Catalytic deoxygenation of epoxides with (Cp*ReO)2(μ-O)2 and catalyst deactivation

Article abstract of DOI:10.1039/a708516h

In situ reduction of Cp*ReO₃ by PPh₃ to form (Cp*ReO)₂(μ-O)₂ allows catalytic deoxygenation of epoxides, however, conproportionation between the Re^V and Re^VII species to form clusters of {(Cp*Re)₃(μ-O)₆}²⁺(ReO ₄^-)₂ and new compound {(Cp*Re)₃(μ²-O)₃(μ ³-O)₃ReO₃}⁺(ReO₄ ^-) leads to removal of rhenium from the catalytic cycle and loss of activity.

Full text of DOI:10.1039/a708516h

View Article Online / Journal Homepage / Table of Contents for this issue
Catalytic deoxygenation of epoxides with (Cp*ReO)2(m-O)2 and catalyst
deactivation
Kevin P. Gable,*† Fedor A. Zhuravlev and Alexandre F. T. Yokochi
Department of Chemistry, Orgeon State University, Corvallis, OR 97331-4003, USA
In situ reduction of Cp*ReO3 by PPh3 to form (Cp*ReO)2(m-
O)2 allows catalytic deoxygenation of epoxides, however,
conproportionation between the ReV and ReVII species to
form clusters of {(Cp*Re)3(m-O)6}2+(ReO42)2 and new com-
not be true under the fairly severe reaction conditions of the
deoxygenation ( > 100 °C). Indeed, heating equimolar amounts
of the two at 110 °C for 14 h formed a precipitate (sealed tube;
isolated yield 5% from CHCl3). This green solid was isolated by
filtration and washing with benzene, and was identified as
{(Cp*Re)3(m-O)6}2+(ReO42)2 1, first characterized by Herr-
mann et al.5 The IR spectrum (908s cm21, 609, 650w cm21) and
1H NMR spectrum (d 2.22, s; CD3CN) were identical to
reported values. As expected, this salt was unreactive toward
epoxides at elevated temperatures. Anion exchange with NaI in
water gave the iodide salt; the perrhenate band disappeared
from the IR spectrum, though the weak m-O bands remained.
Oxidation of the iodide salt with aqueous H2CrO4 (Jones’
reagent) gave a 37% yield of Cp*ReO3 [based on formulating
the salt as {(Cp*Re)3(m-O)6}I2].
pound {(Cp*Re)3(m -O)3(m -O)3ReO3}+(ReO42) leads to re-
moval of rhenium from the catalytic cycle and loss of
activity.
2
3
A primary focus of recent work with metal oxo systems has
been formation of new C–O bonds.1 However, cleavage of C–O
bonds with concomitant formation of one or more new MNO
bonds is also a useful transformation. One example of this is
deoxygenation of epoxides, wherein removal of an oxygen atom
generates a new CNC p bond. Given that most methodology for
making epoxides begins with the alkene, the combination
epoxidation/deoxygenation would provide a useful protection/
deprotection sequence for the multiple bond. Few such
sequences now exist,2 and many of these risk loss of
stereochemical integrity or remain incompatible with other
functional groups in the organic substrate.
We recently observed that (Cp*ReO)2(m-O)2 reacted ster-
eospecifically with epoxides to form alkene plus Cp*ReO3,
presumably via the monomeric form Cp*ReO2.3 Cook and
Andrews recently reported catalytic deoxygenation of vicinal
diols with this rhenium system,4 so we decided to explore the
catalytic deoxygenation of epoxides.
Conditions similar to stoichiometric deoxygenation reactions
were used: toluene solvent, sealed under vacuum, heated to
90–120 °C. An initial study of deoxygenation of 1,2-epoxy-
dodecane using 5 mol% Re gave only traces of alkene after
extensive reaction times. As seen in Table 1, increasing the
porportion of rhenium led to improved yields, but very little
turnover. However, catalytic turnover could be achieved if the
electronic properties of the epoxide were properly tuned.‡ The
optimum turnover was seen in the case of electron-withdrawing
substituents (Table 1).
It was evident from these experiments that catalyst turnover
was being impeded by an interfering reaction. Further, in cases
where poor turnover was observed (Table 1, entries 1, 2, 7), a
green solid was seen to precipitate. The rapid reduction of
Cp*ReO3 to (Cp*ReO)2(m-O)2 at room temperature5a implied
the two compounds did not react with each other, but this might
In epoxide reductions and in conproportionation of Cp*ReO3
with (Cp*ReO)2(m-O)2, this compound was accompanied by
formation of an air-stable purple compound 2 that was insoluble
in benzene. This second compound became predominant at
relatively high rhenium concentrations (Table 1, entries 4, 5);
and it was isolated from the conproportionation reaction by
precipitation from benzene in 28% yield. Aside from the
dramatically different visible spectrum, the only other sig-
nificant spectroscopic difference from 1 was a new IR peak at
736 cm21 and weak shoulder on the very strong perrhenate Re–
O stretch at 908 cm21. A single sharp peak appeared in the 1H
NMR (d 2.23 in CD2Cl2). We obtained high quality crystals of
this compound for X-ray diffraction analysis;6§ the structure of
the cation is shown in Fig. 1.
2
3
This compound is formulated as {(Cp*Re)3(m -O)3(m -
O)3ReO3}+(ReO42), a monoperrhenate salt of a cluster. For-
mally, it can be represented by the coordination of ReO32 to 1,
although a more useful interpretation is to view it as the neutral
+
(Cp*Re)3(m-O)6 coordinated to ReO3 . (The latter is consistent
with the theoretical prediction that compound 1 should be a
ground-state triplet,7 and similar to other reported structures.8)
The ReO3 unit is a slightly distorted octahedral rhenium in a
typical LReO3 environment: the terminal oxo bond lengths are
normal [1.668(10), 1.685(13), 1.707(12) Å] as are the O–Re–O
angles [103.2(6)°, 105.9(6), 104.1(6)°]. The Re3(m-O)6 core is
similar to the trinuclear dication; the Re–Re distances are
2.750(1) and 2.759(1) Å. There is significant distortion of the
a
Table 1 Reaction of epoxides with catalytic Cp*ReO3 + excess PPh3
Conversion
Epoxide
c/m
[Cp*ReO3]/m
t/h
T/°C
(%)
Turnover
1
2
3
4
5
6
7
8
9
1,2-Epoxydecane
0.16
0.14
0.14
0.14
0.14
0.14
0.23
0.12
0.28
0.007
0.007
0.029
0.056
0.14
0.007†
0.009
0.007
0.005
20
18
13
13
13
13
6
116
112
112
112
112
112
116
90
5%
1.1
—
0.9
1.6
0.9
10
—
3.4
12.9
1,2-Epoxydodecane
1,2-Epoxydodecane
1,2-Epoxydodecane
1,2-Epoxydodecane
1,2-Epoxydodecane
2,3-Epoxynorbornane
p-Bromostyrene Oxide
3-Fluoropropylene Oxide
< 5%
20%
70%
> 90%
50%
< 5%
20%
23%
16
3
116
a [PPh3] = 0.17 m except entry 6, [PPh3] = 0.35 m.
Chem. Commun., 1998
799

View Article Online
+
Cp*
O
O
O
Re
Re
Cp*
Cp*
O
Re
O
O
O
PPh3
Re
Re
Re
Re
1
O
O
O
O
O
O
O
O
Re
O
+
O=PPh3
O
O
–
ReO4
O
R
2
R
Scheme 1
National Science Foundation (CHE-9619296, CHE-9015466)
for support of this work. A. F. T. Y. thanks the National Science
Foundation (grant CHE-9711187) for support.
Notes and References
† E-mail: gablek@chem.orst.edu
‡ Examination of the stoichiometric O-atom transfer from substituted
styrene oxides shows that reaction is accelerated by electron deficient
substituents: K. P. Gable and M. A. Gartman, unpublished work.
2
3
§ Crystal data for {Cp*Re)3(m -O)3(m -O)3ReO3}ReO4·NCMe (2·NCMe).
Data were collected on a block shaped (0.5 3 0.3 3 0.2 mm) dark brown
crystal of C32H48NO13Re5 (M = 1585.71) on a Siemens P4 equipped with
graphite monochromated Cu radiation (m
=
30.002 mm21) at room
temperature. Automated search and indexing routines revealed that the
crystals belong to the monoclinic space group P21/c (no. 14) with
a = 11.180(1), b = 15.631(1), c = 22.612(3) Å, b = 98.614(8)°,
U = 3906.8(7) Å3, Z = 4, Dc = 2.696 Mg m23. Of 10677 data collected,
5191 were unique (Rint = 9.10%), and of these 4997 had I > 2s(I). Data
were corrected for the effects of absorption anisotropy by analytical
methods (face indexing). The structure was solved using a Patterson map
search using SHELXS-90, and expanded by Fourier techniques and refined
(full-matrix least-squares refinement on F2) using SHELXL-93. Refine-
ment of 492 parameters using all data yielded final residuals of R1 = 0.661,
Fig. 1 Crystal structure of the cation for 2
two sets of bridging oxo ligands compared to compound 1, with
bond lengths averaging to 1.94 Å (m2) and 2.02 Å (m3). The
O–(ReO3) distances are long at an average of 2.18 Å.
wR2 = 0.1768. The largest residual electron density peaks (3.28 e2 Å23
)
were all very close to the Re atoms and are not of chemical significance.
CCDC 182/789.
Compound 2 is a precursor to the trinuclear cluster 1. Heating
a Me2SO solution of 2 in air to 100 °C for 2 h results in
quantitative conversion. (Under conditions of the conpropor-
tionation or epoxide deoxygenation, Cp*ReO3 can play the
same chemical role as Me2SO in converting the ReO3 unit to
perrhenate.)9 It is still not clear what the origin of the trinuclear
core is, nor the fate of the Cp* ligands lost in formation of 1 and
2. We have seen NMR evidence for monomeric Cp*ReO2,3 but
do not see evidence for a trimeric species. It is possible that
Cp*ReO2 can condense with Cp*ReO3, and that this dimeric
species initiates a cascade resulting in 2. Alternatively, if indeed
the trinuclear ReV cluster forms as an equilibrium aggregate of
Cp*ReO2 units, it may attack Cp*ReO3 irreversibly.
1 W. A. Nugent and J. M. Mayer, Metal–Ligand Multiple Bonds, Wiley-
Interscience, New York, 1988; R. A. Sheldon and J. K. Kochi, Metal
Catalyzed Oxidations of Organic Compounds, Academic Press, New
York, 1981; Organic Synthesis by Oxidation with Metal Compounds, ed.
W. J. Mijs, C. R. H. I. deJonghe, Plenum, New York, 1986; K. P. Gable,
Adv. Organomet. Chem., 1997, 41, 127.
2 Encyclopedia of Reagents for Organic Synthesis, ed. L. A. Paquette,
Wiley, New York, 1995 vol. 3, pp. 1649, 2078; K. G. Molloy, Inorg.
Chem., 1988, 27, 677; P. K. Chowdhury, J. Chem. Res., 1990, (S) 192;
J. March, Advanced Organic Chemistry: Reactions, Mechanisms and
Structure, Wiley, New York, 4th edn., 1992, pp. 1029–1030.
3 K. P. Gable, J. J. J. Juliette and M. A. Gartman, Organometallics, 1995,
14, 3138.
Clearly, the success of the catalytic cycle for epoxide
deoxygenation (or any other system involving this chemistry)
depends on a careful balance of reaction rates (Scheme 1).
Cp*ReO2 must react with the epoxide substrate rapidly, or else
conproportionation will lead to inactivation. The rate of O-atom
transfer from epoxide is controlled by the reactivity of the
substrate, as seen by the impact of substituent on turnover
number. Likewise, Cp*ReO3 must be rapidly reduced to
Cp*ReO2; there is again a competition between reduction and
cluster formation. We tested this by increasing the concen-
tration of PPh3; a 2.4-fold excess led to an increase in
conversion from < 5% to 50% (Table 1, Entry 6)! It must be
noted that this required an almost saturated solution of PPh3.
Although this modification interferes with the practicality of
this catalytic cycle, it shows that the key to an improved system
is to design a kinetically more reactive stoichiometric reductant
that does not interact with the epoxide.
4 G. K. Cook and M. A. Andrews, J. Am. Chem. Soc., 1996, 118, 9448.
5 (a) W. A. Herrmann, R. Serrano, M. L. Ziegler, H. Pfisterer and B. Nuber,
Angew. Chem., Int. Ed. Engl., 1985, 24, 50; (b) W. A. Herrmann,
R. Serrano, U. Ku¨sthardt, E. Guggolz, B. Nuber and M. L. Ziegler,
J. Organomet. Chem., 1985, 287, 329.
6 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467; G. M.
Sheldrick, in Crystallographic Computing 6, ed. H. D. Flack, L. Parkanyi
and K. Simon, Oxford University Press, Oxford, 1993.
7 P. Hoffmann, N. Ro¨sch and H. R. Schmidt, Inorg. Chem., 1986, 25, 4470;
P. Hoffmann and N. Ro¨sch, J. Chem. Soc., Chem. Commun., 1986,
843.
8 K. Wieghardt, C. Pomp, V. Nuber and J. Weiss, Inorg. Chem., 1986, 25,
1659; C. Pomp and K. Wieghardt, Polyhedron, 1988, 7, 2537;
W. A. Hermann, P. W. Roesky, F. E. Ku¨hn, M. Elison, G. Artus,
W. Scherer, C. C. Roma˜o, A. Lopes and J.-M. Basset, Inorg. Chem.,
1995, 34, 4701; J. Xiao, R. J. Puddephatt, L. Manojlovic´-Muir,
K. W. Muir and A. A. Torabi, J. Am. Chem. Soc., 1994, 116, 1129.
9 K. P. Gable, J. J. J. Juliette, C. Li and S. P. Nolan, Organometallics, 1996,
15, 5250.
We wish to thank the donors to the Petroleum Research Fund,
administered by the American Chemical Society, and the
Received in Bloomington, IN, USA, 25th November 1997; 7/08516H
800
Chem. Commun., 1998