Cyclopentadienyl-silsesquioxane titanium complexes: Highly active catalysts for epoxidation of alkenes with aqueous hydrogen peroxide

Article abstract of DOI:10.1021/ic300583g

Titanium complexes bearing an unprecedented tridentate cyclopentadienyl- silsesquioxanate ligand provide a new class of efficient and selective catalysts for epoxidation of olefins with aqueous hydrogen peroxide under homogeneous conditions.

Full text of DOI:10.1021/ic300583g

Article
pubs.acs.org/IC
Cyclopentadienyl−Silsesquioxane Titanium Complexes: Highly
Active Catalysts for Epoxidation of Alkenes with Aqueous Hydrogen
Peroxide
María Ventura,† Marta E. G. Mosquera,†,§ Tomas Cuenca,*,† Beatriz Royo,‡ and Gerardo Jimenez*,†
́
́
†Departamento de Química Inorgan
́
ica, Universidad de Alcala,
́ ́
Campus Universitario, 28871 Alcala de Henares, Spain
‡Instituto de Tecnologia Química e Biolog
́
ica, Universidade Nova de Lisboa Av. de Republica, EAN, 2780-417 Oeiras, Portugal
́
S
* Supporting Information
ABSTRACT: Titanium complexes bearing an unprecedented tridentate cyclopentadienyl−silsesquioxanate ligand provide a new
class of efficient and selective catalysts for epoxidation of olefins with aqueous hydrogen peroxide under homogeneous
conditions.
poxidation of olefinic compounds is a fundamental
E
reaction in industrial organic synthesis,1 because epoxides
are key intermediates for the manufacture of a wide variety of
valuables products, both bulk2 and fine chemicals.3 Despite the
enormous research effort dedicated to the development of
efficient catalysts for the oxidation of olefins, some problems
remain, in particular, those concerning the nature of the oxidant
and the product selectivity. Thus, the search for efficient,
selective, and environmentally benign catalysts for these
oxidation reactions is currently an important synthetic goal.4
In this context, the development of environmentally benign
processes based on nontoxic metals, such as titanium,5 and clean
oxidants deserves special emphasis.6 Among the latter, hydrogen
peroxide is particularly attractive, because it is inexpensive and
environmentally friendly, since it produces water as the only side
product.7
Figure 1. Tridentate cyclopentadienyl−silsesquioxanate ligand.
amount of incompletely condensed silsesquioxane trisilanol,
i
R7Si7O9(OH)3 (R = Bu, 2a; Ph, 2b), in chlorinated solvents
affords the desired derivative Ti(η5-C5Me4SiMe2OR7Si7O11-
κ2O2)Cl (R = iBu, 3a; Ph, 3b) as the sole product (see Scheme 1).
The formation of side products is not observed. These complexes
comprise an unprecedented tridentate ligand cyclopentadienyl−
silsesquioxane disilanolate coordinated to titanium, which involves
the condensation of both functionalities with the consequent
elimination of three equivalents (3 equiv) of HCl.
Although the formation of complexes 3 globally entails
rupturing the Si−Cl bond, as well as two Ti−Cl bonds, such a
transformation does not occur via the direct protonolysis of the
Si−Cl bond, as inferred from the result attained when the
reaction with 2a is carried out in aromatic solvents. In this
reaction, the formation of complex 3a, along with a new corned-
capped cyclopentadienyl derivative Ti(η5-C5Me4SiMe2Cl)-
(iBu7Si7O12-κ3O3) (4a), which is the result of the rupture of
the three Ti−Cl bonds, is detected. Nevertheless, upon heating
the reaction mixture, 4a is quantitatively converted into 3a (see
Scheme 2, as seen by NMR spectroscopy). Therefore, it is
reasonable to assume that complex 4a is the initially formed
compound (kinetic control product) while complex 3a is the
thermodynamic product.13
Metal−silsesquioxane complexes have been the focus of
intensive attention, as well-defined molecular models for
siliceous heterogeneous catalysts.8 Metallasilsesquioxanes not
only offer an unique opportunity to reach an understanding of
metallasilicate catalysts at a molecular level but they also show
an interesting catalytic behavior by themselves.9 In this regard, a
series of titanium silsesquioxane complexes have been reported
to catalyze alkene epoxidation with organic peroxides, exhibit-
ing activities that are comparable, or even superior, to those
shown by the heterogeneous analogues.10 However, the titan-
asilsesquioxanes reported so far feature an important drawback
for its development as homogeneous oxidation catalysts, since
none are capable of performing alkene epoxidation with
aqueous hydrogen peroxide as the oxidant.11
In view of this, our target was synthesizing a novel family of
Ti catalyst based on silsesquioxanes, robust enough to be used
as catalysts with aqueous hydrogen peroxide. We envisioned
that assembling a silsesquioxane fragment with a cyclopentadienyl
ring would form unprecedented tridentate cyclopentadienyl−
silsesquioxanate ligand (see Figure 1) capable to form highly
stable titanium complexes.
Following the synthetic strategy developed in our research
group,12 reaction of the titanium monocyclopentadienyl
derivative Ti(η5-C5Me4SiMe2Cl)Cl3 (1) with an equimolar
Received: March 20, 2012
Published: May 17, 2012
© 2012 American Chemical Society
6345
dx.doi.org/10.1021/ic300583g | Inorg. Chem. 2012, 51, 6345−6349

Inorganic Chemistry
Article
Scheme 1. Synthetic Route to Titanium Complexes 3
Scheme 2. Formation of the Mixture of 3a and 4a
All attempts to isolate 4a were unfortunately unfruitful;
nonetheless, its structural disposition has been unambiguously
formulated by spectroscopic analysis of a sample containing 3a
and 4a.
The spectroscopic behavior of compounds 3 and 4a supports
the proposed structures. 29Si NMR spectra are particularly
informative, indicating a Cs- and a C3v-symmetry disposition,
respectively. The absence of resonances for the silsesquioxane
framework Si atoms above −60 ppm14,15 reveals the
deprotonation and coordination of the three silanol functions.
The most striking spectroscopic features are the chemical shifts
of the SiMe2X moiety in the 1H and 29Si NMR spectra.
Following the tendency observed in our previous works,11
upfield shifted resonances for compounds 3 and downfield
shifted for 4a, compared with those found for 1, confirm the
proposed silsesquioxane cage connectivity within such com-
plexes.
The crystal structure of 3a has been verified by X-ray
diffraction (XRD) study (Figure 2), showing a monomeric
disposition with a pseudo-tetrahedral coordination geometry
around the Ti atom. The silsesquioxane framework is clearly
linked to the cyclopentadienyl ring by one silanolate group,
giving rise to the cyclopentadienyl−silsesquioxane moiety,
which acts as a tridentate ligand coordinating to the Ti atom
through the Cp- and two silsesquioxane silanolate groups. The
condensation of both functionalities does not bring about
relevant distortions within the cyclopentadienyl coordination
manner to titanium, which may be attributed to the flexibility of
the silsesquioxane cage that makes it possible to accommodate
the steric demands of the “Ti(C5Me4SiMe2)Cl” fragment.
The cyclopentadienyl−silsesquioxane complexes 3 efficiently
catalyze the epoxidation of both cyclic and linear alkenes with
aqueous hydrogen peroxide (see Table 1), under mild
conditions.
Figure 2. Perspective ORTEP plot of 3a showing 30% thermal
ellipsoids. Hydrogen atoms and iso-butyl methyl groups are omitted
for clarity. Selected bond distances and angles: Ti(1)−O(2), 1.809(3)
Å; Ti(1)−O(3), 1.810(3) Å; Si(1)−O(1), 1.634(3) Å; O(2)−Ti(1)−
O(3), 105.49(13)°; O(2)−Ti(1)−Cl(1), 101.30(10)°; and O(3)−
Ti(1)−Cl(1), 101.92(11)°.
The efficiency of 3a and 3b as catalysts for the epoxidation of
cyclooctene with H2O2 (30% in water) has been investigated at
55 °C in MeCN (in a substrate:oxidant molar ratio of 1:2).
Complex 3b (0.5 mol %) affords quantitative conversion
(>99%) to the corresponding epoxide in 16 h (Table 1, entry 3).
Under similar conditions, complex 3a results to be less active,
affording an epoxide yield of 71% in 20 h of reaction (Table 1,
entry 1). Increasing the temperature accelerates the catalytic
reactions. At 70 °C, quantitative conversions are obtained within
only 8 h, achieving turnover frequencies (TOF) of 124 (for 3a)
and 160 (for 3b) h−1 (calculated at 30 min). The catalytic
6346
dx.doi.org/10.1021/ic300583g | Inorg. Chem. 2012, 51, 6345−6349

Inorganic Chemistry
Article
a
Table 1. Olefin Epoxidation with H2O2 Catalyzed by 3
b
entry
catalyst
substrate
temperature, T (°C)
% cat.
time, t (h)
yield (%)
turnover number, TON
1
2
3
4
5
6
7
8
9
3a
3a
3b
3b
3b
3b
3b
3a
3b
cyclooctene
cyclooctene
cyclooctene
cyclooctene
cyclooctene
cyclooctene
cyclooctene
1-octene
55
70
55
55
70
70
70
70
70
0.5
0.5
0.5
0.5
0.5
0.2
0.1
0.5
0.5
20
8
71
88
142
176
198
156
198
290
300
136
190
16
24
8
>99
78
c
>99
58
24
24
8
30
68
1-octene
6
91
a
b
Unless noted otherwise, all reactions were performed in MeCN with a substrate:oxidant molar ratio of 1:2. Yield determined by gas
c
chromatography (GC), using mesitylene as an internal standard. Reaction was carried out in methanol.
amount of catalyst can be reduced to 0.1 mol %, achieving turnover
number (TON) values of 300 for catalyst 3b. Interestingly, the
somewhat-difficult epoxidation of 1-octene is effectively catalyzed
by complexes 3a and 3b, at 70 °C, affording epoxide yields of 68%
and 91% in 8 and 6 h, respectively (see Table 1, entries 8 and 9).
All catalytic reactions are entirely selective for epoxidation, with the
corresponding epoxides (1,2-epoxycyclooctane and 1-octene oxide)
being the only products obtained. The use of methanol as a
solvent, instead of acetonitrile, does not improve the catalytic
performance of complexes 3, as shown in Table 1 (entry 4).
The lower activity of 3a, relative to 3b, can be rationalized in
terms of the bigger steric bulk of the silsesquioxane silicon
substituents in 3a (iso-butyl) than in 3b (phenyl), which
hinders the accessibility of the titanium center to the reactants
(alkene and oxidant) in the former complex.
The high efficiency of catalysts 3a and 3b is remarkable, since
Ti silsesquioxane complexes have been reported to be inactive
The kinetic profiles of 3a and 3b show that the conversion of
cyclooctene is initially relatively fast, and no induction periods
are observed, after which the reaction slows down over the
course of time (Figure 3).
Figure 4. Kinetic profiles of cyclooctene epoxidation with H2O2 using
catalyst 3b. Reactions were carried out in MeCN at 70 °C, with a
substrate:oxidant molar ratio of 1:2, and a catalyst content of 5 mol %.
for epoxidation olefins with H2O2 under homogeneous
conditions. To the best of our knowledge, these are the first
homogeneous systems based on Ti silsesquioxane species that
catalyze the epoxidation of olefins with aqueous hydrogen
peroxide.
In summary, a new class of robust titanium cyclo-
pentadienyl−silsesquioxane compounds have been prepared
and used in catalytic epoxidation using H2O2. The novel
titanium complexes show excellent reactivity and selectivity
toward both cyclic and lineal alkenes. The remarkable stability
of these catalysts has been proved by their good performance in
successive epoxidations carried out by recharging the reaction
mixture with new charges of oxidant and substrate.
Figure 3. Kinetic profiles of cyclooctene epoxidation with H2O2 in the
presence of complexes 3a and 3b.
EXPERIMENTAL SECTION
■
General Considerations. All manipulations were performed
under an argon atmosphere, using Schlenk and high-vacuum line
techniques. The solvents were dried and purified with an MBRAUN
solvent purification system. Deuterated solvents were stored over
activated 4 Å molecular sieves and degassed by several freeze−thaw
cycles. NEt3 (Aldrich) was distilled before use and stored over 4 Å
molecular sieves. Silsesquioxane trisilanol R7Si7O9(OH)3 was
purchased from commercial sources (Hybrid Plastics) and used
The catalyst 3b can be used for at least three batches of olefin
epoxidation. Second and third runs have been performed by
recharging the reaction mixture with the same amount of
oxidant and substrate used initially in the first run. The kinetic
curves for the first and second runs are practically coincident,
giving quantitative conversions at similar reaction times.
However, longer reaction time is needed in the third run to
achieve 100% yield of epoxide (Figure 4). Thus, it seems that
the systems formed by 3b/H2O2/MeCN are fairly stable under
catalytic conditions.
16
without further purification. Ti(η5-C5Me4SiMe2Cl)Cl3 was prepared
via a known procedure. C, H, and N microanalyses were performed on
a Perkin−Elmer Model 240B and/or Heraeus CHN-O-Rapid
microanalyzer. Nuclear magnetic resonance (NMR) spectra, at
25 °C, were recorded on a Bruker Model AV400 (1H NMR system
6347
dx.doi.org/10.1021/ic300583g | Inorg. Chem. 2012, 51, 6345−6349

Inorganic Chemistry
Article
at 400 MHz, a 13C NMR system at 100.6 MHz, and a 29Si NMR
system at 79.5 MHz).
and hydrogen atom coordinates in CIF format of complex 3a.
1
1H and H−29Si HMBC spectra for 3a, 3b, and the mixture of
Synthesis of Ti(η5-C5Me4SiMe2OiBu7Si7O11-κ2O2)Cl (3a). A solution
of NEt3 (0.34 mL, 2.45 mmol) in CH2Cl2 (30 mL) was added to a
mixture of Ti(η5-C5Me4SiMe2Cl)Cl3 (0.3 g, 0.8 mmol) and
iBu7Si7O9(OH)3 (0.65 g, 0.8 mmol), finely mashed. After the reaction
mixture was stirred for 12 h, the white solid formed was collected by
filtration, and the volatiles were removed under vacuum. The residue
was extracted into hexane (3 × 15 mL) and the resultant solution
was concentrated (20 mL) and cooled at −20 °C to give 3a (0.56 g,
0.53 mmol, 75% yield) as a microcrystalline solid. Anal. Calcd for
C39H81ClO12Si8Ti: C 44.63, H 7.72; found: C 44.56, H 7.57. 1H NMR
(CDCl3, 400 MHz): δ 0.39 (s, 6H; SiMe2), 0.53, 0.95, 1.82 (m, 14H,
3a and 4a and chromatograms. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
§Correspondence concerning the crystallography data should
be addressed to this author. E-mail: martaeg.mosquera@uah.es.
*Tel.: 34918854767 (G.J.), 34918854655 (T.C.). Fax: 34
918854683. E-mail: gerardo.jimenez@uah.es (G.J.), tomas.
cuenca@uah.es (T.C.).
i
42H, 7H; Bu), 2.11, 2.24 (s, 2 × 6H; C5Me4). 13C NMR (CDCl3): δ
3.08 (SiMe2), 12.66, 14.48 (C5Me4), 22.69, 22.96, 23.30, 24.00, 24.08,
24.13, 24.40, 24.54, 25.80, 25.86, 25.94, 25.96, 25.99, 26.18 (iBu),
127.87, 134.49, 135.08 (C5Me4). 29Si NMR (CDCl3): δ −3.0 (SiMe2),
−66.7, −67.7, −67.8, −68.6, −69.8 (iBu7Si7O12).
Notes
The authors declare no competing financial interest.
Synthesis of Ti(η5-C5Me4SiMe2OPh7Si7O11-κ2O2)Cl (3b). A method
similar to that described for 3a was adopted, using Ph7Si7O9(OH)3
(1.27 g, 1.36 mmol) to give 3b (1.31 g, 1.10 mmol, 81% yield) as a
yellow solid. Anal. Calcd for C53H53ClO12Si8Ti: C 53.55, H 3.99;
found: C 53.48, H 4.45. 1H NMR (CDCl3, 400 MHz): δ 0.43
(s, 6H; SiMe2), 2.05, 2.28 (s, 2 × 6H; C5Me4), 7.27, 7.39, 7.66
(m, 14H, 7H, 14H, C6H5). 13C NMR (CDCl3): δ 3.13 (SiMe2),
12.87, 14.87 (C5Me4), 128.01, 128.14, 128.22, 128.58, 129.39,
130.55, 130.68, 130.80, 130.82, 131.47, 132.03, 134.14, 134.44,
134.49, 134.53, 134.56, 134.63, 135.90, 136.70 (C6H5, C5Me4). 29Si
NMR (CDCl3): δ −2.3 (SiMe2), −77.2, −78.0, −79.0, −79.7,
−79.9 (Ph7Si7O12).
ACKNOWLEDGMENTS
■
We are grateful to Spanish MCYT (Project No. CTQ2011-
23497) and Portuguese FCT (No. POCI 2010), and FEDER
(Project No. PTDC/QUI-QUI/098682/2008) for financial
support. M.V. thanks the University of Alcala
fellowship.
́
for a
REFERENCES
■
(1) Aziridines and Epoxides in Organic Synthesis; Yudin, A. K., Ed.;
Wiley−VCH: Weinheim, Germany, 2006.
(2) Weissermel, K.; Arpe, H.-J. In Industrial Organic Chemistry;
Wiley−VCH: Weinheim, Germany, 1997; pp 143−188.
(3) Sheldon, R. A.; van Bekkum, H. In Fine Chemicals through
Heterogeneous Catalysis; Wiley−VCH: Weinheim, Germany, 2001; pp
473−551.
(4) Al-Maksoud, W.; Daniele, S.; Sorokin, A. B. Green Chem. 2008,
10, 447−451.
(5) Wang, Y.; Wang, M.; Wang, L.; Wang, Y.; Wang, X.; Sun, L. Appl.
Organomet. Chem. 2011, 25, 325−330.
(6) Lane, B. S.; Burgess, K. Chem. Rev. 2003, 103, 2457−2473.
(7) Jones, C. W. Applications of Hydrogen Peroxide and Derivatives.
Catalytic; Clark, J. H., Ed.; Royal Society of Chemistry (RSC):
Cambridge, U.K., 1989.
(8) (a) Feher, F. J.; Budzichowski, T. A. Polyhedron 1995, 14, 3239−
3253. (b) Hanssen, R. W. J. M.; van Santen, R. A.; Abbenhuis, H. C. L.
Eur. J. Inorg. Chem. 2004, 675−683. (c) Severn, J. R.; Chadwick, J. C.;
Duchateau, R.; Friederichs, N. Chem. Rev. 2005, 105, 4073−4147.
(9) (a) Abbenhuis, H. C. L. Chem.Eur. J. 2000, 6, 25−32.
(b) Duchateau, R. Chem. Rev. 2002, 102, 3525−3542. (c) Duchateau,
R. Chem. Rev. 2002, 102, 3525−3542.
(10) (a) Abbenhuis, H. C. L.; Krijnen, S.; van Santen, R. A. Chem.
Commun. 1997, 331−332. (b) Crocker, M.; Herold, R. H. M.; Orpen,
A. G.; Overgaag, M. T. A. J. Chem. Soc., Dalton Trans. 1999, 3791−
3804.
(11) Skowronska-Ptasinska, M. D.; Vorstenbosch, M. L.; van Santen,
R. A.; Abbenhuis, H. C. L. Angew. Chem., Int. Ed. 2002, 41, 637−639.
(12) (a) Maestre, M. C.; Paniagua, C.; Herdtweck, E.; Mosquera, M.
Spectroscopic Data for Ti(η5-C5Me4SiMe2Cl)(iBu7Si7O12-κ3O3) (4a).
1H NMR (CDCl3, 400 MHz): δ 0.73 (s, 6H; SiMe2), 0.50, 0.94, 1.80
i
(m, 14H, 42H, 7H; Bu), 2.03, 2.25 (s, 2 × 3H; C5Me4). 29Si NMR
(CDCl3): δ 15.2 (SiMe2), −66.2, −66.3, −68.7 (iBu7Si7O12).
Single-Crystal X-ray Structure Determination of Compound
3a. Crystal data and details of the structure determination are
presented in the Supporting Information. Suitable single crystals of 3a
for the X-ray diffraction (XRD) study were selected, covered with
perfluorinates polyether oil, and mounted on a Bruker−Nonius Kappa
CCD single-crystal diffractometer. Data collection was performed at
200(2) K. The structure was solved, using the WINGX package,17 by
direct methods (SHELXS-97) and refined using full-matrix least-
squares against F2 (SHELXL-97).18 Al non-hydrogen atoms were
anisotropically refined. Hydrogen atoms were geometrically placed and
left riding on their parent atoms, except for the hydrogen atoms H1 an
H2 that were found in the Fourier map and refined. Full-matrix least-
2
squares refinements were carried out by minimizing ∑w(Fo − Fc2)2
with the SHELXL-97 weighting scheme and stopped at shift/err
<0.001.
Crystal Data for 3a. C39H81ClO12Si8Ti, M = 1050.11, triclinic,
space group P1, a = 11.0572(10) Å, b = 13.5568(8) Å, c =
̅
19.3775(14) Å, α = 92.099(5)°, β = 104.946(6)°, γ = 92.199(6)°, V =
2801.1(4) Å3. Z = 2, Dcalcd = 1.245 g cm−3, F(000) = 1124, λ(Mo Kα)
= 0.71073 Å. At the conclusion of the refinement, wR2(F2 > 2σ(F2) =
0.1391, conventional R [on F values for 12306 reflections] (F2
>
2σ(F2)) = 0.0599 for 558 parameters.
Alkene Epoxidation. The catalytic tests were carried out in a
reaction vessel equipped with a magnetic stirrer and immersed in an oil
bath at the appropriate temperature. A catalyst:alkene:H2O2 ratio of
0.5(0.2, 0.1):100:200 was used, with 2 mL of MeCN. Olefin,
acetonitrile, mesitylene (as internal standard), and the catalyst were
transferred into the reaction vessel, and H2O2 (30% aqueous solution)
was added to the mixture. The course of the reaction was monitored
by quantitative gas chromatography (GC) analysis.
E. G.; Jimen
(b) Maestre, M. C.; Mosquera, M. E. G.; Jacobsen, H.; Jimen
Cuenca, T. Organometallics 2008, 27, 839−849. (c) Paniagua, C.;
Mosquera, M. E. G.; Jacobsen, H.; Jimenez, G.; Cuenca, T.
Organometallics 2009, 28, 6975−6980. (d) Paniagua, C.; Mosquera,
M. E. G.; Cuenca, T.; Jimenez, G. Organometallics 2011, 30, 2993−
3000. (e) Ventura, M.; Ramírez de Arellano, C.; Mosquera, M. E. G.;
Jimen
́
ez, G.; Cuenca, T. Organometallics 2007, 26, 4243−4251.
́
ez, G.;
́
́
́
ez, G.; Cuenca, T. Dalton Trans. 2011, 40, 5728−5733.
ASSOCIATED CONTENT
■
(13) Ventura, M. Unpublished results.
S
* Supporting Information
(14) (a) Tanabe, M.; Mutou, K.; Mintcheva, N.; Osakada, K.
Organometallics 2008, 27, 519−523. (b) Carniato, F.; Boccaleri, E.;
Marchese, L. Dalton Trans. 2008, 36−39.
Tables of crystallographic data, including fractional coordinates,
bond lengths and angles, anisotropic displacement parameters,
6348
dx.doi.org/10.1021/ic300583g | Inorg. Chem. 2012, 51, 6345−6349

Inorganic Chemistry
Article
(15) Giovenzana, T.; Guidotti, M.; Lucenti, E.; Biroli, A. O.; Sordelli,
L.; Sironi, A.; Ugo, R. Organometallics 2010, 29, 6687−6694.
́
(16) Buitrago, O.; Jimenez, G.; Cuenca, T. J. Organomet. Chem. 2003,
683, 70−76.
(17) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837−938.
(18) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
2008, 64, 112−122.
6349
dx.doi.org/10.1021/ic300583g | Inorg. Chem. 2012, 51, 6345−6349