Multifunctional catalysis promoted by solvent effects: Ti-mcm41 for a one-pot, four-step, epoxidation-rearrangement-oxidation-decarboxylation reaction sequence on stilbenes and styrenes

Article abstract of DOI:10.1021/cs501671a

Titanium sites grafted on several siliceous supports are able to act as multifunctional catalytic centers, activating tert-butyl hydroperoxide for oxidation reactions, as redox centers, and promoting rearrangements, as Lewis acids. In the one-pot, four-step conversion of stilbene into benzophenone, the best results were obtained over Ti-MCM41. Under suitable conditions, the catalyst promotes a tandem sequence: alkene epoxidation, epoxide rearrangement, aldehyde oxidation, and oxidative decarboxylation. α,α,α-Trifluorotoluene and a fluorinated glycerol-derived solvent were the optimal solvents for this tandem process, due to their polar aprotic character that allows the efficient oxidation reactions and a poor coordinating ability to prevent any deactivation of the Lewis acid character of the sites. The result of the tandem sequence of reactions is a ketone with loss of a carbon atom that, depending on the starting alkene, is the same result as that of an ozonolysis but under safer and milder conditions. Interesting and new insights on the mechanism of the different reactions involved are also described.

Full text of DOI:10.1021/cs501671a

Research Article
pubs.acs.org/acscatalysis
Multifunctional Catalysis Promoted by Solvent Effects: Ti-MCM41 for
a One-Pot, Four-Step, Epoxidation−Rearrangement−Oxidation−
Decarboxylation Reaction Sequence on Stilbenes and Styrenes
,
†
†
A. Mayoral, Fabio G. Santomauro, and Matteo Guidotti‡
†
‡,§
Jose
†
́
M. Fraile,* Nuria García, Jose
́
́ ́
ISQCH−Instituto de Síntesis Química y Catalisis Homogenea, Universidad de Zaragoza−CSIC, C/Pedro Cerbuna, 12, E-50009
Zaragoza, Spain
‡
Istituto di Scienze e Tecnologie Molecolari, Consiglio Nazionale delle Ricerche, via Golgi 19, I-20133 Milano, Italy
S Supporting Information
*
ABSTRACT: Titanium sites grafted on several siliceous supports are able to act as
multifunctional catalytic centers, activating tert-butyl hydroperoxide for oxidation
reactions, as redox centers, and promoting rearrangements, as Lewis acids. In the
one-pot, four-step conversion of stilbene into benzophenone, the best results were
obtained over Ti-MCM41. Under suitable conditions, the catalyst promotes a
tandem sequence: alkene epoxidation, epoxide rearrangement, aldehyde oxidation,
and oxidative decarboxylation. α,α,α-Trifluorotoluene and a fluorinated glycerol-
derived solvent were the optimal solvents for this tandem process, due to their polar
aprotic character that allows the efficient oxidation reactions and a poor coordinating ability to prevent any deactivation of the
Lewis acid character of the sites. The result of the tandem sequence of reactions is a ketone with loss of a carbon atom that,
depending on the starting alkene, is the same result as that of an ozonolysis but under safer and milder conditions. Interesting
and new insights on the mechanism of the different reactions involved are also described.
KEYWORDS: multifunctional catalysis, titanium catalysts, epoxidation, epoxide rearrangement, oxidative decarboxylation,
tandem catalysis
INTRODUCTION
on the contrary, although they often present a lower oxidative-
cleavage capability, involve different reaction mechanisms and
are particularly suitable to be heterogenized by deposition onto
solid supports or by insertion into porous matrices. Nonheme
iron complexes, both in homogeneous and heterogeneous
phase, have been also used for the cleavage of styrene, β-
methylstyrene, and stilbene, leading to benzaldehyde in all cases
■
The cleavage of alkenes to the corresponding carbonyl or
carboxylic compounds is a widely used method in organic
synthesis, in particular as a strategy to introduce oxygen
functionalities in molecules. These reactions can be carried out
1
,2
by both chemical and enzymatic methods. The most popular
chemical method is ozonolysis, although it is not free from risks
mainly in large scales. Other chemical methods are based on
the use of oxidants in stoichiometric or even overstoichiometric
amounts, leading to the formation of high amounts of
concomitants byproducts. On the other hand, enzymatic
methods are suitable for a limited number of molecules and,
although further developments are envisaged, they still lack of
broad applicability to families of organic compounds. In this
regard, the development of chemical strategies based on the use
of environmentally benign oxidants is a matter of interest.
Oxygen has been shown to cleave double bonds adjacent to
phenyl or naphthyl groups in the presence of N-hydroxyph-
9
when the reaction was carried out in acetonitrile, or to
benzaldehyde dimethyl acetal when using methanol as a
10
solvent. In the last case, it was proposed that the opening
of the epoxide with methanol is the key step for the double
bond cleavage. The carboxylic acid was obtained with aqueous
11
TBHP as the oxidant and FeCl /NaOH as the catalyst. Again,
3
the main oxidation product obtained in the aerobic oxidation of
styrene over supported nanocrystalline gold was benzaldehyde,
12
with lower amounts of the desired epoxide. Benzaldehyde and
benzoic acid were also important side products in the
epoxidation of trans-stilbene over catalytically active cobalt-
13
3
substituted TUD-1 (Co-TUD-1) using molecular oxygen.
thalimide.
4
Titanium-containing heterogeneous catalysts have been used
in the synthesis of adipic acid with tert-butylhydroperoxide
(TBHP) as the terminal oxidant. The cleavage took place
through the formation of the diol and it was necessary to use
From the seminal work of Noyori and co-workers, hydrogen
peroxide has been used, combined with tungstates and
molybdates in the synthesis of adipic acid.
second- and third-row transition metals (in particular, Ru, Os,
and W) have shown good performances as catalytically active
centers for the cleavage of alkenes, since they undergo
oxidation to the related metal-oxo species, which are, in turn,
5
−8
Then, several
Received: October 29, 2014
Revised: April 27, 2015
5
able to perform oxidative cleavage. First-row transition metals,
©
XXXX American Chemical Society
3552
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
Table 1. Textural Properties and Ti Content of the Pristine Silica Supports and of the Final Grafted Catalysts
S (m g−1)
2
V
tot
p
(mL g−1
)
Dp (nm)
a
Ti grafted precursor
Ti content (wt %)
b
UV max (nm)
c
support
MCM41
A
960
930
919
920
905
601
514
512
485
0.75
0.70
0.64
1.04
0.99
1.00
0.91
0.82
0.72
3.7
3.6
3.4
4.4
4.3
9.2
9.1
6.3
Ti-MCM41
Ti-MCM41
HMS
Ti(Cp) Cl2
1.70
1.59
224
220
2
d
Ti(Cp) Cl2
2
Ti-HMS
SBA15
Ti(Cp) Cl2
1.74
1.80
247
229
285
2
Ti-SBA15
Ti(Cp) Cl2
2
SiO (Merck)
2
i
Ti-SiO2
6.0
Ti(O Pr)4
0.96
a
b
c
Pore diameter (determined by Broekhoff and de Boer method). Ti content was determined by ICP-AES. DRS-UV−vis maximum absorption (full
d
DRS-UV−vis spectra are available as Supporting Information). Catalyst after 2 catalytic cycles. The solid underwent calcination in air at 500 °C
after each recover.
Ti−Al−SBA15, a bifunctional catalyst with both redox and acid
catalysts, where the redox and acid character of Ti(IV) sites on
silica are expressed at best.
sites, because monofunctional Ti-SBA15 led exclusively to the
1
4
epoxide. Titanium in an aluminophosphate framework (Ti-
AlPO) was able to convert cyclohexene into adipic acid with
hydrogen peroxide, thanks to a combination of radical and
RESULTS AND DISCUSSION
■
15
Properties of the Grafted Catalysts. Three mesoporous
supports (MCM41, HMS, SBA15), with different textural
properties (Table 1), were used for grafting of bis-
concerted oxidation mechanism. Previously, TS-1 and TS-2
zeotypes exhibited high activity in the conversion of α-
methylstyrene into acetophenone or of styrene into benzalde-
hyde in the presence of aqueous hydrogen peroxide, by
exploiting the oxidizing and Lewis-acid character of Ti(IV)
[
(cyclopentadienyl)dichlorotitanium(IV)], Ti(Cp) Cl , using
2 2
32
the method previously described by Oldroyd et al. In the
grafting procedure, however, pyridine was used instead of
triethylamine as basic additive. Nonordered amorphous Merck
1
6,17
sites.
Again, Ti, in the form of TiO /SiO system, was
2 2
active in the conversion of styrene into benzaldehyde in the
presence of molecular oxygen at 10 bar, via a diol and cleavage
i
28
silica was used as support for the grafting of Ti(O Pr) .
4
The Ti content for the three ordered mesoporous catalysts is
1
8
reaction pathway.
around 1.7−1.8 wt %, whereas for the nonordered Ti-SiO
2
Single-site heterogeneous catalysts prepared by grafting
titanium precursors onto mesoporous silica materials have
material is 0.96 wt %. Such relevant difference is justified to get
a comparable surface dispersion of titanium between Ti-
19
provided interesting results in the epoxidation of alkenes,
MCM41 and Ti-SiO . In fact, taking into account the lower
2
because the pore size in the range of mesopores helps to
circumvent the steric limitations often found in microporous
zeolites and the dispersion of titanium sites can be easily
modulated through the careful dosing of the metal precursor
specific surface area of the latter and considering the surface
density of Ti sites per square nanometer, one finds 0.22 and
−2
0
.24 Ti atoms nm , for Ti-MCM41 and Ti-SiO , respectively.
2
Actually, the concentration and dispersion of the Ti sites on the
i
during the synthesis. Precursors such as Ti(O Pr)4 and
surface proved to be a very important parameter in related
titanocene derivatives were used to prepare this kind of
catalysts, leading to good results in the epoxidation of several
24,33
solids.
For this reason, the loading of Ti on amorphous
silica was tuned in order to approach such value.
2
0−23
alkenes in the presence of TBHP
peroxide.
or hydrogen
The cleavage of α-methylstyrene to acetophe-
none has been described in the epoxidation with hydrogen
The grafting of titanocene on mesostructured silicas led to a
decrease in specific surface area, that is more marked for Ti-
SBA15 (ca. 15%) and less important for Ti-MCM41 and Ti-
HMS (<5%). This behavior is consistent with the data reported
in previous studies, as it is due to the deposition of titanium
species onto the internal surface of the porous solids, and it also
suggests a partial plugging of the microporous fraction of the
2
4−26
27
peroxide in acetonitrile catalyzed by Ti-containing zeolites. In
this case, it was proposed that the epoxide is opened by
hydrogen peroxide to yield a hydroperoxide intermediate that is
then transformed into the acetophenone via Grob fragmenta-
tion. Consistently with this mechanism, alkyl hydroperoxides
were not suitable oxidants for these reactions because they are
worse nucleophiles and hardly open epoxides. However, in
some cases, non-negligible traces of benzaldehyde have been
detected in the epoxidation of styrene with alkyl hydro-
23,34
SBA15 network.
The evaluation of the X-ray diffraction
patterns (see the Supporting Information) of the ordered
mesoporous catalysts show that their structure was maintained
after the grafting. The use of pyridine instead of triethylamine,
as a base to favor the grafting, in fact, minimizes the damages to
the mesoporous ordered structure of the material, as the former
is less aggressive than the latter against the siliceous walls of the
support.
The UV−vis diffuse reflectance spectroscopy (DRS) gives an
indication about the degree of dispersion of the Ti(IV) sites on
the silica supports. The maxima of absorption profiles of Ti-
MCM41 and Ti-SBA15 (Table 1) show that in these solids the
active sites are rather evenly dispersed on the silica surface,
mainly as isolated centers (Ti maxima < 230 nm). On the other
hand, the presence of absorption features above 240 nm has
been attributed to an incipient oligomerization of Ti(IV)
28−30
peroxides under anhydrous conditions,
indicating that
another mechanism must operate. In this manuscript, we
explored the behavior of several substituted styrenes and
stilbenes with TBHP, in the absence of water or nucleophilic
alcohols, using different Ti-silica catalysts obtained by grafting
the catalytically active centers onto various mesoporous solids.
The use of anhydrous TBHP is important to reduce the
undesired hydrolysis of Ti−O−Si bonds and hence to minimize
31
the risk of leaching of the active Ti sites out of the catalysts. It
is then shown how the proper choice of the reaction conditions
makes these systems capable to act as one-pot multistep
3
553
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
3
5
species into small Ti−O−Ti domains. This is the case of Ti-
HMS that shows a broader band with a maximum at 247 nm,
assigned to the presence of Ti−O−Ti dimeric species.
particular, Ti-SBA15 was less selective as catalyst than Ti-HMS
and Ti-MCM41, both for stilbene conversion and selectivity to
benzophenone (50 vs 67−68% after 72 h).
36
However, UV studies on model titanium-silsesquioxanes
have shown that monomeric Ti species can absorb at longer
wavelengths, depending on the coordination sphere of Ti
tetrahedral or octahedral) and the number of bonds to the
silica surface (monopodal, bipodal, and tripodal species). In
particular, the maximum at 285 nm of Ti-SiO is compatible
with octahedral monopodal species. Nevertheless, taking into
account the entire DRS UV−vis spectra (see the Supporting
Information), the presence of a separate TiO phase (in
The formation of 5 from 3 requires the migration of a phenyl
group in a rearrangement-type reaction, probably catalyzed by
acidic centers, and the loss of one carbon atom. The
rearrangement of stilbene epoxide, with migration of phenyl
groups, has been described over strong Brønsted or Lewis acid
(
40,41
2
centers, such as Nafion or chromium(V) oxoporphyrins.
Analogously, minor amounts of 5 during the epoxidation of
trans-stilbene 1 have been detected in several cases over other
42−44
2
classes of catalysts.
Moreover, some decades ago, the
particular anatase) can be ruled out, as no important absorption
noncatalyzed oxidation with O of 1 and thermal rearrange-
37
2
bands were recorded in the range 350−400 nm.
Reaction of Stilbenes. In order to test the behavior of
these catalysts, we first studied the oxidation of trans-stilbene
1) with an excess of anhydrous TBHP (6 equiv) in various
solvents, keeping the reaction mixture temperature at 75 °C
Scheme 1). Table 2 summarizes the results obtained over four
ment to 5 was proposed as a potential process for the
45
production of benzophenone from stilbene, even though it
cannot be considered commercially and environmentally
sustainable according to today’s standards.
In terms of activity for trans-stilbene conversion, expressed as
turnover number after 24 h of reaction, Ti-HMS, Ti-MCM41
(
(
and Ti-SiO showed good results, especially in acetonitrile
2
Scheme 1. Ti Catalyzed Reaction of Stilbenes with tert-Butyl
Hydroperoxide
(
Table 2). Nevertheless, aiming at enhancing and maximizing
the multistep formation of benzophenone, the selectivity to the
desired final product has to be taken into account, and other
solvents were tested. It is indeed well-known that the solvent
plays an important and sometimes decisive role in the behavior
46
of catalysts. Solvent effects have been studied in several
47−51
oxidation reactions, including the epoxidation of styrene.
Over Ti-HMS, three aprotic solvents, to prevent the
possibility of epoxide ring opening, with different polar
character were also tested, namely ethyl acetate, 1,2-dichloro-
ethane and toluene (the reaction mixture was kept at 75 °C for
all solvents). Stilbene (1) conversions depend on the nature of
the solvent and the lowest values were reached in toluene. The
need of polar aprotic solvents as ideal reaction media for the
epoxidation of alkenes in the presence of TBHP over
mesoporous titanosilicates is widely accepted and it has been
different catalysts. First, acetonitrile was the solvent of choice,
as it showed optimal performance in the selective epoxidation
22,23,29,38
of functionalized alkenes with TBHP.
In such solvent,
the primary product of the reaction observed at short reaction
times was the corresponding trans-stilbene epoxide (3), as
expected. This product is due to epoxidation of the CC
double bond catalyzed by titanium(IV) centers in the presence
of alkyl hydroperoxides and such behavior is fully consistent
32,39
with several previous observations.
In addition, thanks the
52
confirmed by several experimental evidence. So, the poor
use of anhydrous TBHP, the reaction proceeded mainly
following a heterolytic oxidation pathway and hence with a
good selectivity to the desired epoxide. However, at longer
reaction times (72 h), the reaction proceeded further and the
major observed product was surprisingly benzophenone (5). In
N
T
53
performance of toluene (the solvent with the lowest E value,
Table 2) in terms of conversion was not unexpected and, in
fact, the conversion after 24 h correlates very well with the
polarity of the solvent (Figure 1) in reactions catalyzed by Ti-
a
Table 2. Results Obtained from the Oxidation of trans-Stilbene (1) with TBHP
b
c
d
%
conv (% select)
TON
24 h
productivity
72 h
N
T
catalyst
solvent
E
24 h
72 h
Ti-SBA15
Ti-HMS
CH CN
0.460
0.460
0.228
0.099
0.327
0.460
0.228
0.327
0.241
0.460
0.228
0.241
70 (18)
87 (40)
62 (20)
54 (9)
89 (50)
99 (67)
98 (46)
58 (86)
88 (94)
98 (68)
99 (46)
92 (97)
93 (96)
77 (45)
nd
18
23
17
15
20
23
15
22
24
21
4
11.6
17.9
12.2
13.5
22.3
18.4
12.6
24.6
24.6
9.6
3
CH CN
3
AcOEt
PhCH3
ClCH CH Cl
74 (55)
85 (37)
55 (22)
79 (58)
86 (63)
76 (35)
15 (13)
46 (94)
2
2
Ti-MCM41
CH CN
3
AcOEt
ClCH CH Cl
2
2
PhCF3
e
Ti-SiO2
CH CN
3
AcOEt
PhCF3
nd
52 (94)
13
13.5
a
b
Reaction conditions: 0.49 mmol stilbene, 50 mg catalyst, 6 equiv TBHP, 75 °C. Conversion of stilbene and selectivity (in parentheses) to
c
−1
d
benzophenone (5), determined by GC. Turnover number (mol converted stilbene mol total Ti in the catalyst) after 24 h reaction. mol of
benzophenone (5) produced mol total Ti in the catalyst. 88.7 mg of catalysts were used.
−1
e
3
554
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
HMS and Ti-MCM41, that arguably show the most similar
sites, as shown by DRS-UV.
proved to be the best solvent to promote the cleavage of the
double bond (Table 2). However, the correlation between
N
T
conversion and E is not so good with this solvent (Figure 1),
as the yield obtained was higher than expected, in agreement
with an influence of other solvation mechanisms. In fact, the
use of mixtures of this solvent with toluene was highly
detrimental, and only a 10% of toluene was enough to reduce
significantly both conversion and selectivity. Moreover, taking
into account the higher boiling point of PhCF , it was possible
3
to run the reaction at higher reflux temperature (102 °C) and
this led to improved benzophenone yields, nearly to
quantitative in only 24 h (Table 3).
Table 3. Results Obtained from the Oxidation of trans-
a
Stilbene (1) over Ti-MCM41 in Fluorinated Solvents
b
c
solvent
PhCF3
T (°C)
t (h)
% conversion
% selectivity
0
60
1
7
46
46
86
93
82
98
40
48
9
Figure 1. Correlation between the stilbene conversion after 24 h (data
from Table 2) catalyzed by Ti-HMS (■, r = 0.9924) and Ti-MCM41
d
2
4
4
1
20
24
72
63
96
(
open symbols, r = 0.9463 with the seven square points) and the
N
80
polarity of the solvent (E parameter).
T
72
1
24
1
24
1
24
d
1
02
29
Selectivity to benzophenone (5) was strongly affected too by
the nature of the solvent and it decreased in the order 1,2-
dichloroethane (94%) > toluene (86%) > acetonitrile (67%) >
ethyl acetate (46%). In previous works, it was shown that there
exists a close relationship between acidity of the titanium
catalytic center and both epoxidation (with H O or TBHP)
97
d
PhCH3
110
110
0
71
11
90
93
e
3
F03F
F13F
41
60
46
96
99
2
2
23,54,55
48
1
and epoxide opening activity,
and this effect depends on
e
d
3
110
9
the structure of the catalytic sites. Indeed, solvents with high
coordinating ability likely reduce the acidity of the catalytic
centers lowering catalytic activity and, in particular, the acid-
promoted rearrangement, first step in the formation of
2
4
4
8
92
99
a
Reaction conditions: 0.49 mmol of stilbene, 50 mg Ti-MCM41, 6
b
c
equiv TBHP. Determined by GC. Selectivity to benzophenone (5).
benzophenone. Differences in Gutmann donor number (D )
N
d
e
5
6,57
Aldehyde was the main product. See ref 60.
values
the selectivity order: 1,2- dichloroethane (0.0) < toluene (0.1)
acetonitrile (14.1) < ethyl acetate (17.1). The order of global
support this hypothesis and qualitatively agree with
Following such strategy, other fluorinated solvents were then
<
tested, namely two ethers of glycerol with 2,2,2-trifluoroethyl
chains (Figure 2), labeled as 3F03F and 3F13F (Table 3).
activity for the transformation of 1 into 5 can be therefore seen
as the balance between the two main features of the solvent: a
marked polar aprotic character, that is optimal for the stilbene
epoxidation step, and a limited coordinating ability, that is good
to enhance the acid-promoted rearrangement at the Ti(IV) site.
The good performance of 1,2-dichloroethane was also
demonstrated in the case of Ti-MCM41 (Table 2), with also
slightly better conversion and selectivity than Ti-HMS. In this
case, the same considerations about the influence of the
solvents are valid, as above.
60
Figure 2. Glycerol-derived fluorinated solvents.
However, only 3F13F was able to provide high benzophenone
yields at temperatures over 110 °C. Both solvents, α,α,α-
trifluorotoluene and 3F13F, are polar enough (E NT values:
In principle, a solvent with aprotic polar character, but with
no coordinating ability and even lower polarizability than 1,2-
dichloroethane should enhance further the activity and
selectivity to benzophenone. α,α,α-Trifluorotoluene possesses
solvent properties that are comparable to those of dichloro-
methane, but with a remarkably lower volatility and a reduced
toxicity. Thanks to the presence of fluorine atoms, PhCF3
shows an intermediate polarity between ethyl acetate and 1,2-
PhCF = 0.24, 3F13F = 0.55) to allow dissolution of reagents
3
and to carry out the epoxidation step, but their low basicity and
polarizability, due to the presence of fluorinated chains, make
them poor solvating agents for the Ti(IV) acid sites. This does
not hinder the effective acidity of the catalysts. On the other
hand, the protic character of 3F03F reduced the epoxidation
yield, perhaps also because of the possible oxidation of the
solvent. Nevertheless, the selectivity to benzophenone was also
high, showing that acidity is not reduced. In spite of the
excellent results reached with the glycerol-derived 3F13F, the
commercial and more easily available PhCF was used to
explore the behavior of other alkenes.
By comparing the performance of the catalysts obtained from
four supports with different textural properties (Tables 1 and
2), it appears that the best results in terms of productivity to 5
N
58
dichloroethane (ET = 0.241, Table 2) does not undergo
oxidation at the benzylic positions and the highly electron-
withdrawing character of the trifluoromethyl group prevents
any electrophilic attack on the aromatic ring. Such relatively
uncommon reaction medium is attracting a growing interest as
a less hazardous alternative to chlorinated organic solvents for
reactions where the Lewis character of the catalyst has to be
3
5
9
enhanced. Confirming this hypothesis, α,α,α-trifluorotoluene
3
555
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
a
Scheme 2. Mechanism of the Tandem Multistep Reactions over Stilbenes
a
The dashed arrow indicates a pathway not described previously.
per Ti site were attained on Ti-MCM41 and Ti-HMS, which
are solids derived from high-surface area ordered mesoporous
silica materials. Indeed, stilbene molecules can freely enter and
be accommodated within the mesopore network of molecular
Finally, a study has been performed on the Ti-MCM41
catalyst to investigate the catalyst recyclability. The solid can be
recovered and reused in new catalytic runs, but a careful
filtration, washing with methanol and calcination under air at
500 °C was necessary to recover the pristine activity and
remove the organic deposits on the catalyst surface (2.8 wt %,
according to TGA measurements of weight loss between 100
and 800 °C, compared to the one of fresh Ti-MCM41). After
this treatment, the initial catalytic activity was recovered in large
32
sieves whose pores are wider than 3 nm, so the mean pore
diameter was not a major limiting factor in defining catalyst
reactivity. Conversely, the higher the specific surface area of the
support, the more dispersed the Ti(IV) sites are on the silica
materials. A better site isolation leads to a more marked acid
character of the Ti(IV) centers and hence a better activity in
part, as the conversion after 24 h is 83%, in PhCF . Conversely,
3
23,54
acid-catalyzed rearrangement reactions.
In addition, a
without such intermediate treatment, the conversion after 24 h
was 6% only and calcination at 300 °C (as in the activation of
the fresh catalyst) improves conversion only up to 26% after 24
h. A thorough removal by calcination of the organic deposits
that gradually foul the catalyst surface is thus necessary. The
textural properties (Table 1) of the catalyst after two cycles
shows that a small loss of specific pore volume took place due
to a minor loss of mesostructure order, but no relevant damage
or collapse of the mesoporous network was observed. In terms
of spectroscopic characterization, even though the absorption
maximum did not changed its position (Table 1), a gradual
higher dispersion of Ti(IV) sites on the silica surface renders
the catalyst less prone to deactivation. In fact, single-site Ti-
silica catalysts usually lose their activity due to the gradual
agglomeration of isolated Ti(IV) centers into larger TiO -like
2
domains, able to homolytically decompose the oxidant and also
61
with lower Lewis acidity. Therefore, the higher the dispersion
of Ti centers in the catalyst, the less the catalyst suffers from
deactivation during the reaction, although this is controlled also
by the role of the solvent, as can be seen by the results with Ti-
SiO . At short reaction time, stilbene conversion is higher in
2
acetonitrile, better solvent for epoxidation, whereas the
selectivity to benzophenone is much better in trifluorotoluene,
as the Lewis acidity is boosted in this solvent. However, in both
cases the catalyst is deactivated and the reactions stop and do
not proceed further after 24 h, in contrast with the results
obtained with the crystalline mesoporous solids. For these
reasons, the order of productivity to the desired product 5 for
the four solids can be depicted as follows: Ti-MCM41 ≈ Ti-
agglomeration of the isolated Ti(IV) sites into larger TiO -like
2
domains occurred and was revealed by the increase of
absorption shoulders around 350−400 nm (see Supporting
Information). These factors may account for the slight loss in
performance observed after two catalytic cycles and they are
consistent with the previous observations by some of us in
recycled Ti-containing silica epoxidation catalysts when TBHP
23,26
or H O were used as an oxidant.
2
2
HMS > Ti-SBA15 > Ti-SiO (that is a nonordered amorphous
Reaction Mechanism. Diphenylacetaldehyde (6) and
benzophenone (5) were the main products in all the cases.
Scheme 2 shows the mechanism going from epoxide to
benzophenone. This mechanism includes an acid-promoted
rearrangement of the epoxide (3) to diphenylacetaldehyde (6).
From this aldehyde, benzophenone can be obtained after loss of
one carbon atom by decarbonylation, either through the
oxidation to diphenylacetic acid (7) and subsequent decarbox-
ylation, or through the formation of diphenylmethyl formate
(8), result of a Baeyer−Villiger oxidation. The evolution of
2
mesoporous solid).
Exploiting the promising reaction conditions over Ti-
MCM41, at 102 °C in PhCF , an overall isolated yield of
3
82% to benzophenone (5) could be obtained starting from a
larger amount (2.0 mmol) of trans-stilbene, via a one-pot,
multistep transformation.
Analogously, a series of tests were run on cis-stilbene (2)
(Scheme 1), in order to investigate the potential influence of
cis−trans configuration of the substrate on the reaction
performance. Nevertheless, no remarkable differences were
detected with respect to the behavior already observed for
trans-stilbene (1), in terms of activity and selectivity. Under
optimized conditions, cis-stilbene yielded a 96% of benzophe-
none with a complete conversion of the alkene after 24 h. This
is a non-negligible result, since cis-stilbene epoxide is typically
far less reactive than trans-stilbene epoxide in epoxide ring
opening reactions promoted by homogeneous or heteroge-
CO in the final step was experimentally observed and
2
63
measured quantitatively.
Further information about the reaction mechanism can be
obtained from the evolution of the reaction with time (Figure
3). In an attempt to slow down the reaction to better observe
the evolution at short reaction time, the experiments were
conducted with a lower amount of catalyst (35 mg). Under
these conditions and in trifluorotoluene, the epoxide (3) was
barely detected along the reaction, indicating that epoxidation
step is slower than the subsequent rearrangement. At the same
6
2
neous acid sites. In this case, on the contrary, either cis- or
trans-isomers gave rise to the same product in a similar way.
3
556
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
from a faster deactivation, as the reaction did not proceed
further after 5−6 h. Conversely, on Ti-SBA15 the alkene
conversion remained constant after 10 h. The catalyst was then
only active for the subsequent oxidation steps, indicating that
these steps take place through a different mechanism, and then
it underwent deactivation only after 24 h. It is difficult to
propose an explanation for differences in catalyst deactivation,
given that, in this case, the reaction steps are influenced by a
large number of factors, whose fine counterbalance addresses
the global reaction trend toward the final desired product.
Those factors are very difficult to evaluate, even in much
simpler reaction systems, as the single epoxidation of alkenes, in
64
which the structure of the support material has shown little or
65
very high influence, depending on the substrate and the
reaction conditions. Nonetheless, the catalyst stability increases
with the surface area and, consequently, with the dispersion of
the Ti-sites.
Figure 3. Reaction of trans-stilbene (1) with TBHP (6 equiv)
catalyzed by Ti-MCM41 (35 mg), in PhCF at 102 °C. The numbers
3
correspond to the different reaction products (see Scheme 2).
The evolution of the intermediates parallels the one observed
in Ti-MCM41, with the epoxide that is hardly detectable.
time, as aldehyde 6 was the major product at short reaction
times, the subsequent oxidation steps were not faster than
rearrangement, but not much slower either considering that the
production of benzophenone (5) was remarkable, even at
relatively low stilbene conversions.
However, at short reaction times, Ti-SiO led to a higher
2
proportion of benzophenone (5). From DRS-UV−vis, this
catalyst contains either a lower amount of isolated Ti-sites or a
higher proportion of monopodal species; as in any of the two
cases the sites should be less acidic than isolated bipodal or
Then, we tested the influence of the support by using Ti-
SiO and Ti-SBA15 prepared following the same grafting
2
66
tripodal ones, rearrangement to aldehyde (6) is comparatively
procedure (Table 2 and Figure 4). The main differences with
slower and the amount of aldehyde present in the reaction
medium is lower as it is more efficiently oxidized. As the nature
of the support did not affect the final selectivity, these tests
confirmed that Ti(IV) is the site where not only the oxidation
step but also the rearrangement occurs, and hence, they behave
as multifunctional sites.
Ti-MCM41 are related to catalyst deactivation. Ti-SiO was
2
fairly active in the first hours of reaction, but then it suffered
Trying to confirm the hypotheses depicted in Scheme 2 by
identifying some of the intermediates, the reaction was carried
out using the exact stoichiometric amount of oxidant needed in
the reaction from 1 to 5, namely 3:1 TBHP/stilbene molar
ratio. The most striking result was the detection of a rather high
amount of benzil (10) (around 18 mol %) at the end of the
reaction (Figure 5). This product might come from the
oxidation of phenyl acetophenone (9), so the less favored
rearrangement is more important in the presence of a lower
amount of oxidant. In view of this, the epoxide (3) was made to
rearrange in the absence of oxidant (Figure 6). As it can be
seen, the amount of ketone (9) was very small and it did not
Figure 4. Reaction of trans-stilbene (1) with TBHP (6 equiv)
catalyzed by 35 mg of Ti-SiO (top) and Ti-SBA15 (bottom), in
Figure 5. Reaction of trans-stilbene (1) with TBHP (3 equiv)
2
PhCF at 102 °C. The numbers correspond to the different reaction
products (see Scheme 2).
catalyzed by Ti-MCM41 (35 mg), in PhCF at 102 °C. The numbers
correspond to the different reaction products (see Scheme 2).
3
3
3
557
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
justify the formation of 18 mol % of benzil, which should be
explained by an alternative mechanism.
the aldehyde under suitable conditions, a possibility disregarded
in most of mechanistic studies of this reaction.
The direct breakdown of diphenylacetaldehyde to benzo-
phenone has been recently described with iodosylbenzene
69
combined with HBF or BF , but the proposed mechanism
4
3
through a five-membered cyclic intermediate including the
hypervalent iodine, does not seem to be compatible with our
system. As shown in Scheme 2, benzophenone (5) could be
obtained by two alternative routes: through diphenylacetic acid
(
7) or benzhydryl formate (8) as intermediates. The broad
peak attributed to 7 was always detected in the intermediate
samples of the reaction medium (even though gas-chromatog-
raphy is not the optimal technique to reveal it). Conversely, the
sharp peak of 8 has never been detected. In order to get
additional information on this point, both 7 and 8 were used as
starting reactants under the typical reaction conditions in the
presence of TBHP and both compounds led to the formation
of benzophenone 5, the formate 8 being far less reactive and
prone to rearrangement than the acid 7. So, both pathways are,
in principle, possible (via oxidation and decarboxylation or via
Baeyer−Villiger and oxidative decarboxylation), but the route
via 7 is more likely, since the formate 8 has never been detected
as an intermediate compound in the reaction mixture (although
it is poorly reactive). Finally, the mechanism leading from 7 to
Figure 6. Rearrangement of trans-stilbene oxide (3) catalyzed by Ti-
MCM41 (35 mg), in PhCF at 102 °C (in the absence of oxidant).
3
The numbers correspond to the different reaction products (see
Scheme 2).
In order to shed light on this route, the oxidation of the
aldehyde 6 was performed with a substoichiometric amount of
TBHP (2 equiv) (Figure 7). More than 40 mol % of benzil
5
is not fully clear, but it is probable that an oxidative
decarboxylation of the carboxylic acid occurs in the presence of
a Lewis acid (here Ti(IV)), comparable to the Ruff−Fenton
70
oxidative degradation of aldonic acids with H O , usually
2
2
carried out with Cu or Fe catalysts in solution, although one
71
case with Ti-zeolites has been also described.
Several additional tests were performed in order to ascertain
the role of Ti(IV) centers in the reaction. First, titanium(IV)
i
isopropoxide, Ti(O Pr) , (a common Ti precursor for
4
enantioselective epoxidation reactions in solution) was used
as a homogeneous catalyst in the presence of trans-stilbene and
tested in α,α,α-trifluorotoluene, acetonitrile, and toluene. No
detectable reactant conversion was recorded both in the
presence and in the absence of 3A drying molecular sieves,
i
as, in principle, Ti(O Pr) might be degraded or deactivated by
4
traces of moisture in the reaction mixture.
Figure 7. Reaction of diphenylacetaldehyde (6) with TBHP (2 equiv)
In the same way, the Ti-free siliceous MCM41 support was
unable to epoxidize trans-stilbene. An average stilbene
conversion of 10% was obtained, but with a negligible
formation of stilbene epoxide (3) and no detectable formation
of benzophenone (5). On the contrary, its intrinsic very weak
Lewis acidity conferred some activity in the epoxide rearrange-
ment (5% conversion to aldehyde 6 from epoxide 3 after 1 h)
in the absence of oxidant, although much lower than the one
recorded over Ti-MCM41 under the same conditions (89%
conversion after 1 h). These results show the very low activity
of the support in both reactions and confirm the multifunc-
tional role of Ti sites.
catalyzed by Ti-MCM41 (35 mg), in PhCF at 102 °C. The numbers
correspond to the different reaction products (see Scheme 2).
3
(
(
10) was thus obtained, indicating than the precursor ketone
9) is produced by migration of a phenyl group onto the
coordinated aldehyde through a unexpected reaction pathway
dashed arrow from 6 to 9 in Scheme 2) and that, in the
(
presence of a low amount of oxidant, this reaction efficiently
competes with the oxidation pathway.
The rearrangement of epoxides to carbonyl compounds
Finally, the heterogeneous nature of the catalysts in this
multistep transformation was confirmed by a series of tests, in
which the solid catalysts were thoroughly removed from the hot
liquid mixture by high-speed centrifugation. These tests proved
that no leaching of homogeneous and catalytically active
titanium species took place under the reaction conditions tested
(see the Supporting Information).
Thus, the observed reactivity is attributed to the peculiar
chemical environment of evenly dispersed Ti(IV) sites
deposited onto a silica matrix rather than to the presence of
(
Meinwald rearrangement) is a reaction promoted by Brønsted
and Lewis acids. Several aspects of this reaction, including its
mechanism, have been recently studied.
67,68
Starting from
stilbene oxide, the reaction leads to diphenylacetaldehyde (6)
as the kinetically favored product through the migration of a
phenyl group from the intermediate zwitterion. No traces of
deoxybenzoin (9) are detected, because of the higher activation
energy of the hydrogen migration. Our results agree with the
kinetic preference of the aldehyde. However, they show the
reversibility of this rearrangement, leading to the ketone from
68
3
558
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
Ti(IV) itself in the reaction mixture and the contribution of the
purely siliceous supports to reactivity is negligible.
Scheme 6. Ti-Catalyzed Reaction of Triphenylethylene with
tert-Butyl Hydroperoxide
Reaction with Other Substituted Styrenes. In the case
of α-phenylstyrene (11), the higher stability of the carbocation
makes easier the rearrangement to the aldehyde through
hydrogen migration. Under the same conditions, 100% yield of
benzophenone was again obtained in 24 h (Scheme 3).
Scheme 3. Ti-Catalyzed Reaction of α-Phenylstyrene with
tert-Butyl Hydroperoxide
both oxidant and acid characteristics. In the present case,
titanium centers on the mesoporous silica support were, at the
same time, efficient oxidation sites that activate tert-butyl
hydroperoxide and good Lewis acids that trigger epoxide
rearrangement (Meinwald reaction). In this way, the catalyst
promotes a tandem one-pot, four-step sequence of reactions,
namely alkene epoxidation, epoxide rearrangement, oxidation
of the obtained aldehyde, and finally oxidative decarboxylation
of the carboxylic acid. This behavior can be optimized by the
use of a suitable solvent with polar aprotic character, under
anhydrous conditions, to favor the oxidation reactions, and with
a poor coordinating ability, to prevent any deactivation of the
Lewis acid character of the Ti(IV) sites. Moreover, high
temperatures above 100 °C are an additional requirement to
run the tandem reaction efficiently in reasonable times.
The number of solvents that fulfill all those requirements is
limited, and fluorinated aprotic solvents are a good choice in
this regard. In particular, α,α,α-trifluorotoluene has been
selected as the optimal solvent in this work to perform the
tandem oxidation−rearrangement sequence on stilbenes (cis
and trans), substituted styrenes (α-phenylstyrene, trans-β-
methylstyrene, α-methylstyrene), and triphenylethylene. Such
solvent is easily commercially available and is being considered
as a more sustainable and less hazardous alternative to
chlorinated solvents.
We have also explored the oxidation of trans-β-methylstyrene
12) (Scheme 4) and, from it, phenylacetone (14) was
(
Scheme 4. Ti-Catalyzed Reaction of trans-β-Methylstyrene
with tert-Butyl Hydroperoxide
obtained as the major product, indicating that migration of
hydrogen is favored over methyl migration from the
intermediate epoxide (13). 2-Phenylpropanal (15) and the
final decarboxylated product, acetophenone (16), were also
detected in minor amounts, as side products.
Starting from α-methylstyrene (17) the reaction was slower,
although total conversion was reached at the end of the test.
Acetophenone (16), with one carbon atom less than the
starting reactant, was the major product with 72% selectivity,
and 2-phenylpropionic acid (19) accounted for the rest of 28%
selectivity (Scheme 5).
Lastly, starting from triphenylethylene (20) the reaction
proceeded with a slower rate too, probably because of steric
hindrance. It should be noted that hydrogen migration is
indeed favored and, after 24 h, a 55% conversion was reached
with a benzophenone (5) to triphenylacetic acid (23) ratio of
In the best case, trans- or cis-stilbene was completely
(conversions >98%) and selectively (selectivity >96%) trans-
formed into benzophenone in 24 h, as the unique product.
Furthermore, in the cases of α-phenylstyrene, α-methylstyrene,
and triphenylethylene, the result is the same as that of an
ozonolysis, but under far safer and milder conditions.
EXPERIMENTAL DETAILS
■
Materials. Reagents were purchased from Sigma-Aldrich
and used as received without further purification. The solvents
were thoroughly dried over activated drying 3A molecular
sieves (Sigma-Aldrich). Anhydrous tert-butylhydroperoxide
(TBHP; Aldrich, 5.5 M solution in decane) was used as
oxidant.
1
0:1 (Scheme 6). It is worth highlighting that the result of this
reaction is similar to an ozonolysis, as in the case of α-
methylstyrene. However, following the present experimental
protocol, CC bond cleavage can be obtained under far milder
and simpler conditions.
Ti-MCM41, Ti-HMS, and Ti-SBA15 were prepared by
grafting Ti(Cp) Cl on the corresponding support following
2
2
CONCLUSIONS
Titanium sites on Ti(IV)-silica catalysts, prepared by grafting
technique, may act as multifunctional catalytic centers, with
32
■
the method reported in the literature. Ti-SiO
2
was prepared
by grafting Ti(OiPr) on Merck silica following a described
4
28
procedure.
Scheme 5. Ti-Catalyzed Reaction of α-Methylstyrene with tert-Butyl Hydroperoxide
3
559
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
Catalyst Characterization. Titanium content of the
prepared catalysts was determined by inductively coupled
plasma-optical emission spectroscopy on a ICAP 6300 Duo
Description of the heterogeneity tests and character-
ization data (XRD patterns and DRS-UV−vis spectra) of
supports and catalysts (PDF)
(Thermo Fisher Scientific) instrument after mineralization of
the samples with a mixture of hydrofluoric (aq 40%) and
fuming nitric acid.
AUTHOR INFORMATION
■
Corresponding Author
E-mail: jmfraile@unizar.es.
Present Address
F.G.S.: Ecole Polytechnique Federale de Lausanne, Labo-
́ ́
ratoire de Spectroscopie Ultrarapide, ISIC, Faculte
de Base, CH-1015 Lausanne, Switzerland.
Thermogravimetric analysis (TGA) was performed on a
high-temperature thermal balance (PerkinElmer; 7HT−Pyris
Manager) in the temperature range between 50 and 1000 °C,
*
§
−1
under dry air with a temperature rate of 5 °C min .
The textural characterization of the mesoporous materials
was performed on a Micromeritics ASAP 2010 instrument. The
́
des Sciences
Notes
calcined samples were outgassed at 250 °C, until a stable static
vacuum of 3 × 10−3 Torr was reached. Mesopore diameters
The authors declare no competing financial interest.
were calculated from the desorption branch of the nitrogen
isotherms by the Broekhoff and de Boer (BdB) method.
Catalytic Tests. The solid catalysts (50.0 mg or 35 mg in
the case of the studies of products evolution with time) were
pretreated prior to use in the catalytic reactions. Ti-MCM41,
Ti-HMS, and Ti-SBA15 were activated under argon flow,
raising the temperature from room temperature to 300 °C at a
ACKNOWLEDGMENTS
Financial support from the Spanish Ministerio de Economia y
Competitividad (Project CTQ2011-28124) and the Diputacion
́
General de Aragon (E11 Group cofinanced by the European
Regional Development Funds) is gratefully acknowledged.
M.G. thanks the European Research Institute of Catalysis
(ERIC; http://www.eric-aisbl.eu/) for a short-mobility grant.
The authors are grateful to Dr. Anne Galarneau and Marie-
France Driole (Institute Charles Gerhardt, CNRS Montpellier)
for providing the samples of mesoporous silica supports.
■
́
́
−1
rate of 10 °C min , 1 h at 300 °C, and then cooling under
vacuum. In the case of Ti-SiO , the catalyst was activated at 140
2
°
C under vacuum for 12 h.
The pretreated catalyst was added to a solution of alkene
(
0.488 mmol) and mesitylene (0.323 mmol, 45 μL, as internal
REFERENCES
■
standard) in 10 mL of dry solvent. The suspension was stirred
for 3 min and the zero-point sample was withdrawn. After that,
tert-butyl hydroperoxide (2.93 mmol, 532 μL) was added and
the reaction was stirred at the required temperature. The
reaction was monitored by GC (FID detector), using a cross-
linked methyl silicone column (HP1; 30 m × 0.25 mm × 0.35
(
1
1) Lemieux, R. U.; Von Rudloff, E. Can. J. Chem. 1955, 33, 1701−
709.
2) Rajagopalan, A.; Lara, M.; Kroutil, W. Adv. Synth. Catal. 2013,
(
3
55, 3321−3325.
(3) Lin, R.; Chen, F.; Jiao, N. Org. Lett. 2012, 14, 4158−4161.
(4) Sato, K.; Aoki, M.; Noyori, R. Science 1998, 281, 1646−1647.
−1
(
5) Zhu, W.; Li, H.; He, X. Catal. Commun. 2008, 9, 551−555.
μm), oven temperature program: 80 °C (0 min), 20 °C min
−1
(6) Ren, S.; Xie, Z.; Cao, L.; Xie, X.; Qin, G.; Wang, J. Catal.
Commun. 2009, 10, 464−467.
7) Wen, Y.; Wang, X.; Wei, H.; Li, B.; Jin, P.; Li, L. Green Chem.
012, 14, 2868−2875.
8) Vafaeezadeh, M.; Hashemi, M. M. Catal. Commun. 2014, 43,
to 180 °C (0 min), and 5 °C min to 250 °C (2 min).
Products were identified by GC-MS and in some cases by
comparison with commercial samples. In order to determine
the isolated yield of benzophenone, 2.0 mmol of trans-stilbene
were reacted over 200 mg of Ti-MCM41 catalyst, in α,α,α-
trifluorotoluene, at solvent reflux, for 24 h, with 6.0 mmol of
TBHP, in the absence of the internal standard. The solvent was
removed from the raw reaction mixture at the rotary
evaporator, whereas tert-butanol and the traces of unreacted
TBHP were eliminated by conventional liquid chromatography
on a short silica-gel column.
(
2
(
169−172.
(9) Bilis, G.; Christoforidis, K. C.; Deligiannakis, Y.; Louloudi, M.
Catal. Today 2010, 157, 101−106.
(
2
(
1
10) Ray, R.; Chowdhury, A. D.; Lahiri, G. K. ChemCatChem 2013, 5,
158−2161.
11) Shaikh, T. M.; Hong, F.-E. Adv. Synth. Catal. 2011, 353, 1491−
496.
(12) Hughes, M. D.; Xu, Y. J.; Jenkins, P.; McMorn, P.; Landon, P.;
The solid catalysts were recovered and reused in two new
catalytic cycles. After each recovery, the materials were carefully
filtered, washed with fresh methanol and calcined under air flow
at 300 or 500 °C. Then, they underwent a new catalytic run as
described above.
Enache, D. I.; Carley, A. F.; Attard, G. A.; Hutchings, G. J.; King, F.;
Stitt, E. H.; Johnston, P.; Griffin, K.; Kiely, C. J. Nature 2005, 437,
1
(
3
(
132−1135.
13) Quek, X. Y.; Tang, Q.; Hu, S.; Yang, Y. Appl. Catal., A 2009,
61, 130−136.
14) Lapisardi, G.; Chiker, F.; Launay, F.; Nogier, J. P.; Bonardet, J.-
In order to check the leaching of catalytically active titanium
species, in a series of tests, after 1 h of reaction (at ca. 50% of
substrate conversion), the solid catalysts were removed from
the hot liquid mixture by high-speed centrifugation and the
resulting solutions were tested for further reaction (see also the
Supporting Information).
L. Catal. Commun. 2004, 5, 277−281.
(15) Lee, S. O.; Raja, R.; Harris, K. D. M.; Thomas, J. M.; Johnson, B.
F. G.; Sankar, G. Angew. Chem., Int. Ed. 2003, 42, 1520−1523.
(16) Reddy, J. S.; Khire, U. R.; Ratnasamy, P.; Mitra, R. B. J. Chem.
Soc., Chem. Commun. 1992, 1234−1235.
17) Zhuang, J.; Ma, D.; Yan, Z.; Liu, X.; Han, X.; Bao, X.; Zhang, Y.;
Guo, X.; Wang, X. Appl. Catal., A 2004, 258, 1−6.
18) Nie, L.; Xin, K. K.; Li, W. S.; Zhou, X. P. Catal. Commun. 2007,
8, 488−492.
19) Kholdeeva, Q. A.; Trukhan, N. N. Russ. Chem. Rev. 2006, 75,
11−432.
(20) Fraile, J. M.; García, J. I.; Mayoral, J. A.; Proietti, M. G.; Sanchez,
(
(
ASSOCIATED CONTENT
■
(
4
*
S
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/cs501671a.
́
M. C. J. Phys. Chem. 1996, 100, 19484−19488.
3
560
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561

ACS Catalysis
Research Article
(
1
(
21) Maschmeyer, T.; Rey, F.; Sankar, G.; Thomas, J. M. Nature
995, 378, 159−162.
22) Marchese, L.; Gianotti, E.; Dellarocca, V.; Maschmeyer, T.; Rey,
F.; Coluccia, S.; Thomas, J. M. Phys. Chem. Chem. Phys. 1999, 1, 585−
92.
23) Guidotti, M.; Ravasio, N.; Psaro, R.; Ferraris, G.; Moretti, G. J.
Catal. 2003, 214, 242−250.
24) Fraile, J. M.; García, J. I.; Mayoral, J. A.; Vispe, E. J. Catal. 2001,
04, 146−156.
25) Fraile, J. M.; García, J. I.; Mayoral, J. A.; Vispe, E. Appl. Catal., A
003, 245, 363−376.
26) Guidotti, M.; Pirovano, C.; Ravasio, N.; Laz
Mayoral, J. A.; Coq, B.; Galarneau, A. Green Chem. 2009, 11, 1421−
427.
27) Adam, W.; Corma, A.; Martínez, A.; Renz, M. Chem. Ber. 1996,
29, 1453−1455.
28) Cativiela, C.; Fraile, J. M.; García, J. I.; Mayoral, J. A. J. Mol.
bonds. EN is the normalized form of E (30), a solvent polarity
T
T
parameter, proposed by Dimroth and Reichardt, based on the
solvatochromic shift of a betaine dye in solution (Reichardt, C.
Solvents and Solvent Effects in Organic Chemistry; VCH: Weinheim,
1988). From IUPAC, this parameter reflects the ionizing power of the
solvent, that is the tendency of a particular solvent to promote
5
(
ionization of an uncharged or, less often, charged solute. In fact, for
N
T
(
2
(
2
protic (hydrogen bond donor, HBD) solvents the value of E
increases because this parameter reflects a mixture of polarity and
HBD character of the solvent.
(54) Fraile, J. M.; García, J. I.; Mayoral, J. A.; Vispe, E.; Brown, D. R.;
(
́
aro, B.; Fraile, J. M.;
Fuller, G. J. Phys. Chem. B 2003, 107, 519−526.
(55) Bhaumik, A.; Tatsumi, T. J. Catal. 1999, 182, 349−356.
(56) Gutmann, V. Coord. Chem. Rev. 1976, 18, 225−255.
(57) Marcus, Y. J. Solution Chem. 1984, 13, 599−624.
1
(
1
(
(58) Reichardt, C. Chem. Rev. 1994, 94, 2319−2358.
(59) Maul, J. J.; Ostrowski, P. J.; Ublacker, G. A.; Linclau, B.; Curran,
D. P. Benzotrifluoride and Derivatives: Useful Solvents for Organic
Synthesis and Fluorous Synthesis. Modern Solvents in Organic Synthesis;
Topics in Current Chemistry; Springer, 1999; Vol. 206, pp 79−106;
DOI: 10.1007/3-540-48664-x_4.
Catal. A: Chem. 1996, 112, 259−267.
(
29) Iwamoto, M.; Tanaka, Y. Catal. Surv. Jpn. 2001, 5, 25−36.
(
30) Liu, J.; Wang, F.; Qi, S. S.; Gua, Z.; Wu, G. New J. Chem. 2013,
3
(
7, 769−774.
31) Gianotti, E.; Bisio, C.; Marchese, L.; Guidotti, M.; Ravasio, N.;
Psaro, R.; Coluccia, S. J. Phys. Chem. C 2007, 111, 5083−5089.
32) Oldroyd, R. D.; Thomas, J. M.; Maschemeyer, T.; MacFaul, P.
(
60) Aldea, L.; Fraile, J. M.; García, J. I.; García-Marín, H.; Mayoral, J.
A.; Perez, P. Green Chem. 2010, 12, 435−440.
61) Fraile, J. M.; García, J.; Mayoral, J. A.; Proietti, M. G.; San
M. C. J. Phys. Chem. 1996, 100, 19484−19488.
62) Han, J. H.; Hong, S. J.; Lee, E. Y.; Lee, J. H.; Kim, H. J.; Kwak,
H.; Kim, C. Bull. Korean Chem. Soc. 2005, 26, 1434−1436.
63) The evolution of CO from the reaction medium was confirmed
́
(
́
chez,
(
A.; Snelgrove, D. W.; Ingold, K. U.; Wayner, D. D. D. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 2787−2790.
(
(
33) Capel-Sanchez, M. C.; Campos-Martin, J. M.; Fierro, J. L. G.; de
(
Frutos, M. P.; Padilla Polo, A. Chem. Commun. 2000, 855−856.
34) Tiozzo, C.; Bisio, C.; Carniato, F.; Gallo, A.; Scott, S. L.; Psaro,
R.; Guidotti, M. Phys. Chem. Chem. Phys. 2013, 15, 13354−13362.
35) Gianotti, E.; Frache, A.; Coluccia, S.; Thomas, J. M.;
Maschmeyer, T.; Marchese, L. J. Mol. Catal. A: Chem. 2003, 204−
05, 483−489.
36) Fraile, J. M.; García, J. I.; Mayoral, J. A.; Vispe, E. J. Catal. 2005,
2
by N bubbling and trapping the effluent gas in aqueous Ba(OH)
(
2
2
solution. The gravimetric analysis of the formed BaCO was roughly
3
consistent with the number of moles of CO expected (0.74 mol of
(
2
CO per mol of benzophenone found vs 1.0 mol expected).
2
(
64) Guidotti, M.; Ravasio, N.; Psaro, R.; Ferraris, G.; Moretti, G. J.
Catal. 2003, 214, 242−250.
65) Guidotti, M.; Ravasio, N.; Psaro, R.; Gianotti, E.; Coluccia, S.;
Marchese, L. J. Mol. Catal. A: Chem. 2006, 250, 218−225.
66) The role of geometrical constraints on the acidity of Al sites on
2
(
(
2
(
33, 90−99.
37) Liu, Z.; Davis, R. J. J. Phys. Chem. 1994, 98, 1253−1261.
38) Kholdeeva, O. A.; Maksimov, G. M.; Fedotov, M. A.; Grigoriev,
(
(
silica has been demonstrated by both experimental and theoretical
methods, with higher acidity of the tripodal species: Fraile, J. M.;
García, J. I.; Mayoral, J. A.; Pires, E.; Salvatella, L.; Ten, M. J. Phys.
Chem. B 1999, 103, 1664−1670.
V. A. React. Kinet. Catal. Lett. 1994, 53, 331−337.
39) Gajan, D.; Guillois, K.; Delichere, P.; Basset, J. M.; Candy, J. P.;
Caps, V.; Coperet, C.; Lesage, A.; Emsley, L. J. Am. Chem. Soc. 2009,
31, 14667−14669.
40) Surya Prakash, G. K.; Mathew, T.; Krishnaraj, S.; Marinez, E. R.;
Olah, G. A. Appl. Catal., A 1999, 181, 283−288.
41) Garrison, J. M.; Bruice, T. C. J. Am. Chem. Soc. 1989, 111, 191−
98.
42) Stultz, L. K.; Binstead, R. A.; Reynolds, M. S.; Meyer, T. J. J. Am.
(
́
1
(
(
67) Fraile, J. M.; Mayoral, J. A.; Salvatella, L. J. Org. Chem. 2014, 79,
993−5999.
68) Jamalian, A.; Rathman, B.; Borosky, G. L.; Laali, K. K. Appl.
5
(
(
1
(
Catal., A 2014, 486, 1−11.
(
(
18.
69) Havare, N.; Plattner, D. A. Org. Lett. 2012, 14, 5078−5081.
70) Stapley, J. A.; BeMiller, J. N. Carbohydr. Res. 2007, 342, 407−
Chem. Soc. 1995, 117, 2520−2532.
(
(
4
43) Muzart, J.; Ajjou, A. New J. Chem. 1994, 18, 731−736.
44) Ko, N. H.; Kim, H. S.; Kang, S. O.; Cheong, M. Bull. Korean
(71) Hourdin, G.; Germain, A.; Moreau, C.; Fajula, F. Catal. Lett.
2
000, 69, 241−244.
Chem. Soc. 2008, 29, 682−684.
45) Holtz, H. D. Producing benzophenone and related products. US
Patent 3,899,537, 1975.
46) Van der Waal, J. C.; Van Bekkum, H. J. Mol. Catal. A: Chem.
997, 124, 137−146.
47) Koohestani, B.; Ahmad, A. L.; Bhatia, S.; Ooi, B. S. Current
Nanosci. 2011, 7, 781−789.
48) Wang, G. J.; Liu, G. Q.; Xu, M. X.; Yang, Z. X.; Liu, Z. W.; Liu,
Y. W. Appl. Surf. Sci. 2008, 255, 2632−2640.
49) Tang, B.; Lu, X.-H.; Zhou, D.; Lei, J.; Niu, Z.-H.; Fan, J.; Xia, Q.-
(
(
1
(
(
(
H. Catal. Commun. 2012, 21, 68−71.
(50) Saux, C.; Pierella, L. Appl. Catal., A 2011, 400, 117−121.
(51) Pardeshi, S. K.; Pawar, R. Y. Mater. Res. Bull. 2010, 45, 609−615.
(52) Kholdeeva, O. A. In Liquid Phase Oxidation via Heterogeneous
Catalysis; Clerici, M. G.; Kholdeeva, O. A., Eds.; Wiley: NJ, 2013; Ch.
4
.
(
53) Solvent polarity is a commonly used term that refers to the
ability of a solvent for solvating charged or dipolar species. This
concept involves several properties of the solvent, such as dipolar
moment, but also specific molecular interactions, such as hydrogen
3
561
DOI: 10.1021/cs501671a
ACS Catal. 2015, 5, 3552−3561