Journal of Catalysis 251 (2007) 249–257

www.elsevier.com/locate/jcat

Investigation of pore-size effects on base catalysis using

amino-functionalized monodispersed mesoporous silica spheres
as a model catalyst
Tomiko M. Suzuki ∗ , Masami Yamamoto, Keiko Fukumoto, Yusuke Akimoto, Kazuhisa Yano ∗
Toyota Central Research & Development Labs. Inc., Nagakute, Aichi 480-1192, Japan
Received 11 May 2007; revised 10 August 2007; accepted 11 August 2007

Abstract
The effects of pore size on base catalysis have been studied using amino-functionalized monodispersed mesoporous silica spheres (NH2 -
MMSS). By changing the surfactant that is used for the template and also the synthetic conditions, NH2 -MMSS with similar particle diameters
(560–600 nm) and different pore sizes (0–2.66 nm) were successfully prepared. Because the NH2 -MMSS that were obtained had the same shapes,
particle diameters and mesopore alignments, any differences in catalytic activity could be substantially attributed to differences in their pore sizes.
We employed the nitroaldol condensation between benzaldehyde derivatives and nitromethane as a base catalysis reaction. It was found for the first
time that the reaction mostly proceeds inside the mesopores (the effectiveness factor: 0.63) and that the optimum pore size for the NH2 -MMSS
was affected by changing the type and number of the substituent groups on the reactants. In addition, NH2 -MMSS was found to be an excellent
catalyst due to the radial alignment of the mesopores compared to the other types of mesoporous silica.
© 2007 Elsevier Inc. All rights reserved.

Keywords: Mesoporous silica; Monodispersed spheres; Amino-functionalized; Pore size; Base catalyst

1. Introduction of various MCM-41 with different pore sizes for the acetal-
ization of cyclohexanone with methanol. They showed that the
In recent years, various ordered mesoporous materials have catalytic activity was strongly dependent on the pore size of the
been synthesized by using surfactants as templates. Meso- catalysts, and was maximized when the pore size was around
porous silica materials with uniform pore sizes, high surface 1.9 nm. Because the diameters of cyclohexanone and the prod-
areas and high concentrations of surface hydroxyl groups are of uct were 0.75 and 0.96 nm, respectively, the phenomenon could
considerable research interest in fields such as catalysis, adsorp- not be explained by the concept of shape-selective catalysis.
tion, separation, ion exchange and sensor design. In particular, They also reported that the catalytic activity was not dependent
their application as heterogeneous catalysts is of great interest on the molecular size of the reactant, but on the pore size of the
from the view point of “green” chemistry. In general, hetero- catalyst [2].
geneous catalysts have advantages in terms of separation and Tanchoux et al. demonstrated that a sharp increase in cat-
regeneration from reaction solutions. alytic activity for the double bond isomerization of 1-hexene
There have been several investigations into the effect of pore was observed for MCM-41 with a pore size of 3.4 nm. They
size on catalytic activity using mesoporous silica with irregu- reported that this result was caused by a compensation effect be-
lar shapes [1–3]. Iwamoto et al. studied the catalytic activity tween diffusion limitations and confinement activity enhance-
ment [3]. It became clear that MCM-41 had the optimum pore
* Corresponding authors. Fax: +81 561 63 6507. size in acid catalytic reactions.
E-mail addresses: [email protected] (T.M. Suzuki), However, because the mesoporous silicas that have previ-
[email protected] (K. Yano). ously been used as catalysts have had irregular shapes and pore
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.08.010
250 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257

arrangements, it is quite difficult to accurately evaluate the ef- The amino-moieties were incorporated by a grafting method,
fect of pore size on the catalytic activity without any influence in which organically functional groups were attached onto the
from other effects. In order to accurately study the effect of the silica surface via a reaction between the hydroxyl groups on the
pore size of the mesoporous silica catalyst on catalytic activity, surface of the mesoporous silica and silane compounds.
it is essential to use porous materials that have the same shapes, 3-Aminopropyltrimethoxysilane (APTMS), 4-hydroxybenz-
sizes and pore arrangements. aldehyde, 4-butoxybenzaldehyde, 4-n-octoxybenzaldehyde,
In our laboratory, we have succeeded in the synthesis of 4-methoxybenzaldehyde, 2,4,6-trimethoxybenzaldehyde and
hexagonally ordered and well-defined highly monodispersed 3,4,5-trimethoxybenzaldehyde were purchased from Aldrich.
mesoporous silica spheres (hereafter abbreviated as “MMSS”) Decyltrimethylammonium bromide (C10 TMABr), tetradecyl-
from tetramethoxysilane (TMOS) and n-alkyltrimethylammo- trimethylammonium chloride (C14 TMACl), hexadecyltrimeth-
nium halide (Cn TMAX, X = Cl, Br) [4–6]. Because these ylammonium chloride (C16 TMACl), octadecyltrimethylammo-
MMSS have radially aligned mesopores [6], it is expected that nium chloride (C18 TMACl), and tetramethoxysilane (TMOS)
accessibility for reactants and the release of products would be were purchased from Tokyo Kasei. Dococyltrimethylammo-
enhanced if they were used as catalysts. nium chloride (C22 TMACl) was purchased from Lion Akzo
Recently, aminopropyl-functionalized mesoporous silica Co., Ltd. 1 M sodium hydroxide solution, methanol, hy-
materials such as MCM, FSM, and SBA have been reported drochloric acid, dry toluene, nitromethane, tetrahydrofuran
to be effective as base catalysts for Knoevenagel condensa- (THF), acetone-d6 and decane were purchased from Wako
tions [7–10], aldol condensations [11–14], Michael additions Inc. Monodispersed silica spheres were purchased from Nip-
[10,11,15] and epoxidation reactions [16]. pon Shokubai Co., Ltd. All of the chemicals were used as-re-
We synthesized MMSS that were particles with almost the ceived.
same diameters and different pore sizes. These were function- The syntheses were carried out under basic conditions, as
alized as base catalysts by incorporating an aminopropyl group follows.
using a grafting method. The purpose of this work is to under- For example, in the case of C16 -MS, 7.04 g of C16 TMACl
stand the “pure” effect of pore size on base catalysis when using and 6.84 ml of 1 M sodium hydroxide solution were dis-
amino-functionalized MMSS as a model catalyst. solved in 1600 g of a methanol/water (50/50 = w/w) solution
When a reaction is conducted with micro/meso-porous cat- (methanol ratio: 0.5). 5.28 g (34.7 mmol) of TMOS was added
alysts, diffusion could govern the reaction because the sizes of to the solution with vigorous stirring at 298 K. After the addi-
reactants and pores are relatively very close [17]. To date, there tion of the TMOS, the clear solution suddenly turned opaque
have been many reports using mesoporous silica as a catalyst. and resulted in a white precipitate. After 8 h of continuous stir-
However, in many cases, it has been expected that only a part ring, the mixture was aged overnight. The white powder was
of mesopores works as a catalyst due to diffusion limitation. then filtered and washed three times with distilled water, after
Therefore, it is very interesting to know the effectiveness fac- which it was dried at 318 K for 72 h. The powder that was ob-
tor of mesoporous silica as a catalyst. This can be only realized tained was calcined in air at 823 K for 6 h to remove the organic
by using MMSS with the same pore size and different particle species.
diameter as a catalyst. In the case of pore-expanded MMSS (C18 PS-MS), 2 g of
We have synthesized five types of amino-functionalized C18 -MS before calcination was suspended in 120 g of a 0.1 M
MMSS with almost the same particle diameters and different C22 TMACl water–ethanol solution. The mixture was sealed and
pore sizes by changing the template surfactant and the syn- placed in an oven at 353 K for 7 days without stirring. The
thetic conditions, and their base catalytic performances were resulting white powder was filtered and washed with distilled
evaluated. It was found that the optimum pore size changed de- water and then dried. The powder that was obtained was cal-
pending on the molecular structure of the reactant. We have also cined in air at 823 K for 6 h to remove the organic species.
conducted a reaction with amino-functionalized MMSS with Amino moiety grafting was performed on calcined and evac-
the same pore size and different particle diameters. The effec- uated (323 K) samples of MMSS, on FSM [18] and on non-
tiveness factor of spherical mesoporous silica was evaluated for porous monodispersed silica spheres (SS).
the first time. For example, in the case of C16 -MS-AP, 0.7 g of calcined
C16 -MS was added to a mixture of 70 g of dehydrated toluene
and 1.88 g of APTMS, and the solution was then refluxed at
2. Experimental
363 K for 15 h. The product was dried overnight at 318 K after
filtration.
2.1. Chemicals and catalyst synthesis Hereafter in this paper, the samples that were obtained are
abbreviated as Cx PS-MS, where x is 10, 14, 16 or 18. x de-
We synthesized MMSS with the same particle diameters notes the number of carbon atoms of alkyl chain in the sur-
and different pore sizes by using n-alkyltrimethylammonium factant that was used as a templating agent, and PS indicates
halide (Cn TMAX: n = 10–18, X = Cl, Br) as a templating that the pore size of the material was expanded by the sur-
agent and a tetramethoxysilane (TMOS) as a silica source [5,6]. factant exchange method [19]. Amino-functionalized materi-
The particle sizes of the MMSS were adjusted by changing the als are denoted by attaching “AP” to the end of the sample
methanol/water ratio in the solvent. names.
 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257 251

2.2. Sample characterization

Powder X-ray diffraction measurements were carried out

on a Rigaku Rint-2200 X-ray diffractometer using CuKα ra-
diation. Scanning electron micrographs (SEMs) were obtained
using a Sigma-V (Akashi Seisakusho). The surfaces of the sam-
ples were coated with gold before the measurements. The av-
erage particle diameter was calculated from the diameters of
50 particles in an SEM picture. The standard deviation was also
calculated, from which the distribution of the particle diameters
was derived. The nitrogen adsorption isotherms were measured
using a Belsorp-mini II (BEL Japan) at 77 K. The samples
were evacuated at 373 K under 10−3 mmHg before measure- Fig. 1. Relationship between the methanol ratio and average particle diameter
ment. The pore diameters were calculated using the Barrett– of MMSS obtained using quaternary ammonium salts with different alkyl-chain
lengths. (a) C10 TMABr, (b) C14 TMACl, (c) C16 TMACl and (d) C18 TMACl.
Joyner–Halenda (BJH) method. By considering the linearity of
a Brunauer–Emmett–Teller (BET) plot, the specific surface ar-
eas were calculated using adsorption data in the P /P0 range 3. Results and discussion
from 0.05 to 0.13. Transmission electron micrographs were ob-
tained using a Jeol-200CX TEM at an acceleration voltage of 3.1. Synthesis of MMSS with the same particle diameters and
200 kV. 29 Si magic-angle-spinning (MAS) nuclear magnetic different pore sizes
resonance (NMR) and 13 C cross-polarization (CP) NMR analy-
ses were carried out on a Bruker AVANCE 400 spectrometer at In order to clarify the relationship between the pore size of
79.49 MHz for 29 Si and at 100.61 MHz for 13 C. The 29 Si MAS a mesoporous silica catalyst and its catalytic activity, the pro-
NMR spectra were measured at 60 s repetition delay and 3 µs vision of MMSS samples with the same particle diameters and
pulse width. The 13 C CP MAS NMR spectra were measured at different pore sizes is necessary.
2 s repetition delay, 2 ms contact time, and 2.8 µs 1 H 90◦ pulse. We have previously succeeded in the synthesis of MMSS
The chemical shifts for 29 Si and 13 C NMR were referenced to with different pore sizes by using various surfactants as tem-
tetramethylsilane and glycine, respectively. For both measure- plates. However, no attempts have been made to synthesize
ments, the spinning rate was 5 kHz. N elemental analyses (EA) MMSS with the same particle diameters. From a previous re-
were carried out on an Elementer vario EL elemental analyzer. port [6], it is obvious that the particle diameter of MMSS
tends to increase as the number of carbon atoms in the sur-
2.3. Catalytic reactions factant increases. To overcome this tendency, the requirement
for precise control of the reaction conditions during MMSS
Reactions were carried out in a round-bottomed flask synthesis is essential. Because the particle diameter of the
equipped with a reflux condenser. The nitroaldol condensa- MMSS is significantly affected by the methanol ratio in the
tion was typically carried out as follows: a reaction mixture solvent [5], syntheses were conducted under various methanol
consisting of 50 mg of the catalyst and 0.61 g (5.0 mmol) of ratios. Fig. 1 shows the relationship between the methanol ratio
4-hydroxybenzaldehyde in 10 ml of nitromethane was heated and the average particle diameter of MMSS samples obtained
at 363 K with constant stirring for 1 h. The mixture was filtered from quaternary ammonium salts with different alkyl-chain
and the catalyst on the filter was washed thoroughly with chlo- lengths (C10 TMABr, C14 TMACl, C16 TMACl, C18 TMACl).
roform. The filtrate was evaporated under reduced pressure. The The methanol ratios with which the monodispersed particles
solid that was obtained was completely dissolved in a sufficient were obtained changed drastically depending on the type of
amount of acetone-d6 . An internal standard, THF (≈10 mmol), surfactant that was used. Although C16 -MS (MMSS synthe-
was added into the acetone-d6 solution. The product was ana- sized using C16 TMACl as a template) could be obtained over
lyzed by studying its 1 H NMR spectra on a Jeol JNM-LA500 a wide methanol ratio range (0.45–0.6), C14 -MS and C18 -MS
NMR spectrometer [14]. were only obtained in narrow methanol ratio ranges. It was
In case when 4-methoxybenzaldehyde was used as a reac- found that by precisely controlling the methanol ratio, MMSS
tant, the reaction was monitored periodically by withdrawing samples with almost the same particle diameters could be ob-
an aliquot of solution from the solution in order to determine tained using quaternary ammonium salts with different alkyl-
a reaction rate. The samples were filtered and analyzed by gas chain lengths.
chromatograph (Agilent, 6890) provided with a 30 m capillary
column of DB-1, using decane as an internal standard for mass 3.2. Characterization of Cx -MS-AP
balance. These products were characterized by MS (Agilent,
6890-5973) provided with a 30 m capillary column of DB-5MS. Amino-functionalized MMSS samples were obtained by
In order to compare reaction rates of various MMSS, we eval- grafting APTMS onto Cx -MS. Fig. 2 shows SEM images of
uated TOF (turn over frequency) from the data for the reaction Cx -MS-AP. As can be seen from Fig. 2, the particles were
time of 10 min. highly monodispersed, and their average particle diameters
252 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257

Fig. 3. XRD patterns of the samples: (a) C10 -MS-AP, (b) C14 -MS-AP, (c) C16 -
MS-AP, (d) C18 -MS-AP and (e) C18 PS-MS-AP.

isotherms of the samples are similar to those without organ-

ically functional groups (MMSS) [6], but C10 -MS-AP hardly
adsorbed any nitrogen. This result agreed well with the result
Fig. 2. Scanning electron micrographs of (a) C10 -MS-AP, (b) C14 -MS-AP, from the XRD measurements. Most of the mesopores in C10 -
(c) C16 -MS-AP, (d) C18 -MS-AP and (e) C18 PS-MS-AP. Scale bar represents MS-AP were filled with amino-moieties. As with the results of
1.1 µm. the XRD measurements, the pore sizes of Cx -MS-AP increased
as the alkyl-chain length of the surfactant increased. C18 PS-
were between 560 and 600 nm. Table 1 summarizes the syn- MS-AP that was synthesized from pore-expanded C18 PS by the
thetic conditions and properties of the particles. The regularity surfactant exchange method had the largest pore size of 2.66 nm
of the mesopores was studied by X-ray diffraction measure- (Table 1).
ments (Fig. 3). The peak intensities of d100 for all samples The N contents of the samples are also shown in Table 1. Al-
except for C10 -MS-AP were over 10,000, indicating that the though the N content of C18 PS-MS-AP (which has the greatest
Cx -MS-AP had well-ordered hexagonal regularity. In the case pore size) was a little larger (2.1 mmol/g), most of the samples
of C10 -MS-AP, the intensity of d100 was very low. It is con- had almost the same N content (1.71–1.82 mmol/g), indicating
sidered that because most of the mesopores of C10 -MS were that nearly equal amounts of amino groups were incorporated
filled with amino-moieties, the contrast between the mesopores into MMSS samples with different pore diameters.
and the silica walls became less, leading to low d100 intensity. For comparison, grafting of amino-moieties was conducted
From the XRD results shown in Table 1, it was confirmed that for FSM-16, which is mesoporous silica having irregular shape,
the value of d100 increased with increasing alkyl-chain length and monodispersed silica spheres without pores. They are de-
of the surfactant. noted as FSM-16-AP and SS-AP hereafter. FSM-16-AP had the
The nitrogen adsorption–desorption isotherms were mea- pore size of 1.98 nm which was little bit larger than C18 -MS-
sured and the pore-size distributions were analyzed by the BJH AP and N content of 1.74 mmol/g. SS-AP had almost the same
method. Fig. 4 shows the results for Cx -MS-AP samples. The particle diameter of MMSS samples (550 nm).

Table 1
Properties of aminopropyl-functionalized MMSS
Sample Surfactant Methanol Average Standard XRD Nitrogen adsorption N
concentra- ratio diameter deviation d100 Intensity Pore SBET Pore content
tion (mM) (nm) (%) (nm) (cps) size (m2 /g) volume (mmol/g)
(nm) (ml/g)
C10 -MS-AP 13.7 0.25 590 10.1 2.56 3400 – 280 0.03 1.71
C14 -MS-AP 13.7 0.44 600 5.4 3.18 10,100 1.34 730 0.31 1.79
C16 -MS-AP 13.4 0.50 580 6.2 3.36 12,500 1.53 870 0.40 1.78
C18 -MS-AP 13.8 0.59 560 6.3 3.67 14,500 1.89 850 0.50 1.82
C18 PS-MS-APa – – 590 5.6 4.09 16,200 2.66 810 0.74 2.10
FSM-16-AP – – – – 3.71 8900 1.98 600 0.41 1.74
SS-AP 550 4.0 0.13
a This was obtained by aminopropyl grafting on pore-expanded MMSS [19].
 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257 253

(A)

Fig. 6. 13 C CP MAS NMR spectrum of C16 -MS-AP. (* Signals from residual

ethoxy group. ** Signals from residual methoxy group.)

n = 2–4, Q4 : δ = −110 ppm, Q3 : δ = −100 ppm, Q2 : δ =

−92 ppm) and two peaks derived from Tm (Tm = RSi(OSi)m -
(OH)3−m , m = 1–3, T3 : δ = −65 ppm, T2 : δ = −57 ppm) were
observed. The spectrum that was obtained is the same as those
of organically functionalized SBA-15 or MCM-41 obtained by
the grafting method [20,21]. The presence of the Tm peaks
indicated the incorporation of amino-moieties into the silica
skeleton. In the 13 C CP MAS NMR spectrum, peaks derived
(B)
from the C atoms of the incorporated aminopropyl groups were
observed. This result also supports the conclusion that amino-
Fig. 4. (A) Nitrogen adsorption–desorption isotherms and (B) pore-size distri-
butions of (a) C10 -MS-AP, (b) C14 -MS-AP, (c) C16 -MS-AP, (d) C18 -MS-AP
moieties were successfully incorporated into the mesopores.
and (e) C18 PS-MS-AP. Other than main peaks, there are some small peaks observed.
Two small peaks at δ = 59, 16 ppm were assigned to the ethoxy
carbon atoms assumed to be generated by the partial esterifi-
cation of the surface silanol with ethanol which was used as
solvent for washing [22]. A small peak at 50 ppm was assigned
to unreacted methoxy carbon atoms of organic trimethoxysi-
lane [23].
The radial alignment of the mesopores was confirmed by
TEM observations (Fig. 7). It was observed that the hexago-
nal mesopores were arrayed in a radial pattern [19,24].
These experimental results led us to the conclusion that five
types of amino-functionalized MMSS with nearly the same par-
ticle diameters and different pore sizes had been successfully
prepared.

3.3. Evaluation of catalytic activity

The catalytic activities of Cx -MS-AP were studied by the ni-

troaldol condensation of nitromethane and benzaldehyde deriv-
atives. Firstly, 4-methoxybenzaldehyde was used as the reac-
Fig. 5. 29 Si MAS NMR spectrum of C16 -MS-AP. tant. Fig. 8 shows time course of yields of nitroaldol condensa-
tion of nitromethane and 4-methoxybenzaldehyde using C14 -
The 29 Si and 13 C CP MAS NMR spectra of C16 -MS-AP MS-AP, C18 -MS-AP, and FSM-16-AP as a catalyst. Table 2
were measured and its chemical structure was investigated summarizes initial rate and TOF (turnover frequency, min−1 )
(Figs. 5 and 6). In the 29 Si MAS NMR spectrum, three clear determined from data for 10 min reaction, and TON (turnover
resonance peaks derived from Qn (Qn = Si(OSi)n (OH)4−n , number) determined from data for 1 h reaction for various cat-
254 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257

Fig. 8. Changes in the activity of various types of catalysts in nitroaldol con-

densation reaction.

alysts. From Fig. 8 it is obvious that most of the reactions

finished just before 1 h. These results indicate that compari-
son of their catalytic activity by the use of TON determined
from the data for 1 h reaction is appropriate for this reaction
system. In order to determine reaction rates, we used GC for
analysis. However, yields determined by GC fluctuated and we
used NMR for determining TON, because the data obtained
from NMR is more reliable than the data from GC due to poor
volatile property of the product. It turned out that the initial TOF
and TON largely affected by the pore size of MMSS. C16 -MS-
AP (pore size: 1.53 nm) and C18 -MS-AP (pore size: 1.89 nm)
showed high catalytic activity. Initial TOF values of these sam-
ples were 1.3 and 1.2 min−1 , which were higher than the val-
ues for the other catalyst. The final yields of these samples
were also higher than those of the other catalysts, indicating
that initial TOF and TON correlated well for Cx -MS-AP sam-
ples. Meanwhile, amino-grafted FSM-16 (pore size: 1.98 nm)
Fig. 7. Transmission electron micrographs of (a) C16 -MS-AP and (b) C18 PS-
showed smallest initial TOF of 0.8 min−1 . However, yield in-
MS-AP. Scale bar represents 50 nm. creased gradually, and the final TON became almost the same
value with that of C18 -MS-AP. FSM-16 is mesoporous silica
having an irregular shape, and most of mesopores are aligned
in the same direction. This means that most of its outer surface
Table 2 is covered with silica walls, leading to limited access for the re-
Results of nitroaldol condensations of 4-methoxybenzaldehyde with ni- actants. On the contrary, in MMSS, the mesopores are aligned
tromethane in the presence of catalyst (363 K) radially from the center to the outside of a particle. Therefore,
Catalyst Initial rate Initial TOFa TONb reactants can enter into the mesopores from all directions. It
(mmol g−1 −1
cat min ) (min−1 ) (yield, %) may be concluded that Cx -MS-AP is one of the best catalysts
C10 -MS-AP 0 0 3 (5) for the nitroaldol reactions of benzaldehyde derivatives among
C14 -MS-AP 1.8 1.0 41 (74) other mesoporous silicas.
C16 -MS-AP 2.3 1.3 54 (96)
It turned out that the catalytic activity was affected by the
C18 -MS-AP 2.2 1.2 52 (94)
C18 PS-MS-AP 2.2 1.0 44 (92) pore size of MMSS, that is, the reaction is not in totally dif-
FSM-16-AP 1.3 0.8 52 (90) fusion control. If sizes of mesopores are close to the sizes of
a Turnover frequency = mmol product/mmol organic groups during initial reactants, a reaction could be mainly controlled by diffusion. In
10 min. TOF was determined by GC analysis. this case, a product is generated mostly on the surface of a cata-
b Turnover number = mmol product/mmol organic groups during 1 h. TON lyst. Therefore, it is very important to know what percentage of
was determined by NMR analysis. mesopores is effectively used for the reaction. Then, the effec-
 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257 255

Fig. 9. Changes in the activity of MMSS catalyst having different particle di-
ameter in nitroaldol condensation reaction.

Table 3
Results of nitroaldol condensations of 4-substituted benzaldehyde with ni-
tromethane in the presence of Cx -MS-AP catalyst (363 K, 1 h)

Fig. 10. Relationship between the pore size of particles and turnover number
for nitroaldol condensation. The pore size of C10 -MS-AP was estimated to be
zero from the nitrogen adsorption measurement.

the results obtained from the reaction between nitromethane

Catalyst Yield (%) TONa
and 4-hydroxybenzaldehyde (3a). MMSS without any amino
a b c a b c
groups (C16 -MS) showed no catalytic activity, suggesting that it
C10 -MS-AP 2 4 4 1 2 2 is the amino-moieties that act as catalysts. The yield increased
C14 -MS-AP 52 49 55 29 27 31
C16 -MS-AP 86 100 92 48 56 52
as we increased the pore size of the Cx -MS-AP, and reached
C18 -MS-AP 78 98 86 43 54 47 a maximum (86%) for C16 -MS-AP. C10 -MS-AP showed very
C18 PS-MS-AP 75 98 91 36 47 43 low catalytic activity; its yield was 2% and its TON was 1. Be-
C16 -MS 0 – – 0 – – cause the mesopores of C10 -MS-AP are considered to be filled
SS-AP 5 – – 38 – – with amino-moieties, the reactants could not reach the active
a Turnover number = mmol product/mmol organic groups during 1 h.
sites (= amino-moieties) inside the mesopores. When C14 -MS-
AP (pore size: 1.34 nm) or C16 -MS-AP (1.53 nm) were used as
tiveness factor of MMSS was evaluated from the kinetic data the catalysts, the TON increased to 29 or 48, respectively. It be-
measured by use of catalysts with different particle diameters came clear that the mesopores require a certain capacity for the
but the same pore size. Fig. 9 shows time course of yields of enhancement of catalytic activity. However, when C18 -MS-AP
nitroaldol condensation of nitromethane and 4-methoxybenz- (pore size: 1.89 nm) or C18 PS-MS-AP (2.66 nm), which have
aldehyde using C16 -MS-AP (particle diameter: 580 nm) and larger mesopores, were used as catalysts, the TON decreased
C16 -MS-AP-730nm (particle diameter: 730 nm). Initial rates to 43 or 36, respectively. Meanwhile, when SS-AP, which has
determined from the reaction time of 10 min for C16 -MS-AP nearly the same particle diameter with MMSS samples and no
and C16 -MS-AP-730 nm were 2.3 and 1.8 mmol g−1 −1
cat min , mesopores, was used as a catalyst, the yield was as low as 5%.
respectively. From these rate and particle diameter data, the This result strongly supports the assumption that the reaction
effectiveness factor of MMSS was evaluated to be 0.63. This occurred mainly in mesopores, but outer surfaces. Fig. 10 shows
result suggests that MMSS with smaller particle diameter could the relationship between the pore size of the samples and the
be a candidate for an ideal catalyst or catalyst support because TON for the reaction. The pore size of C10 -MS-AP was es-
higher effectiveness factor can be expected for smaller particles. timated to be zero from the result of the nitrogen adsorption
measurement. It can be seen from Fig. 10 that there is an op-
3.4. Influence of molecular size timum pore size for the reaction. In Fig. 10, results obtained
when 4-methoxybenzaldehyde was used as a reactant were also
In order to investigate the effect of the molecular size of shown and the tendency was the same with the results for the
the reactant on the catalytic activity, firstly, para-benzaldehyde other reactants.
derivatives incorporating alkoxy groups with different alkyl- Then, the study was conducted by using para-substituted re-
chain lengths were examined. Products for 1 h reaction were actants with longer substituents. The reactants that were used
analyzed, and yields and TON were determined. Table 3 shows were 4-butoxybenzaldehyde (3b) and 4-n-octoxybenzaldehyde
256 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257

Table 4
Results of nitroaldol condensations of multi-substituted benzaldehyde with nitromethane in the presence of Cx -MS-AP catalyst (363 K, 1 h)

Catalyst Yield (%) TONa

4-OMe 2,4,6-OMe 3,4,5-OMe 4-OMe 2,4,6-OMe 3,4,5-OMe
a b c a b c
C10 -MS-AP 5 21 1 3 12 1
C14 -MS-AP 74 69 27 41 39 15
C16 -MS-AP 96 62 77 54 35 43
C18 -MS-AP 94 83 88 52 46 48
C18 PS-MS-AP 92 75 90 44 36 43
a Turnover number = mmol product/mmol organic groups during 1 h.

(3c). The results of the reactions are also shown in Table 3 aldehydes such as 2,4,6-trimethoxybenzaldehyde (5b) and
and Fig. 10. In all cases, C16 -MS-AP exhibited the best cat- 3,4,5-trimethoxybenzaldehyde (5c) were examined. It is ex-
alytic performance among those tested, as seen for 4-hydroxy- pected that introduction of substituents on meta or ortho po-
benzaldehyde and 4-methoxybenzaldehyde. It is surprising that sition leads to the change in minimum kinetic diameter [17,26],
the optimum pore size remained unchanged for the reaction, which brings to the change in an optimum pore size for the
even though the molecular sizes of the reactants changed dras- reaction. Table 4 and Fig. 11 show the results of the catalytic
tically. Based on the Molecular Mechanics model, the sizes activities for the reaction. When multi-substituted aldehydes
of the molecules along their longer axes are about 0.67 nm were used as a reactant, C18 -MS-AP, which has a pore size of
for 4-hydroxybenzaldehyde, 0.82 nm for 4-methoxybenzalde- 1.89 nm, exhibited the highest catalytic activity. The optimum
hyde, 1.15 nm for 4-butoxybenzaldehyde, and 1.70 nm for pore size for the nitroaldol condensation shifted to higher value
4-n-octoxybenzaldehyde. The pore size of C16 -MS-AP as de- by increasing the number of substituent groups in the reactant.
termined from the adsorption measurement was 1.53 nm, which The molecular size of the reactants became more bulky as the
is smaller than the molecular size of 4-n-octoxybenzaldehyde number of methoxy groups increased, and then the catalytic
on its longer axis. However, all of the benzaldehyde deriva- activity of C16 -MS-AP fell sharply. On the contrary, the ac-
tives had the same molecular size, ca. 0.43 nm, for their shorter tivity of C18 -MS-AP and C18 PS-MS-AP hardly changed. The
axes. It is assumed that the benzaldehyde derivatives entered dimensions of the molecule between the 1-position and the 4-
into mesopores parallel to the pore directions and reacted with position for the longer axis is about 0.82 nm and the shorter
the amino groups that existed inside mesopores. The pore size axis is 0.43 nm for 4-methoxybenzaldehyde. The shorter axis
of C16 -MS-AP is considered to be the most suitable for the re- increases to 0.82 nm for the multi-substituted aldehydes be-
action of amino groups and para-substituted aldehyde moieties cause the methoxy groups are disposed along both sides of
from the view point of optimal space and density of the amino- the benzene ring. We assume from the foregoing experiments
moieties. that the aldehydes entered into the mesopores parallel to the
The size of the substituent group affected the catalytic activ- pore directions. Larger pores are considered to be necessary to
ities of the Cx -MS-AP samples. When 4-butoxybenzaldehyde enable more-bulky reactants to enter smoothly into the meso-
(3b) was used as the reactant, the value of TON was the highest pores.
in many cases, that is, Cx -MS-AP exhibits some selectivity for
the reaction [25]. It is thought that hydrophobic environment 4. Conclusions
was formed in the mesopores of MMSS by the incorporation of
organically functional groups, leading to easy infiltration of re- Precise control of the reaction conditions led to the success-
actants with an alkyl-chain into the mesopores. However, given ful synthesis of amino-functionalized monodispersed meso-
the diffusion of the reactant, there is an optimum size for the re- porous silica spheres with nearly the same particle diameters
action to take place. The optimum pore size was not affected at and different pore sizes or with the same pore size and different
all by the para-substitution of the reactant. particle diameters. By using the amino-functionalized monodis-
In order to investigate further the effects of the molecular persed mesoporous silica spheres as catalysts, the effect of pore
size of the reactant on the catalytic activity, multi-substituted size on base catalysis has been studied. It was found for the
 T.M. Suzuki et al. / Journal of Catalysis 251 (2007) 249–257 257

Acknowledgments

The authors thank Dr. Yasutomo Goto for assistance with

NMR measurements. The authors also thank Dr. Takao Masuda
and Dr. Teruoki Tago of Hokkaido University for fruitful dis-
cussion on the reaction rate of MMSS catalysts.

References

[1] B. Chiche, E. Sauvage, F.D. Renzo, I.I. Ivanova, F. Fajula, J. Mol. Catal.
A 134 (1998) 145.
[2] M. Iwamoto, Y. Tanaka, N. Sawamura, S. Namba, J. Am. Chem. Soc. 125
(2003) 13032.
[3] S. Pariete, P. Trens, F. Fajula, F.D. Renzo, M. Tanchoux, Appl. Catal.
A 307 (2006) 51.
[4] K. Yano, Y. Fukushima, Bull. Chem. Soc. Jpn. 75 (2002) 1977.
[5] K. Yano, Y. Fukushima, J. Mater. Chem. 13 (2003) 2577.
[6] K. Yano, Y. Fukushima, J. Mater. Chem. 14 (2004) 1579.
[7] E. Angeletti, C. Canepa, G. Martinetti, P. Venturello, Tetrahedron Lett. 29
(1988) 2261.
[8] D.J. Macquarrie, D.B. Jackson, Chem. Commun. (1997) 1781.
[9] Y. Kubota, Y. Nishizaki, H. Ikeya, M. Saeki, T. Hida, S. Kawazu,
M. Yoshida, H. Fujii, Y. Sugi, Microporous Mesoporous Mater. 70 (2004)
135.
[10] X. Wang, K.S.K. Lin, J.C.C. Chan, S. Cheng, J. Phys. Chem. B 109 (2005)
1763.
[11] F. Bigi, S. Carloni, R. Maggi, A. Mazzacani, G. Satori, Stud. Surf. Sci.
Fig. 11. Relationship between the pore size of particles and turnover number
Catal. 130 (2000) 3501.
for nitroaldol condensation.
[12] G. Demicheli, R. Maggi, A. Mazzacani, P. Righi, G. Sartori, F. Bigi, Tetra-
hedron Lett. 42 (2001) 2401.
[13] Y. Kubota, K. Goto, S. Miyata, Y. Goto, Y. Fukushima, Y. Sugi, Chem.
first time that an optimum pore size exists, which changes de- Lett. 32 (2003) 234.
pending on the size and shape of the reactants. In addition, by [14] S. Huh, H.T. Chen, J.W. Wiench, M. Pruski, V.S.Y. Lin, J. Am. Chem.
changing particle diameter, the effectiveness factor of MMSS Soc. 126 (2004) 1010.
was evaluated to be 0.63. Amino-functionalized monodispersed [15] Y. Kubota, H. Ikeya, Y. Sugi, T. Yamada, T. Tatsumi, J. Mol. Catal. A 249
(2006) 181.
mesoporous silica spheres were found to be excellent catalysts
[16] A.C. Blanc, D.J. Macquarrie, S. Valle, G. Renard, C.R. Quinn, D. Brunel,
for nitroaldol condensation due to the radial alignment of their Green Chem. 2 (2000) 283.
mesopores when compared to other types of mesoporous silica, [17] O. Jimenez, T.E. Müller, W. Schwieger, J.A. Lercher, J. Catal. 239 (2006)
such as FSM-16. Because the amino-functionalized monodis- 42.
persed mesoporous silica spheres were identical particles ex- [18] S. Inagaki, Y. Fukushima, K. Kuroda, J. Chem. Soc. Chem. Commun.
(1993) 680.
cept for the sizes of their mesopores, other external causes, such
[19] M. Mizutani, Y. Yamada, K. Yano, Chem. Commun. (2007) 1172.
as particle diameter and shape, and mesopore alignment, can be [20] H. Yoshitake, T. Yokoi, T. Tatsumi, Chem. Mater. 14 (2002) 4603.
ignored. It is believed that mesoporous materials such as these [21] S. Shylesh, A.P. Singh, J. Catal. 244 (2006) 52.
will be useful as basic research materials for studies of catal- [22] M.P. Kapoor, Q. Yang, Y. Goto, S. Inagaki, Chem. Lett. 32 (2003)
ysis if a reaction is not under diffusion control. Moreover, by 914.
[23] J.C. Hicks, C.W. Jones, Langmuir 22 (2006) 2676.
designing the pore properties of the MMSS through the intro-
[24] T. Nakamura, M. Mizutani, H. Nozaki, N. Suzuki, K. Yano, J. Phys. Chem.
duction of hydrophobic, hydrophilic, basic, or acidic moieties, C 111 (2007) 1093.
this material can now be considered as a candidate for a new [25] T.F. Degnan, J. Catal. 216 (2003) 32.
type of substrate-selective catalyst. [26] F.E. Imbert, N. Gnep, M. Guisnet, J. Catal. 172 (1997) 307.