction
temperature
(K)
Epoxide
conversion
(%)
Product selectivity (%)
TOF
Reaction
temperature
(K)
Epoxide
conversion
(%)
Product selectivity (%)
TOF
−1 c
−1 c
(h
)
(h
)
A
B
A
B
298
308
323
38.1
68.5
88.7
91.3 (9.0)
89.8
86.3
87.1
83.6 (83.9)
10.2
13.7
12.9
16.4 (16.1)
68
272
293
303
308
323
343
20.2
31.4
50.9
77.1
88.0
97.6
96.0
96.2
95.3
94.4
93.5
92.3
4.0
3.8
4.7
5.6
6.5
7.7
85
122
158
163
132
214
324
370
410
d
343
a
Reaction conditions: catalyst, 25 mg; styrene oxide, 10 mmol; styrene oxide:aniline (molar ratio), 1:1; reaction time, 4 h.
Reaction conditions: catalyst, 50 mg; styrene oxide, 20 mmol; styrene oxide:aniline (molar ratio), 1:1; reaction time, 2 h.
b
c
d
Turnover frequency (TOF) = moles of epoxide converted per mol of –SO H or Ti in the catalyst per hour.
3
Values in parentheses are those with no catalyst.
Fig. 4. Aminolysis of styrene oxide. (i) Influence of reaction time: reaction conditions—styrene oxide, 40 mmol; aniline, 40 mmol; Ti-MCM-41, 100 mg; tempera-
ture, 308 K. (ii) Influence of temperature: reaction conditions—styrene oxide, 20 mmol; aniline, 20 mmol; Ti-MCM-41, 50 mg. (iii) Reaction over SBA-15-pr-SO H.
3
Influence of reaction time: reaction conditions—styrene oxide, 20 mmol; aniline, 20 mmol; SBA-15-pr-SO H, 50 mg.
3

L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
157
Table 6
Alcoholysis of styrene oxide over SBA-15-pr-SO H and Ti-MCM-41
a
3
Run
No.
Alcohol
SBA-15-pr-SO H
Ti-MCM-41
3
Styrene oxide
conversion
(mol%)
Selectivity of
product A
TOF
Styrene oxide
conversion
(mol%)
Product selectivity
(mol%)
TOF
−1 b
−1 b
(h
)
(h
)
A
B
1
2
3
4
5
6
Methanol
Ethanol
Propanol
n-Butanol
Benzyl alcohol
Cyclohexanol
25.6
16.8
10.4
9.2
6.2
7.2
100
100
100
100
100
100
23
15
9
8
5
27.8
15.2
12.4
11.1
88.0
87.3
92.8
92.9
12.0
12.7
7.2
29
16
13
11
7.1
6
8.1
78.9
21.1
8
a
Reaction conditions: SBA-15-pr-SO H: catalyst, 25 mg; styrene oxide and alcohol, 5 mmol each; temperature, 298 K; reaction time, 4 h. Ti-MCM-41: catalyst,
3
50 mg; styrene oxide, 10 mmol; alcohol, 20 mmol; temperature, 333 K; reaction time, 4 h.
b
Turnover frequency (TOF) = moles of styrene oxide converted per mol of –SO H or Ti in the catalyst per hour.
3
Table 7
Kinetic data for aminolysis of styrene oxide with aniline over SBA-15-pr-
a
SO H and Ti-MCM-41 catalysts
3
Catalyst
Reaction tem-
perature (K)
Rate constant
Activation energy
−1 −1
−1
(L mol
s
)
(E ) (kJ mol
)
a
−4
−3
−3
−3
SBA-15-pr-SO H
298
308
313
318
7.96 × 10
2.70 × 10
2.75 × 10
3.58 × 10
59
3
−3
−3
−3
Ti-MCM-41
a
313
318
323
7.31 × 10
9.15 × 10
11.31 × 10
37
Reactions were conducted taking 1:1 molar ratio of styrene oxide to aniline.
Amount of the catalyst: 25 mg (SBA-15-pr-SO H) or 50 mg (Ti-MCM-41).
3
sorption studies of one substrate, a known quantity of SBA-15-
pr-SO3H (50 mg) was initially activated at 393 K for 4 h and
then dispersed at 298 K in a known concentration of the sub-
strate (0.5 mmol) dissolved in a solvent (5 mL). After an hour,
the concentration of the substrate in the liquid portion was es-
timated by gas chromatography, and the amount adsorbed was
determined. When dichloromethane was used as a solvent, the
amount of styrene oxide and aniline adsorbed on the catalyst
surface were found to be 0.20 and 0.21 mmol/g catalyst, re-
spectively (Table 4). Similar values were obtained when toluene
was used as a solvent (Table 4). However, in the polar, apro-
tic acetonitrile, the amounts of epoxide as well as the amine
adsorbed on the catalyst were considerably lower (0.06 and
0.13 mmol/g catalyst, respectively). Polar solvents compete
with the substrate molecules for adsorption (on sulfonic acid
groups) and thereby lower catalytic activity. Note that there ex-
ists an inverse correlation between the dielectric constant of
the solvent and catalytic activity over Ti-MCM-41 (Table 4).
In further studies, dichloromethane was selected as the solvent
of choice, because it did not compete appreciably with the sub-
strate molecules for adsorption on the catalyst surface (Table 4).
The influence of substrate adsorption on proton-exchange
capacity of the catalyst was investigated by the titrimetric
method (Section 2.4). When different amounts (0.25–1.5 mmol)
of styrene oxide or aniline were adsorbed on the sulfonic acid-
functionalized SBA-15, the ion-exchange capacity of the cata-
Fig. 5. Change in ion exchange capacity with the amount of preadsorbed epox-
ide or amine on SBA-15-pr-SO H.
3
lyst decreased from 0.73 mEq/g for SBA-15-pr-SO3H to about
0.3 mEq/g for the epoxide- or amine-adsorbed catalyst (Fig. 5).
Both the epoxide and aniline adsorb on the sulfonic acid func-
tionality.
In some experiments, aniline was preadsorbed on SBA-15-
pr-SO3H. Styrene oxide was then adsorbed. The catalytic ac-
tivities of these aniline-preadsorbed catalysts were also deter-
mined simultaneously. Epoxide adsorption and catalytic activ-
ity decreased with increasing amounts of preadsorbed amine.
Epoxide adsorption on the acid site was apparently essential for
the reaction (Fig. 6a). In another set of experiments, different
amounts of styrene oxide were initially added onto the SBA-
15-pr-SO3H surface. Thereafter, the amine adsorption capacity
and catalytic activity of those modified catalysts were deter-
mined. The amount of aniline adsorption and catalytic activity
increased with increasing amount of epoxide added (Fig. 6b).
Aniline was consumed due to reaction with the epoxide already
activated on the acidic site, yielding the product β-amino al-
cohol. There is a correlation between styrene oxide conversion
and the amount of epoxide preadsorbed (Fig. 6c). The amounts

158
L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
Fig. 6. SBA-15-pr-SO H. (a) Variation in styrene oxide adsorption and conversion with preadsorbed aniline. (b) Variation in aniline adsorption and styrene oxide
3
conversion with preadsorbed styrene oxide. (c) Influence of preadsorbed amine on epoxide adsorption and catalytic activity in the reaction of styrene oxide (SO)
with aniline.
Table 8
Epoxide and amine adsorptions on SBA-15-pr-SO H and Ti-MCM-41 catalysts
a
3
Material
Amount of epoxide adsorbed (mmol/g)
Amount of amine adsorbed (mmol/g)
Epichloro-
hydrin
Styrene
oxide
Cyclo-hexene
oxide
Piperidine
Aniline
4-Cl-aniline
SBA-15
SBA-15-pr-SO H
Ti-MCM-41
0.09
0.17
0.42
0.04
0.20
1.25
0.17
0.34
0.43
0.02
0.03
0.77
0.01
0.21
0.55
0.09
0.23
1.41
3
a
50 mg of activated SBA-15 material was dispersed, for 1 h, in 0.5 mmol of styrene oxide or aniline substrate dissolved in 5 mL of solvent. The solid was, then,
separated and the concentration of the substrate in the liquid portion was determined by gas chromatography. The amount adsorbed on the catalyst surface was
determined by difference. 100 mg of Ti-MCM-41 was dispersed in 1 mmol of styrene oxide or aniline dissolved in 10 mL of CH Cl .
2
2
of different epoxides or amines adsorbed per gram of SBA-15-
pr-SO3H (in one-substrate adsorption experiments) are given
in Table 8. Adsorptions were also done on “bare” SBA-15. The
amount of substrate adsorbed on the catalyst surface varied with
the nature of the substrate molecule. Diffusion constraints were
not expected over these mesoporous catalysts. Both epoxides
and amines adsorbed even on “bare” SBA-15, the former more
than the latter (Table 8). Functionalization of SBA-15 surface
with propylsulfonic acid moieties enhanced the adsorption of
epoxide as well as amine substrates, the latter more than the
former (Table 8). The amount of substrate adsorbed increased
with its increasing concentration in the solution; however, it
was always lower than the molar equivalent of propylsulfonic
acid moieties present on the catalyst surface. Surprisingly, the
amount of epoxide or amine adsorbed was significantly greater
on the weakly Lewis-acidic Ti-MCM-41 than on SBA-15-pr-
SO3H (Table 8). The larger concentration of acidic, surface Ti
atoms (compared with the number of grafted –SO3H groups),
rather than the strength of acidity, is probably responsible for
the greater adsorption capacity as well as catalytic activity of
T-MCM-41.
amines [46], we also conducted competitive adsorption ex-
periments wherein the sulfonic acid-functionalized SBA-15
(50 mg) was suspended in equimolar (0.5 mmol) amounts of
epoxide (cyclohexene oxide, epichlorohydrin, or styrene ox-
ide) and amine (aniline or o-toluidine) dissolved in 5 mL of
dichloromethane. The adsorption of both the epoxide and amine
were estimated; Table 9 compares the relative adsorption of the
various epoxides and amines and also gives the corresponding
TOFs. The importance of carrying out competitive adsorption
of the reactants (in addition to their independent adsorption
data) can be demonstrated by comparing the data in Tables 8
and 9. In independent adsorption experiments, the amount of
epoxides adsorbed on SBA-15-pr-SO3H varied as epichlorohy-
drin (0.17) < styrene oxide (0.20) < cyclohexene oxide (0.34)
(Table 8). However, in the presence of an amine (aniline),
a reverse variation was observed: epichlorohydrin > styrene
oxide > cyclohexene oxide (relative adsorption ratios of 1.8,
1.1, and 0.8, respectively) (Table 9). The corresponding relative
rates (TOFs of 165, 68, and 43 h−1) correlate better with data
from the competitive adsorption experiments and illustrate the
perils involved in drawing conclusions from adsorption of in-
dividual reactants without taking into account the competition
from the others or the solvent. The conversion vs time plots for
Because adsorption is an essential prerequisite in many of
the liquid-phase bimolecular reactions between epoxides and

L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
159
Table 9
Competitive adsorption of epoxide and amine over SBA-15-pr-SO H
sible. In almost all of our studies (Tables 2–6), the A isomer
predominated. Substituent groups on the epoxide (like R in
Scheme 1) influenced the orientation of the incoming moiety
(RꢀNH and RꢀOH) by steric as well as electronic effects [46];
bulky R groups favored the B isomer. Under basic or neutral
conditions, when a SN2 attack of the reagent molecule (like
RꢀNH2 or RꢀOH) on the epoxide ring carbon atom led to prod-
ucts, the B isomer was dominant. Under acidic conditions, on
the other hand, electronic effects (inductive and conjugative)
determine the orientation [46]. These electronic effects can sta-
bilize a positive charge on the carbon atom (of the epoxide ring)
adjacent to R. This is facilitated if the reaction occurs by a SN1
mechanism in which the transition state for the rate-determining
step carries a partial positive charge on the carbon atom adja-
cent to R. Protonation of the epoxide oxygen (by the –SO3H
group over SBA-15-pr-SO3H) or its coordination to Ti4+ (in
Ti-MCM-41) apparently favors the SN1 mechanism on both of
these acidic catalysts and the consequent predominance of the
A isomer. This picture is supported by our results over SBA-15-
pr-SH and SBA-15, mentioned earlier, wherein the B isomer
predominated over the A isomer. Propylthiol and silanol are
weakly acidic and near-neutral functional groups, respectively,
compared with propylsulfonic acid (refer to the ion-exchange
capacities in Section 3.2.2); thus, the significant formation of
the B isomer over SBA-15-pr-SH and bare SBA-15 is under-
standable.
a
3
Adsorbate substrates
Amount adsorbed Relative
TOF
−1 b
(mmol/g
adsorption ratio:
(h
)
catalyst)
epoxide/amine
Epoxide Amine
Epichlorohydrin + aniline
Styrene oxide + aniline
0.10
0.08
0.06
0.07
0.13
0.04
1.8
1.1
0.8
0.8
165
68
43
Cyclohexene oxide + aniline 0.10
Styrene oxide + o-toluidine
0.03
34
a
50 mg of sulfonic acid functionalized SBA-15 was suspended for 1 h in
equimolar amounts (0.5 mmol) of epoxide and amine dissolved in 5 mL of
dichloromethane. The catalyst was, then, separated and the concentration of
the substrate in the liquid portion was determined by gas chromatography. The
amount adsorbed on the catalyst surface was determined by difference.
b
Turnover frequency (TOF) = moles of epoxide converted per mol of
–SO H in the catalyst per hour.
3
both Ti-MCM-41 and SBA-15-pr-SO3H were of sigmoid shape
(Fig. 4), further confirming that the mechanism involves a com-
petition between reactants for adsorption.
Fig. 7 further illustrates the competition between the epoxide
(styrene oxide) and the amine or alcohol for the active site. The
rate of the reaction decreased when the basicity of the amine
or alcohol increased. These findings from the reaction of epox-
ides with amines are similar to our earlier results [32,33] in the
reaction of epoxides with carbon dioxide to yield cyclic carbon-
ates over amine-functionalized Ti-SBA-15. We had concluded
that whereas CO2 molecules are activated at basic amine sites,
Lewis-acidic Ti4+ ions are the sites for adsorption and activa-
tion of the epoxide molecules [32,33]. The decreased styrene
oxide conversion observed in high-dielectric constant solvents
(Table 4) also supports the above picture. Highly polar sol-
vents compete with the epoxide for adsorption on the acidic
site. Thus, it is not surprising that the highest conversions were
observed in the absence of any solvent (Table 4, row 1). As
shown in Scheme 1, two regioisomers are theoretically pos-
4. Conclusion
Here the aminolysis of epoxides over Brönsted-acidic SBA-
15 functionalized with propylsulfonic acid and Lewis-acidic
Ti-MCM-41 is reported for the first time. These mesoporous,
solid acid catalysts are highly active and selective at ambient
temperatures and solvent-free conditions. A range of β-amino
alcohols were synthesized with high regioselectivity and stere-
oselectivity by reacting symmetric or unsymmetric epoxides
Fig. 7. Correlation between catalytic activity (turnover frequency; TOF) and pK of amine in aminolysis of styrene oxide over (a) SBA-15-pr-SO H and (b)
a
3
Ti-MCM-41. (c) Correlation between TOF and pK of alcohols in alcoholysis of styrene oxide over SBA-15-pr-SO H.
a
3

160
L. Saikia et al. / Journal of Catalysis 252 (2007) 148–160
with aromatic and aliphatic amines. Both of these catalysts ex-
hibited significantly higher catalytic activity than the hitherto
known acid catalysts for these reactions. Ti-MCM-41 was the
more reactive of the two. Both epoxide and amine competed for
adsorption on surface acid sites. Polar solvent molecules, when
present, also competed for adsorption with the reactant mole-
cules.
[19] J.K. Satyarthi, L. Saikia, D. Srinivas, P. Ratnasamy, Appl. Catal. A
Gen. 330 (2007) 145.
[20] J.A. Melero, R. van Grieken, G. Morales, Chem. Rev. 106 (2006) 3790.
[21] A. Vinu, K.Z. Hossain, K. Agira, Nanosci. Nanotechnol. 5 (2005) 347.
[22] I.K. Mbraka, D.R. Radu, V.S.Y. Lin, B.H. Shanks, J. Catal. 219 (2003)
329.
[23] W.D. Bossaert, D.E. De Vos, W.M. Van Rhijin, J. Bullen, P.J. Grobet, P.A.
Jacobs, J. Catal. 182 (1999) 156.
[24] I. Diaz, C. Márquez-Alvarez, F. Mohino, J. Pérez-Pariente, E. Saste, Mi-
croporous Mesoporous Mater. 44–45 (2001) 295.
[25] B.C. Wilson, C.W. Jones, Macromolecules 37 (2004) 9709.
[26] L. Saikia, D. Srinivas, P. Ratnasamy, Appl. Catal. A Gen. 309 (2006) 144.
[27] L. Saikia, D. Srinivas, P. Ratnasamy, Microporous Mesoporous Mater. 104
(2007) 225.
Acknowledgments
L.S. and J.K.S. acknowledge the Council of Scientific and
Industrial Research (CSIR), New Delhi for a fellowship award.
[28] P. Ratnasamy, D. Srinivas, H. Knözinger, Adv. Catal. 48 (2004) 1.
[29] B. Notari, Adv. Catal. 41 (1996) 253–334.
References
[30] R. Srivastava, D. Srinivas, P. Ratnasamy, Catal. Lett. 91 (2003) 133.
[31] R. Srivastava, D. Srinivas, P. Ratnasamy, Stud. Surf. Sci. Catal. C 154
(2004) 2703.
[32] R. Srivastava, D. Srinivas, P. Ratnasamy, J. Catal. 233 (2005) 1.
[33] R. Srivastava, D. Srinivas, P. Ratnasamy, Microporous Mesoporous
Mater. 90 (2006) 314.
[34] D. Srinivas, R. Srivastava, P. Ratnasamy, Catal. Today 93 (2004) 127.
[35] R. Garro, M.T. Navarro, J. Primo, A. Corma, J. Catal. 233 (2005) 342.
[36] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka,
G.D. Stucky, Science 279 (1998) 548.
[37] F. Feng, G.E. Fryxell, L.-Q. Wang, A.Y. Kim, K.M. Kemner, Science 276
(1997) 923.
[38] L. Mercier, T.J. Pinnavaia, Adv. Mater. 9 (1997) 500.
[39] A. Stein, B.J. Melde, R.C. Schroden, Adv. Mater. 12 (2000) 1403.
[40] W.M. Van Rhijn, D.E. De Vos, B.F. Sels, W.D. Bossaert, P.A. Jacobs,
Chem. Commun. (1998) 317.
[41] D. Das, J.-F. Lee, S. Cheng, Chem. Commun. (2001) 2178.
[42] E. Cano-Serrano, J.M. Campos-Martin, J.L.G. Fierro, Chem. Commun.
(2003) 246.
[43] A. Thangaraj, R. Kumar, P. Ratnasamy, J. Catal. 131 (1991) 294.
[44] D. Margolese, J.A. Melero, S.C. Christiansen, B.F. Chmelka, G.D. Stucky,
Chem. Mater. 12 (2000) 2448.
[45] B. Sow, S. Hamoudi, M.H. Zahedi-Niaki, S. Kaliaguine, Microporous
Mesoporous Mater. 79 (2005) 129.
[1] O. Mitsunobu, in: B.M. Trost, I. Fleming (Eds.), Comprehensive Organic
Synthesis, vol. 6, Pergamon Press, New York, 1991, p. 88.
[2] E.J. Corey, F.-Y. Zhang, Angew. Chem. Int. Ed. Engl. 38 (1999) 1931.
[3] D.J. Ager, I. Prakash, D.R. Schaad, Chem. Rev. 96 (1996) 835.
[4] P. O’Brien, Angew. Chem. Int. Ed. Engl. 38 (1999) 326.
[5] G. Li, H.-T. Chang, K.B. Sharpless, Angew. Chem. Int. Ed. Engl. 35
(1996) 451.
[6] R.M. Hanson, Chem. Rev. 91 (1991) 437.
[7] P.A. Crooks, R. Szyudler, Chem. Ind. (London) (1973) 1111.
[8] A. Kamal, B. Rajendra Prasad, A. Malla Reddy, M. Naseer, A. Khan,
Catal. Commun. 8 (2007) 1876.
[9] M. Vijender, P. Kishore, P. Narender, B. Satyanarayana, J. Mol. Catal. A
Chem. 266 (2007) 290.
[10] A.T. Placzek, J.L. Donelson, R. Trivedi, R.A. Gibbs, S.K. De, Tetrahedron
Lett. 46 (2005) 9029.
[11] J. Iqbal, A. Pandey, Tetrahedron Lett. 31 (1990) 575.
[12] R.I. Kureshy, S. Singh, N.H. Khan, S.H.R. Abdi, E. Suresh, R.V. Jasra,
J. Mol. Catal. A Chem. 264 (2007) 162.
[13] F. Carrée, R. Gil, J. Collin, Org. Lett. 7 (2005) 1023.
[14] J.S. Yadav, B.V.S. Reddy, A.K. Basak, A.V. Narasaiah, Tetrahedron
Lett. 44 (2003) 1047.
[15] P.-Q. Zhao, L.-W. Xu, C.G. Xia, Synlett (2004) 846 and 1122.
[16] N. Azizi, M.R. Saidi, Org. Lett. 7 (2005) 3649.
[17] R. Garcia, M. Martinez, J. Aracil, Chem. Eng. Technol. 22 (1999) 12.
[18] M.W.C. Robinson, R. Buckle, I. Mabbett, G.M. Grant, A.E. Graham,
Tetrahedron Lett. 48 (2007) 4723.
[46] R.E. Parker, N.S. Isaacs, Chem. Rev. 59 (1959) 737.