Electrostatically immobilised BOX and PYBOX metal catalysts: application to ene reactions

Article abstract of DOI:10.1016/j.tetlet.2008.12.046

Copper(I), copper(II) and scandium(III) triflate complexes of BOX and PYBOX ligands have been electrostatically immobilised on silica. The performance of the immobilised catalysts, in the two carbonyl-ene reactions studied, compares very well with that of

Full text of DOI:10.1016/j.tetlet.2008.12.046

Tetrahedron Letters 50 (2009) 979–982
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Electrostatically immobilised BOX and PYBOX metal catalysts: application
to ene reactions
Chiara McDonagh, Patrick O’Leary *
National University of Ireland, Galway, School of Chemistry, SMA C˙ T (Synthetic Methods: Asymmetric Catalytic Transformations), Ireland
a r t i c l e i n f o
a b s t r a c t
Article history:
Copper(I), copper(II) and scandium(III) triflate complexes of BOX and PYBOX ligands have been electro-
statically immobilised on silica. The performance of the immobilised catalysts, in the two carbonyl-ene
reactions studied, compares very well with that of their homogeneous equivalents. The immobilised cat-
alysts were successfully reused a number of times. In the case of a scandium complex, a variation in
enantioselectivity of up to 73% was observed on immobilisation compared to its use homogeneously.
The reason for this variation in enantioselectivity is explored with the aid of molecular modelling.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 12 November 2008
Revised 3 December 2008
Accepted 10 December 2008
Available online 16 December 2008
Keywords:
Copper
Scandium(III)
Bis(oxazolinyl)pyridine
Bis(oxazoline)
Electrostatic immobilisation
Ene reaction
Silica
Molecular modelling
Enantioselective catalysis in both research and industry makes
extensive use of chiral catalysts derived from chiral ligands and
metal salts. Many such catalyst systems offer advantages in terms
of activity and selectivity. However, particularly in reactions where
large catalyst loadings are required, ligands which are synthesised
by multistep sequences from amino acids or expensive metal salts,
can make the use of such catalysts impractical on a large scale.
Immobilisation and reuse strategies are particularly advantageous
in these cases. Immobilisation of catalysts on inorganic materials
such as silica or similar materials via covalent bonds or by other
methods is an active area of research.1
the homogeneous catalysis of ene type reactions for the synthesis
of homopropargylic alcohols. We decided to apply the catalysts to
3
the synthesis of homoallylic alcohols via carbonyl-ene reactions.
The carbonyl-ene reaction is a synthetically powerful carbon–car-
bon bond forming reaction, which forms the useful homoallylic
alcohol functional group and can create two stereocentres.4 Its
use has been described in the enantioselective synthesis of many
biologically active compounds as well as of many other chiral com-
5
pounds. A diverse range of catalysts has been described for this
6
reaction. Bisoxazoline- and bis(oxazolinyl)pyridine-based cata-
lysts are used typically in amounts of ꢀ10 mol % and so the ability
to reuse the catalyst would present a significant advantage. The
complexes of BOX Ligand 1 and PYBOX ligands 2 and 3 (Fig. 1) were
used in this study.
We have recently reported an extremely facile method for the
immobilisation of copper(I) and copper(II) complexes of bisoxazo-
2
line (BOX) and bis(oxazolinyl)pyridine (PYBOX) ligands on silica.
These immobilised complexes have been shown to be catalytically
active in the Diels–Alder cycloaddition2a and the direct addition of
terminal alkynes to imines leading to propargylamines.2 In previ-
ous studies, we have also reported on the recyclability of these
The phenyl-substituted copper(II)BOX1 catalyst was generated
in dichloromethane from the ligand and copper(II) triflate, and this
solution was added to silica (silica gel 40–63 lm, chromatographic
2
catalysts.
To expand the range of reactions to which our immobilised cat-
alysts are applied, we studied the carbonyl-ene reaction. In addi-
tion, we expanded the scope of the technique by applying it to
scandium(III) complexes which, Evans has reported, are active in
*
Corresponding author. Tel.: +353 91 492476; fax: +353 91 525700.
E-mail address: patrick.oleary@nuigalway.ie (P. O’Leary).
Figure 1.
0
040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.12.046

9
80
C. McDonagh, P. O’Leary / Tetrahedron Letters 50 (2009) 979–982
grade), which had been dried under vacuum for 2 h at 70 °C. The
heterogeneous mixture was stirred until the blue colour disap-
peared from the solvent and the silica became coloured (ꢀ1 min).
The organic layer was decanted and the solid was washed with
two aliquots of the solvent to be used in the catalytic reaction. Typ-
ically, theoretical loadings of 2.8% copper were used in this study.
The electrostatically immobilised Cu(II)BOX1 triflate was tested in
Table 2
Immobilised scandium(III)PYBOX catalysis of the carbonyl-ene reaction of
ylstyrene 4 and ethyl glyoxylate 5
a
-meth-
Catalyst type
Ligand
Cycle
Conva (%)
Ee (%) , R/S
b
Homogeneous
Heterogeneous
Heterogeneous
Heterogeneous
Homogeneous
Heterogeneous
Heterogeneous
Heterogeneous
2
2
2
2
3
3
3
3
N/A
1st
2nd
3rd
N/A
1st
79
87
87
94
100
94
15, R
5, R
5, R
18, R
21, S
52, S
45, S
33, S
the reaction of a-methylstyrene 4 and ethyl glyoxylate 5 by simply
mixing the alkene, ethyl glyoxylate and the catalyst suspension un-
der nitrogen. All reactions were carried out at reflux to allow for
comparison. After the reaction was complete, the crude reaction
mixture was removed from the catalyst by decanting off the liquid
and washing the solid with the reaction solvent (2 Â 5 ml). The so-
lid catalyst was now ready to be reused and the product was iso-
lated from the solution by removing the solvent under reduced
pressure. Cu(I)BOX1 triflate was also prepared immobilised and
tested in a similar manner. The results are shown in Table 1.
The immobilised copper(II) catalyst was as active as its homo-
geneous counterpart in the first cycle, however, over the course
of three uses the conversion dropped from 62% to 42%. In addition,
the enantioselectivity obtained was less in the case of the hetero-
geneous catalyst, 75% ee versus 86% ee for the homogeneous cata-
lyst. The enantioselectivity also fell on reuse. The decrease in
conversion and selectivity is likely to be due to a small level of
2nd
3rd
87
72
a
With respect to the alkene determined by 1H NMR analysis.
Determined by GC (CYCLODEX B).
b
catalyst gave 100% conversion but only 21% ee. When the catalyst
was immobilised electrostatically on silica the activity of the cata-
lysts was largely maintained but there was a large increase in the
enantioselectivity. The ee was raised from 21% to 52% when the
catalyst was immobilised on chromatographic grade silica. More-
over, the ee remained higher through three cycles than that ob-
tained when the catalyst was used homogeneously. There was
some fall off in the catalyst activity across the three cycles which
is likely to be due to a small degree of leaching.
This result was potentially very useful as it represented a way of
enhancing the enantioselectivity achieved by a catalyst just by
adding silica. We tested the generality of the effect by carrying
out another ene reaction using our immobilised catalysts. We re-
acted methylene cyclohexane 7 with ethyl glyoxylate 5 using a
variety of homogeneous and heterogeneous metal complexes as
catalysts. The results are shown in Table 3.
2
b
leaching as well as possible surface contamination of the catalyst.
The copper(I) catalyst was significantly less active on immobili-
sation, 49% conversion versus 95% conversion. The enantioselectiv-
ity achieved by the immobilised copper(I) complex is comparable
to that obtained with the homogeneous catalyst 83% ee versus
8
7% ee, and was largely maintained over three cycles.
These initial results encouraged us to expand the scope of the
The reactions conducted using copper(I) and copper(II) BOX1
catalysts gave results largely consistent with those seen with the
previous ene reaction. The ees obtained with the homo- and heter-
ogeneous catalysts were broadly similar. The % conversions ob-
tained using the copper(II) catalyst were maintained on
immobilisation; however, as with the previous reaction, the cop-
per(I) catalyst activity was severely reduced on immobilisation.
When scandium(III) was used with the phenyl-substituted PY-
BOX2, the results were very disappointing. This again is consistent
with the previous reaction studied.
study and we decided to investigate the immobilisation of scan-
dium(III)PYBOX catalysts and their activity on immobilisation.
The catalysts derived from PYBOX ligands 2 and 3 and scan-
dium(III) triflate were prepared, immobilised and reacted in a sim-
ilar manner to that described for the copper BOX catalysts. The
results are shown in Table 2.
The homogeneous scandium(III)PYBOX2 complex was an active
catalyst for the ene reaction, however, the ee achieved was very
poor. When the catalyst was immobilised on silica the activity
was maintained but the enantioselectivity remained very poor.
Given our previous experience where phenyl and alkyl BOX cat-
alysts gave very different performances in Diels–Alder reactions on
The scandium(III) iso-propyl-substituted PYBOX3 again gave
very interesting results in the catalysis of this reaction. Used homo-
geneously, it was very active but not a very enantioselective cata-
lyst, whereas on immobilisation the catalyst activity fell slightly
but the product was isolated in 60% ee. Furthermore, the major
enantiomer of the product following the homogeneous reaction
was R and following the heterogeneous reaction it was S. The
enantioselectivity of the reaction swings by 73% on addition of
an achiral silica substrate to immobilise the homogeneous catalyst.
We had previously seen how immobilisation of a copper(II)BOX
catalyst by this method affected the enantioselectivity achieved in
Diels–Alder reactions, which is likely to be due to a change in the
2
a
immobilisation we decided to test the immobilised alkyl PYBOX3
scandium(III) complex in the ene reaction. Homogeneously, the
Table 1
Immobilised copper BOX1 catalysis of the canbonyl-ene reaction
7
geometry about the metal on immobilisation. Jørgensen’s work
in the area of the effect of additional coordinating ligands or re-
agents on the geometry of metal BOX systems could be adapted
to explain those results. In the case of the scandium PYBOX results
reported here, it is more difficult to explain the result.
Catalyst typea
Copper salt
Cycle
Convb (%)
eec (%)
Homogeneous
Heterogeneous
Heterogeneous
Heterogeneous
Homogeneous
Heterogeneous
Heterogeneous
Heterogeneous
Cu(OTf)
Cu(OTf)
Cu(OTf)
Cu(OTf)
CuOTf
2
2
2
2
N/A
1st
2nd
3rd
N/A
1st
61
62
55
42
95
49
56
39
86
75
78
66
87
83
83
74
Evans rationalised the results obtained in the annulation of alle-
nylsilanes on the basis of a model of the scandium(III) PYBOX inter-
mediate which had an octahedral scandium with the PYBOX
3
a
nitrogens bound in three adjacent equatorial sites. One apical site
was occupied by the one remaining bound triflate and the remain-
ing two sites by the ethyl glyoxylate. He argued that the relative
strength of the trans influence of the pyridine of the PYBOX and
the triflate determined whether ethyl glyoxylate bound with the
aldehyde apically or equatorially.
CuOTf
CuOTf
CuOTf
2nd
3rd
a
Theoretical loading of copper on silica is 2.8% for all heterogeneous catalysis.
With respect to the alkene determined by H NMR analysis.
Determined by GC (CYCLODEX B); S-Enantiomer in each case.
b
c
1

C. McDonagh, P. O’Leary / Tetrahedron Letters 50 (2009) 979–982
981
Table 3
Catalysis of the carbonyl-ene reaction of methylene cyclohexane 7 and ethyl glyoxylate 5
Homo %Conv , (%ee)b
a
Hetero run 1%Conv , (%ee)b
a
Hetero run 2 %Conv , (%ee)b
a
Metal Source
Ligand
Cu(OTf)
CuOTf
2
BOX1
BOX1
PYBOX2
PYBOX3
98 (82)
96 (79)
30 (79)
57 (0)
69 (72)
23 (77)
69 (5)
100 (79)
100 (10)
100 (13, R)
Sc(OTf)
Sc(OTf)
3
3
87 (60, S)
22 (52, S)
a
With respect to the alkene determined by 1H NMR analysis.
Determined by GC (CYCLODEX B).
b
One possible explanation for the observed swing in the enanti-
the surface of the silica. We positioned the gem silanol within
hydrogen bonding distance to the apical triflate sulfonate oxygen.
Interestingly, when the aldehyde is apical the model will not
converge and quickly fails, however, when the aldehyde is equato-
rial the calculation converged. The optimised structure is shown in
Figure 3. The gem silanol is positioned directly above the pyridine
ring and is within hydrogen bonding distance of the two sulfonate
oxygens not coordinated to the scandium.
oselectivity achieved when the immobilised catalyst was used may
be that coordination of the ethyl glyoxylate to the scandium centre
is affected by the interaction of the triflate with the silica solid sup-
port. If this interaction altered the relative trans influence of the
pyridine and triflate ligand it may alter the binding mode of the
aldehyde. We used molecular modelling using the GAUSSIAN-03 pro-
8
gram to investigate this possible explanation for our observed re-
sults. The structures discussed were optimised using B3LYP density
functional calculations (6-31G basis set).
It appears that, as proposed above, the coordination of the gem
silanol to the triflate alters the preferred geometry of coordination
of the ethyl glyoxylate from aldehyde apical (no silanol present) to
aldehyde equatorial (silanol present). This would expose a differ-
ent face for reaction and explain the reversal seen in the ene reac-
tion of methylene cyclohexane. It also provides a rationale for why
the enantioselectivity of the ene reaction of a-methylstyrene alters
as it does when an immobilised catalyst is used. The fact that the
homogeneously catalysed scandium(III)PYBOX3 ene reactions of
methylene cyclohexane and a-methylstyrene with ethyl glyoxy-
*
We first optimised the geometries of the octahedral scan-
dium(III)PYBOX3 triflate-ethyl glyoxylate complex with the alde-
hyde apical A and equatorial B (Fig. 2). A and B represent likely
intermediates in the homogeneously catalysed ene reaction. In A,
the re face of the aldehyde is exposed which would lead to reaction
giving the R ene product. In B, the si face of the aldehyde is exposed
which would lead to reaction giving the S ene product. The total
À1
energy of complex A was marginally lower (4.3 kJ mol ) indicat-
ing that reaction through this complex might be expected to pre-
dominate. Interestingly, the optimised structure in both cases
had a seven coordinate scandium, two sites being occupied by sul-
fonate oxygens.
late give opposite, albeit poor, selectivities (R vs S) is as yet
unexplained.
In conclusion, we have managed to successfully apply immobi-
lised copper- and scandium-based BOX and PYBOX catalysts to the
carbonyl-ene reaction. In the cases of copper-based catalysts, the
performance on immobilisation largely reflected that in a homoge-
neous context. The scandium-based catalyst gave variable results
on immobilisation. One catalyst, however, on immobilisation gave
enantioselectivities that were dramatically different to its homoge-
neous counterpart. This result is thought to be due to a change in
We next modelled the situation when silica was present. Given
the limited pre-drying, we applied to our silica the dominant sila-
9
nol species that is likely to be the gem silanol. A gem silanol might
be expected to interact with the available sulfonate oxygens
through hydrogen bonding. We modelled how complex A or B
might interact with a bulky gem silanol, which we use to represent
Figure 2. Optimised geometries of possible octahedral scandium(III)PYBOX3 triflate-ethyl glyoxylate complexes.

9
82
C. McDonagh, P. O’Leary / Tetrahedron Letters 50 (2009) 979–982
tional facilities and support. The training and support of Prof. John
Simmie in the use of ICHEC and GAUSSIAN-03 is gratefully
acknowledged.
References and notes
1.
(a) Song, C. E.; Lee, S.-G. Chem. Rev. 2002, 102, 3495; (b) Yang, H.; Zhang, L.; Su,
W.; Yang, Q.; Li, C. J. Catal. 2007, 248, 204; (c) Bogza, M.; Oeser, T.; Blümel, J. J.
Organomet. Chem. 2005, 690, 3383; (d) Rechavi, D.; Lemaire, M. Chem. Rev. 2002,
1
02, 3467.
2.
(a) O’Leary, P.; Krosveld, N. P.; De Jong, K. P.; van Koten, G.; Klein Gebbink, R. J.
M. Tetrahedron Lett. 2004, 45, 3177–3180; (b) McDonagh, C.; O’Conghaile, P.;
Klein Gebbink, R. J. M.; O’Leary, P. Tetrahedron Lett. 2007, 48, 4387.
3
4
.
.
(a) Evans, D. A.; Sweeney, Z. K.; Rovis, T.; Tedrow, J. S. J. Am. Chem. Soc. 2001, 123,
12095; (b) Evans, D. A.; Wu, J. J. Am. Chem. Soc. 2005, 127, 8006.
(a) Mikami, K.; Shimizu, M. Chem. Rev. 1992, 92, 1021; (b) Berrisford, D. J.; Bolm,
C. Angew. Chem., Int. Ed. Engl. 1995, 34, 1717; (c) Grachan, M. L.; Tudge, M. T.;
Jacobsen, E. N. Angew. Chem., Int. Ed. 2008, 47, 1469; (d) Mikami, K.; Terrada, M..
In Comprehensive Asymmetric Catalysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H.,
Eds.; Springer: Berlin, 1999; Vol. 3,. Chapter 32.3.
5
.
.
(a) Nakagawa, K.; Okano, T.; Ozono, K.; Kato, S.; Kubodera, N.; Ohba, S.; Itoh, Y.;
Mikami, K. J. Fluorine Chem. 2007, 128, 654; (b) Muratake, H.; Natsume, M.
Tetrahedron 2006, 62, 7056; (c) Alcaide, B.; Pardo, C.; Rodriguez-Ranera, C.;
Rodriquez-Vicente, A. Org. Lett. 2001, 3, 4205.
(a) Luo, H.-K.; Khim, L. B.; Schumann, H.; Lim, C.; Jie, T. X.; Yang, H.-Y. Adv. Synth.
Catal. 2007, 349, 1781; (b) Mikami, K.; Koizumi, Y.; Osawa, A.; Terada, M.;
Takayama, H.; Nakagawa, K.; Okano, T. Synlett 1999, 1899.
6
Figure 3. Molecular model of the interaction of a scandium(III)PYBOX3 triflate-
ethyl glyoxylate complex with a gem silanol.
7
.
.
Thorhauge, J.; Roberson, M.; Hazell, R. G.; Jørgensen, K. A. Chem. Eur. J. 2002, 8,
1888.
8
Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.;Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.;
Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; K. eith, T.;
Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A., GAUSSIAN 03,
Revision D.01; Gaussian, Wallingford, CT, 2004.
the coordination chemistry during the reaction. We are now
undertaking a more substantial modelling study in an attempt to
identify catalysts and reactions which may be similarly effected
by the presence of a silica support.
Acknowledgements
The authors gratefully acknowledge NUI-Galway for a fellow-
ship (CMD). This publication has emanated from research con-
ducted with the financial support of Science Foundation Ireland
(
RFP). The authors wish to acknowledge the SFI/HEA Irish Centre
for High-End Computing (ICHEC) for the provision of computa-
9. Hlatky, G. Chem. Rev. 2000, 100, 1347.