Facile and rapid immobilization of copper(II) bis(oxazoline) catalysts on silica: Application to Diels-Alder reactions, recycling, and unexpected effects on enantioselectivity

Article abstract of DOI:10.1016/j.tetlet.2004.03.001

Two chiral copper(II) bis(oxazoline) complexes have been immobilized on silica via electrostatic interactions using a remarkably straightforward procedure. The immobilized catalysts were tested in a standard Diels-Alder reaction and gave surprising results. Where the immobilized Cu((S,S)-phenyl-box) (OTf)₂ catalyst was used, the predominant enantiomer formed was the opposite of that produced using the same catalyst in a homogeneous reaction. This is a startling result given that the only difference is the electrostatic immobilization of the catalyst on amorphous silica. The activity of the catalyst in a hetero Diels-Alder reaction was also tested. This catalyst was also recycled, successfully maintaining a similar activity to the homogeneous analogue through a number of cycles.

Full text of DOI:10.1016/j.tetlet.2004.03.001

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 3177–3180
Facile and rapid immobilization of copper(II) bis(oxazoline)
catalysts on silica: application to Diels–Alder reactions,
recycling, and unexpected effects on enantioselectivity
a
Patrick O’Leary, Nico P. Krosveld, Krijn P. De Jong, Gerard van Koten
a
b
a
a,
*
and Robertus J. M. Klein Gebbink
a
Utrecht University, Debye Institute, Department of Metal Mediated Synthesis, Padualaan 8, 3584 CH Utrecht, The Netherlands
b
Utrecht University, Department of Inorganic Chemistry and Catalysis, Sorbonnelaan 16, 3584 CA Utrecht, The Netherlands
Received 6 January 2004; revised 19 February 2004; accepted 1 March 2004
Abstract—Two chiral copper(II) bis(oxazoline) complexes have been immobilized on silica via electrostatic interactions using a
remarkably straightforward procedure. The immobilized catalysts were tested in a standard Diels–Alder reaction and gave sur-
prising results. Where the immobilized Cu((S,S)-phenyl-box)(OTf) catalyst was used, the predominant enantiomer formed was the
2
opposite of that produced using the same catalyst in a homogeneous reaction. This is a startling result given that the only difference
is the electrostatic immobilization of the catalyst on amorphous silica. The activity of the catalyst in a hetero Diels–Alder reaction
was also tested. This catalyst was also recycled, successfully maintaining a similar activity to the homogeneous analogue through a
number of cycles.
ꢀ
2004 Elsevier Ltd. All rights reserved.
1
. Introduction
chiral rhodium–phosphine hydrogenation catalyst was
reported via hydrogen bonding of the triflate counter
anion with the MCM-41 silica surface. The electro-
statically immobilized catalyst had similar activity to the
catalyst in homogeneous solution with good recyclabil-
ity and low levels of leaching.
4
Immobilization of catalysts represents an important
area of research. The ability to recycle expensive chiral
species is attractive at the research and application level
and, as a result, research into immobilization of catalytic
1
species on supports is ongoing.
Immobilization of copper(II) bis(oxazoline) catalysts
has been achieved in both a covalent and noncovalent
Chiral copper(II) bis(oxazoline) catalysts have been
applied to the homogeneous catalysis of a wide variety
of reactions. In many cases, products of high enantio-
5;6
manner. Typically the methods used have involved
complex procedures and in some cases significant
alterations to the ligand. As triflate counterions are
often used in combination with copper(II) bis(oxazoline)
cations we envisaged that a process similar to that
employed by De Rege with a rhodium hydrogenation
2
purity have been obtained. Bis(oxazoline) ligands,
however, depending on the amino acid used, can be
relatively expensive to prepare and the catalysts are
3
often used in amounts up to 10 mol % or indeed greater.
4a
Immobilization of these catalysts thus presents a
worthwhile goal, which would expand the ability to
reuse them. Recently, the immobilization of a cationic
catalyst, and subsequently by Thomas with both rho-
4b
dium and palladium hydrogenation catalysts, could be
used to immobilize these catalysts. The principle
advantages of this immobilization technique are: the
absence of any complexity in the immobilization pro-
cess, the widespread availability of the silica inorganic
support (chromatography grade silica), the lack of any
need to alter the catalyst or preceding ligand in any way
to facilitate the immobilization and the rapidity with
which the immobilization can be completed.
Keywords: Copper(II) bis(oxazoline); Electrostatic immobilization;
Diels–Alder; Silica; Enantioselectivity reversal.
*
Corresponding author. Tel.: +31-30-2533054; fax: +31-30-2523615;
e-mail: r.j.m.kleingebbink@chem.uu.nl
0
040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.03.001

3178
P. O’Leary et al. / Tetrahedron Letters 45 (2004) 3177–3180
Figure 1. CopperðIIÞ bis(oxazolines) used in this study.
Scheme 1. Sketch of the immobilization process employed in this study (attachment via H bonds can be envisaged in a number of ways).
2
. Results and discussion
The phenyl substituted copper(II) bis(oxazoline) catalyst
(Fig. 1) was first generated in dichloromethane from
the commercially available ligand and anhydrous copper-
II) triflate. Immobilization of this catalyst was achieved
1
(
by addition of this solution to silica (ACROS silica gel
for chromatography, pore diameter 6 nm), which had
been dried at 70 ꢁC under vacuum for 2 h (Scheme 1).
Stirring of the homogeneous solution with the silica for
a period of 2 min resulted in complete immobilization of
the complex. The solution, which was initially a light
green, was decolorized and the silica took on the same
color. The solvent was decanted and the solid washed
with two aliquots of dichloromethane and then stored
under fresh dichloromethane until use. Typically, theo-
retical loadings of 1.3% copper were used in this study.
Scheme 2. Diels–Alder reaction between ethyl (E)-2-oxopent-3-enoate
and ethyl vinyl ether catalyzed by electrostatically immobilized Cu(II)
catalyst.
went to completion within 30 min and the product 5 was
obtained in 18% ee. To compare the performance of the
electrostatically immobilized catalyst with that of the
catalyst in homogeneous solution, a temperature profile
of the enantiopurity of the product formed in the reac-
tion was obtained. These results are shown graphically
in Figure 2. Interestingly there is little difference between
the performance of the catalyst in solution and the same
catalyst electrostatically immobilized on silica. The
enantiomeric excess of the product formed in the reac-
tion was slightly reduced in the case of the supported
catalyst compared to the solution reaction but the most
striking element of this study was the retention of
activity on reuse, indeed in some cases a higher ee was
observed during the second cycle. The supported cata-
lyst was found to retain its activity for four cycles.
Whereas most of the reactions conducted for this study
used 10 mol % of the catalyst, when 5 mol % was used, a
similar performance was observed both in terms of ee
and recyclability.
The electrostatically immobilized catalyst was initially
tested in the inverse electron demand hetero Diels–Alder
reaction between ethyl (E)-2-oxopent-3-enoate 3 and
ethyl vinyl ether 4 as described by Evans using 10 mol %
2d
of catalyst (Scheme 2). The major product of the
reaction was the endo product 5 and the major enan-
tiomer was the (2S,4S) product. At 0 ꢁC, the reaction
Representative experimental procedure: To 10 mol % of the catalyst 1
(
(
homogeneous or heterogeneous as appropriate) in dichloromethane
ꢀ
5 mL) was added 30 mg of dry 4 A molecular sieves and ethyl (E)-2-
ꢀ4
oxopent-3-enoate 3 (27 mg, 1.90 · 10 mol) in dichloromethane
2 mL). This mixture was then allowed to equilibrate at the reac-
tion temperature at which point ethyl vinyl ether 4 (0.075 mL,
(
ꢀ4
7
.84 · 10 mol) was added via syringe. The reaction mixture was
Next, the tert-butyl copper(II) bis(oxazoline) 2 was
immobilized by the same method and the resulting
heterogenized catalyst was tested in the same Diels–
Alder reaction (see Table 1). This catalyst has been
stirred until completion at which point it was filtered through a short
silica column (where a solid catalyst was used the reaction mixture
was first decanted from the solid). The volatiles were removed under
1
vacuum. Reaction completion was confirmed by H NMR spectro-
2d
described as effective for this reaction and indeed the
homogeneous reaction at room temperature gave a
product with a 56% ee [the major enantiomer has the
configuration (2R,4R)]. The corresponding electrostati-
cally immobilized catalyst also gave a rapid reaction,
scopy and the enantiomeric excess of the product measured by HPLC
on a Chiralpak OD column. Detection was at 254 nm. Absolute
configuration was assigned by comparison with the work of Evans
2d
et al. Recovered solid catalysts were washed with two aliquots of
dichloromethane before reuse.

P. O’Leary et al. / Tetrahedron Letters 45 (2004) 3177–3180
3179
Figure 2. Results for the hetero Diels–Alder reaction between ethyl (E)-2-oxopent-3-enoate and ethyl vinyl ether catalyzed by 1 (homogeneous and
heterogeneous). The ee is that of the major endo product.
Table 1. Results for the hetero Diels–Alder reaction between ethyl (E)-
2-oxopent-3-enoate and ethyl vinyl ether catalyzed by 2 (homogeneous
and heterogeneous)
a
No. Catalyst
type
Cycle
T (ꢁC)
Time for
completion
Ee (%)
1
2
3
4
5
Homo
Homo
Hetero
Hetero
Hetero
20
ꢀ78
20
40 min
56
95
24
17
b
1
2
1
20 min
12 h
20
Scheme 3. Diels–Alder reaction between acryloyloxazolidinone 6 and
cyclopentadiene 7 catalyzed by 1 and 2 (homogeneous and hetero-
geneous).
c
ꢀ78->ꢀ5 16 h
41
a
Determined by HPLC on a Chiralpak OD column and detection at
54 nm. The major enantiomer has the configuration (2R,4R)
2
2
d
(
assigned by analogy to the work of Evans).
2d
The results obtained in this study are summarized in
Table 2.
b
c
As reported by Evans et al.
Only 25% conversion.
We were surprised to note a very unusual phenomenon
in this study. When catalyst 1 was used in homogeneous
solution an ee of 20% was observed. The same catalyst,
immobilized electrostatically to a silica support, gave an
ee of 33%. Most surprising is that the major enantiomer
of the product in the homogeneous reaction is S and in
the heterogeneous reaction the major enantiomer is the
however, the ee of the resulting product was only 24%.
The catalyst was found to change color during the
reaction resulting in a brown solid being recovered after
the reaction. The recycled catalyst was found to be much
less active and the enantiomeric excess of the product in
a second cycle was poor. The rate of reaction also fell in
the second cycle. A reaction conducted overnight, using
freshly prepared immobilized catalyst, with an initial
temperature of ꢀ78 ꢁC rising to ꢀ5 ꢁC was found to be
only 25% complete and gave product with an ee of 42%
8
R (assigned by analogy to the work of Evans et al. ).
This result means that using the same enantiomer of the
ligand we can produce different enantiomers of the
product on the basis of the interaction of the catalyst
à
with an achiral support.
(
Table 1, entry 5). It is quite evident from these results
A reversal of this type has been previously observed,
when a magnesium bis(oxazoline) catalyst was treated
with water. In the anhydrous case, a tetrahedral tran-
that the behavior of the immobilized tert-butyl catalyst
is very different to that of its phenyl analogue. The
phenyl-substituted catalyst maintained its activity,
enantioselectivity and was highly recyclable, whereas the
tert-butyl catalyst was obviously less stable and a less
effective catalyst when immobilized by this technique.
9
sition state is envisaged and in the hydrated case, the
transition state is thought to be octahedral. This change
in co-ordination chemistry gives rise to the reversal of
enantioselectivity. A related explanation could be
envisaged in the general case of copper catalysts, but
hydrated species have not been reported to give a
reversal in enantioselectivity for copper bis(oxazolines).
Thus it appears that the nature of the interaction with
Next, we turned our attention to the Diels–Alder reac-
tion between N-acryloyloxazolidinone 6 and cyclopenta-
diene 7 (Scheme 3). This reaction is a well-known test
reaction for enantioselective catalysts of the Diels–Alder
7
reaction. The reaction was conducted at ꢀ30 ꢁC in
dichloromethane using an excess of the cyclopentadiene.
The endo diasteromer 8 is known to be the major
product of this reaction and we determined the enantio-
selectivity of the catalyst in the synthesis of this product.
à
Using the homogeneous catalyst the endo: exo ratio was ꢁ90:10
however the corresponding heterogeneously catalyzed reaction gave
the product with ꢁ80:20.

3180
P. O’Leary et al. / Tetrahedron Letters 45 (2004) 3177–3180
Table 2. Results for the Diels–Alder reaction between acryloyloxazolidinone 6 and cyclopentadiene 7 catalyzed by 1 and 2 (homogeneous and
heterogeneous)
a
Temperature (ꢁC)
Catalyst, 10 mol %
R
Conversion (%)
Ee (%)
Major enantiomer
ꢀ
ꢀ
ꢀ
ꢀ
30
30
50
50
Homogeneous
Heterogeneous
Homogeneous
Heterogeneous
Ph
Ph
100
100
50
20
33
59
57
S
R
S
S
t
Bu
Bu
t
43
a
Of the major endo diastereomer. Determined by HPLC on a Chiralpak OD column and detection at 215 nm. The absolute configuration was
8
assigned by analogy to the work of Evans et al.
the silica in our case results in a different reactivity of the
copper species compared to its hydrated analogue.
References and notes
1
2
. Song, C. E.; Lee, S.-G. Chem. Rev. 2002, 102, 3495–3524.
. (a) Fache, F.; Schulz, E.; Tommasino, M. L.; Lemaire, M.
Chem. Rev. 2000, 100, 2159–2231; (b) Evans, D. A.; Faul,
M. M.; Bilodeau, M. T.; Anderson, B. A.; Barnes, D. M. J.
Am. Chem. Soc. 1993, 115, 5328–5329; (c) Evans, D. A.;
Johnson, J. S. Acc. Chem. Res. 2000, 33, 325–335; (d)
Evans, D. A.; Johnson, J. S.; Olhava, E. J. J. Am. Chem.
Soc. 2000, 122, 1635–1649; (e) Evans, D. A.; Burgey, C. S.;
Paras, N. A.; Vojkovsky, T.; Tregay, S. W. J. Am. Chem.
Soc. 1998, 120, 5824–5825.
3. Conclusion
The electrostatically immobilized bis(oxazoline) cata-
lysts prepared in this study are the first to be immobi-
lized on silica by this method. A hetero Diels–Alder
reaction has been catalyzed and gave products in up to
3
. Denmark, S. E.; Nakajima, N.; Nicaise, J.-C.; Fauchner,
A.-M.; Edwards, J. P. J. Org. Chem. 1995, 60, 4884–4892.
. (a) De Rege, F. M.; Morita, D. K.; Ott, K. C.; Tumas, W.;
Broene, R. D. Chem. Commun. 2000, 1797–1798; (b) Raja,
R.; Thomas, J. M.; Jones, M. D.; Johnson, B. F. G.;
Vaughan, D. E. W. J. Am. Chem. Soc. 2003, 125, 14982–
14983.
4
2% ee. This is the highest ee of which we are aware in a
Diels–Alder cycloaddition obtained by reaction of a
noncovalently immobilized bis(oxazoline) catalyst. It is
apparent from our results that careful selection of the
ligands used, is important. The phenyl substituted
bis(oxazoline) 1 maintained its activity on immobiliza-
tion and through a number of recycles, whereas this was
not the case for catalyst 2.
4
5. (a) Burguete, M. I.; Fraile, J. M.; Garc ꢁı a, J. I.; Garc ꢁı a-
ıas, C. I.; Luis, S. V.; Mayoral, J. A.
Verdugo, E.; Herrer
ꢁ
J. Org. Chem. 2001, 66, 8893–8901; (b) Rechavi, D.;
Lemaire, M. J. Mol. Catal. A: Chem. 2002, 182–183, 239–
In a standard homo Diels–Alder reaction, interaction of
the catalyst with an achiral inorganic support altered its
reactivity sufficiently to produce the previously less
favored enantiomer in excess. This obviously has very
significant implications in a wide area of chemistry as
traditionally one enantiomer is much more expensive to
produce than the other. This development may allow
use of the easily available chiral pool to produce not
only straightforward products, but also the less avail-
able opposite enantiomers. Such a simple method, which
offers such flexibility in the enantiomer produced is of
considerable interest and we will be studying this effect
in detail.
2
2
47; (c) Hallman, K.; Moberg, C. Tetrahedron: Asymmetry
001, 12, 1475–1478; (d) Chow, H.-F.; Mak, C. C. J. Org.
Chem. 1997, 62, 5116–5127; (e) Clarke, R. J.; Shannon, I. J.
Chem. Commun. 2001, 1936–1937; (f) Corma, A.; Garcia,
H.; Moussaif, A.; Sabater, M. J.; Zniber, R.; Redouane, A.
Chem. Commun. 2002, 1058–1059.
6. (a) Fraile, J. M.; Garc ꢁı a, J. I.; Harmer, M.; Herrer ꢁı as, C. I.;
Mayoral, J. A. J. Mol. Catal. A: Chem. 2001, 165, 211–218;
(b) Fern
C. I.; Mayoral, J. A.; Salvatella, L. Catal. Commun. 2001,
, 165–170; (c) Fraile, J. M.; Garc ꢁı a, J. I.; Mayoral, J. A.;
ꢁa ndez, A. I.; Fraile, J. M.; Garc ꢁı a, J. I.; Herrer ꢁı as,
2
Tarnai, T. Tetrahedron: Asymmetry 1998, 9, 3997–4008; (d)
Fraile, J. M.; Garc ꢁı a, J. I.; Mayoral, J. A.; Tarnai, T.
Tetrahedron: Asymmetry 1997, 8, 2089–2092.
7
8
. (a) Owens, T. D.; Souers, A. J.; Ellman, J. A. J. Org. Chem.
2
003, 68, 3–10; (b) Comprehensive Asymmetric Catalysis;
Jacobsen, E. J., Pfaltz, A., Yamamoto, H., Eds.; Springer:
New York, 1999.
Acknowledgements
. Evans, D. A.; Miller, S. J.; Lectka, T.; von Matt, P. J. Am.
Chem. Soc. 1999, 121, 7559–7573.
9. Zanoni, G.; Castronova, F.; Franzini, M.; Vidari, G.;
This work was supported by the Netherlands’ Research
School Combination-Catalysis (NRSC-C).
Giannini, E. Chem. Soc. Rev. 2003, 32, 115–129.