Polymer-supported chiral Co(salen) complexes: Synthetic applications and mechanistic investigations in the hydrolytic kinetic resolution of terminal epoxides

Article abstract of DOI:10.1021/ja984410n

This paper describes the synthesis of polystyrene- and silica-bound chiral Co(salen) complexes and their application in asymmetric catalysis. A general method for the covalent attachment of salen complexes to both types of support has been devised, and the corresponding immobilized cobalt derivatives are shown to be efficient and highly enantioselective catalysts for the hydrolytic kinetic resolution (HKR) of terminal epoxides. These systems provide practical solutions to certain technical difficulties associated with the isolation of reaction products from the HKR. Removal of the supported catalyst by filtration and repeated recycling is demonstrated with no loss of reactivity or enantioselectivity. The enantioselective addition of phenols to terminal epoxides mediated by this catalyst system provides a facile, high-yielding synthesis of the corresponding enantioenriched aryl ethers. The immobilized catalysts have been adapted to a continuous flow process for the generation of reaction products in high yield and ee, requiring only very simple techniques for product purification. The mechanism by which these catalysts perform highly efficient and enantioselective epoxide ring opening has been addressed using a silica- bound Co(salen) complex. A dramatic correlation between the degree of catalyst site-isolation and reaction rate has been observed, consistent with a cooperative bimetallic mechanism in these reactions.

Full text of DOI:10.1021/ja984410n

J. Am. Chem. Soc. 1999, 121, 4147-4154
4147
Polymer-Supported Chiral Co(Salen) Complexes: Synthetic
Applications and Mechanistic Investigations in the Hydrolytic Kinetic
Resolution of Terminal Epoxides
D. Allen Annis and Eric N. Jacobsen*
Contribution from the Department of Chemistry and Chemical Biology and the Institute of Chemistry and
Cell Biology, HarVard UniVersity, Cambridge, Massachusetts 02138
ReceiVed December 22, 1998
Abstract: This paper describes the synthesis of polystyrene- and silica-bound chiral Co(salen) complexes and
their application in asymmetric catalysis. A general method for the covalent attachment of salen complexes to
both types of support has been devised, and the corresponding immobilized cobalt derivatives are shown to be
efficient and highly enantioselective catalysts for the hydrolytic kinetic resolution (HKR) of terminal epoxides.
These systems provide practical solutions to certain technical difficulties associated with the isolation of reaction
products from the HKR. Removal of the supported catalyst by filtration and repeated recycling is demonstrated
with no loss of reactivity or enantioselectivity. The enantioselective addition of phenols to terminal epoxides
mediated by this catalyst system provides a facile, high-yielding synthesis of the corresponding enantioenriched
aryl ethers. The immobilized catalysts have been adapted to a continuous flow process for the generation of
reaction products in high yield and ee, requiring only very simple techniques for product purification. The
mechanism by which these catalysts perform highly efficient and enantioselective epoxide ring opening has
been addressed using a silica-bound Co(salen) complex. A dramatic corrrelation between the degree of catalyst
site-isolation and reaction rate has been observed, consistent with a cooperative bimetallic mechanism in these
reactions.
Introduction
with lower enantioselectivities or efficiencies than the solution-
phase counterparts. Also, the practical advantages gained by
covalent attachment to a support are often outweighed by the
added complexity associated with synthesizing the appropriately
modified chiral ligands.7
The recently developed chiral, salen-based catalysts for the
enantioselective ring-opening of meso epoxides and kinetic
resolution of terminal epoxides8 are appealing candidates for
immobilization on solid support. The catalysts are readily
prepared from inexpensive components, and are amenable to
The covalent attachment of homogeneous catalysts to in-
soluble polymer supports has been studied widely as an attractive
strategy for extending the practical advantages of heterogeneous
catalysis to homogeneous systems.1 The potential benefits of
heterogenization include facilitation of catalyst separation from
reagents and reaction products, simplification of methods for
catalyst recycle, and the possible adaptation of the immobilized
catalyst to continuous-flow processes. The use of polymer-
supported reagents in combinatorial synthesis is also a topic of
growing interest.2 The possibility of promoting selective reac-
tions between libraries of reacting partners with subsequent
separation of catalyst by mechanical means renders polymer-
bound catalysts worthy of investigation in this regard.3 Ideally,
attachment of catalysts to a solid support can also provide insight
into reaction mechanism, particularly with respect to issues of
intercatalyst interaction.4
(6) For recent examples and related efforts, see: (a) Han, H.; Janda, K.
D Tetrahedron Lett. 1997, 38, 1527-1530. (b) Wan, K. T.; Davis, M. Nature
1994, 370, 449-450. (c) Lohray, B. B.; Nandanan, E.; Bhushan, V.
Tetrahedron Asym. 1996, 7, 2805-2808. (d) Altava, B.; Burguete, M. I.;
Escudar, B.; Luis, S. V.; Salvador, R. V. J. Org. Chem. 1997, 62, 3126-
3134. (e) Nanadanan, E.; Sudalai, A.; Ravindranathanan, T. Tetrahedron
Lett. 1997, 38, 2577-2580. (f) Huang, W. S.; Hu, Q. S.; Zheng, X. F.;
Anderson, J.; Pu, L. J. Am. Chem. Soc. 1997, 119, 4313-4314. (g)
Kamahori, K.; Ito, K.; Itsuno, S. J. Org. Chem. 1996, 61, 8321-8324. (h)
Han, H.; Janda, K. D. J. Am. Chem. Soc. 1996, 118, 7632-7633. (i)
Salvadori, P.; Pini, D.; Petri, A. J. Am. Chem. Soc. 1997, 119, 6929-6930.
(j) Vidal-Ferran, A.; Bampos, N.; Moyano, A.; Pericas, M. A.; Riera, A.;
Sanders, J. K. M. J. Org. Chem. 1998, 63, 6309-6318.
(7) It should be noted, however, that these problems are at least partly
avoided in the recently developed methods for immobilizing chiral catalysts
noncovalently within membranes or polymer matrixes. See: (a) Vancelecom,
I. F. J.; Tas, D.; Parton, R. F.; Van de Vyver, V.; Jacobs, P. A. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1346-1348. (b) Janssen, K. B. M.; Laquiere,
I.; Dehaen, W.; Parton, R. F.; Vankelecom, I. F. J.; Jacobs, P. A. Tetrahedron
Asym. 1997, 8, 3481-3487.
(8) (a) Water: Tokunaga, M.; Larrow, J. F.; Kakiuchi, F.; Jacobsen, E.
N. Science 1997, 277, 936-938. (b) Trimethylsilyl azide: Martinez, L. E.;
Leighton, J. L.; Carsten, D. H.; Jacobsen, E. N. J. Am. Chem. Soc. 1995,
117, 5897-5898. Schaus, S. E.; Larrow, J. F.; Jacobsen., E. N. J. Org.
Chem. 1997, 62, 4197-4199. Larrow, J. F.; Schaus, S. E.; Jacobsen., E. N.
J. Am. Chem. Soc. 1996, 118, 7420-7421. (c) Carboxylic acids: Jacobsen,
E. N.; Kakiuchi, F.; Konsler, R. G.; Larrow, J. F.; Tokunaga, M. Tetrahedron
Lett. 1997, 38, 773-776. (d) Thiols: Wu, M.; Jacobsen, E. N. J. Org. Chem.
1998, 63, 5252-5254.
Whereas immobilization of achiral catalysts onto polymer
supports has enjoyed growing application,5 the development of
practical polymer-immobilized asymmetric catalysts has proven
highly challenging.6 Immobilization often results in catalysts
(1) For a recent review see: Shuttleworth, S. J.; Allin, S. M.; Sharma,
P. K. Synthesis 1997, 1217-1239.
(2) (a) Kaldor, S. P.; Siegel, M. G. Curr. Opin. Chem. Biol. 1997, 1,
101. (b) Thompson, L. A.; Ellman, J. A. Chem. ReV. 1996, 96, 555-600.
(3) For a discussion of the relevance of absolute stereocontrol in library
synthesis, see: Annis, D. A.; Helluin, O.; Jacobsen, E. N. Angew. Chem.,
Int. Ed. Engl. 1998, 37, 1907-1909.
(4) (a) Collman, J. P.; Belmont, J. A.; Brauman, J. I. J. Am. Chem. Soc
1983, 105, 7288-7294. (b) Tollner, K.; Popovitv-Biro, R.; Lahav, M.;
Milstein, D. Science 1997, 278, 2100-2102.
(5) See ref 1 and also: (a) Kobayashi, S.; Nagayama, S. J. Am. Chem.
Soc. 1998, 120, 2985-2986. (b) Nagayama, S.; Endo, M.; Kobayashi, S.
J. Org. Chem. 1998, 63, 6094-6095. (c) Kobayashi, S.; Nagayama, S. J.
Org. Chem. 1996, 61, 2256-2257.
10.1021/ja984410n CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/20/1999

4148 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
Annis and Jacobsen
Scheme 1
Figure 1.
modification for attachment to a solid support.9,10 These systems
have been demonstrated to be quite efficient, even at large
(>100 kg) scale,11 and the catalysts are recyclable without loss
of selectivity or reactivity. Moreover, mechanistic studies
indicate that the epoxide ring-opening reactions proceed through
a mechanism involving cooperative interaction between catalyst
units,12 so enforcement of a high local concentration of catalyst
by attachment to a high-loading support may be beneficial to
their reactivity.
Herein we report the synthesis of polystyrene- and silica-
bound chiral Co(salen) complexes and their application to the
kinetic resolution of terminal epoxides. These systems allow
highly simplified catalyst separation from product mixtures, and
recovered catalysts are demonstrated to retain full activity and
enantioselectivity. In addition, the effects on reactivity of varying
catalyst loading on the solid support have provided a new tool
for studying the mechanism of epoxide ring-opening.
Addition of commercial 90 µm beads of hydroxymethylpoly-
styrene, derivatized as the corresponding 4-nitrophenyl carbon-
ate, to a solution of 3 in DMF in the presence of diisopropyl-
ethylamine afforded the carbonate-linked immobilized ligand
as bright yellow beads. These were rinsed thoroughly with CH2-
Cl2 and dried in vacuo, and metal insertion was accomplished
by adding a solution of Co(OAc)2 in MeOH/toluene. Oxidation
to the Co(III) state occurred upon rinsing the resulting beads
with AcOH/toluene in air. Complex 5 was isolated as dark red
beads after thorough rinsing and drying in vacuo; elemental
analysis indicated incorporation of 160 µmol Co(salen) per gram
of resin.
An alternative, and more practical, synthesis of 5 by resin
capture of 3 generated from a crude ligand preparation is
outlined in Figure 2. Synthesis of the salen ligand precursor by
utilization of an excess of di-tert-butyl salicaldehyde relative
to 2,5-dihydroxy-3-tert-butylbenzaldehyde yielded a statistical
6:1 ratio of 3 to diphenolic ligand 4. Addition of the carbonate-
derived polystyrene resin to the mixture allowed for the selective
capture of the hydroxy-terminated ligands 3 and 4, while the
soluble tetra-tert-butyl-substituted ligand 2 was washed away
from the polymer-bound product. Complex 5 prepared in this
manner displayed identical reaction rates and enantioselectivy
to material prepared using purified 3, but required no chro-
matographic purification of unsymmetrical salen ligands for its
preparation. Incorporation of a minor amount of the diphenolic
ligand 4 in the resin-bound catalyst, to the extent that it does
occur, has no apparent deleterious effect on catalyst reactivity
or enantioselectivity.
Polystyrene-bound complex 5 was examined for its efficacy
in the hydrolytic kinetic resolution of epichlorohydrin, 6. This
substrate is susceptible to chloride-catalyzed racemization,14,15
and its isolation from catalyst by distillation of HKR product
mixtures can lead to diminution of epoxide ee, particularly when
carried out at large scale.16 Therefore, an improved method for
the removal of the catalyst prior to the isolation of enantioen-
riched 6 could hold practical significance, and attachment of
the catalyst to a solid support could allow for such a separation
by simple filtration. The results of a series of experiments in
which polystyrene-bound complex 5 was used and recycled four
times in the HKR of 6 are summarized in Figure 3. Complete
consumption of the reactive enantiomer occurred within 3 h
Results and Discussion
A. Synthesis and Applications of Polystyrene-Bound
Chiral Co(Salen) Complexes. The Co(salen) complex 1,
derived from the chiral salen ligand 2, has been demonstrated
to be a highly efficient and enantioselective catalyst for the
hydrolytic kinetic resolution (HKR) of terminal epoxides.8a
With the goal of accessing variants of 1 covalently linked to
polymer supports, we have developed an effective method for
the covalent attachment of the corresponding mono-phenol 3
to polystyrene beads (Scheme 1). Compound 3 has been utilized
previously for the construction of soluble dimeric catalysts.13
In that context, it was shown that tethering the catalyst through
the 5-substituent of the salicaldehyde has no adverse effect on
the enantioselectivity of the asymmetric ring-opening of terminal
epoxides.
(9) Synthesis of 2: Larrow, J. F.; Jacobsen, E. N.; Gao, Y.; Hong, Y.;
Nie, X.; Zepp, C. M. J. Org. Chem. 1994, 59, 1939.
(10) Several previous efforts to immobilize (salen)Mn-based epoxidation
catalysts have been described. However, the tendency of (salen)Mn
complexes to undergo decomposition under conditions of catalysis limits
the attractiveness of immobilization strategies for these particular systems.
See ref 7a and: (a) Minutolo, F.; Pini, D.; Salvadori, P. Tetrahedron Lett.
1996, 37, 3375-3378. (b) De, B. B.; Lohray, B. B.; Sivaram, S.; Dhal, P.
K. Tetrahedron: Asymm. 1995, 6, 2105-2108. (c) Das, B. C.; Iqubal, J.
Tetrahedron Lett. 1997, 38, 2903-2906. (d) Chang, S. Ph.D. Thesis,
Harvard University, 1996. (e) Ogunwumi, S. G.; Bein, T. J. Chem. Soc.,
Chem. Commun. 1997, 901. (f) Angelino, M. D.; Laibinis, P. E. Macro-
molecules 1998, 22, 7581-7587. (g) Canali, L.; Cowan, E.; Gibson, C. L.;
Sherrington, D. C.; Deleuze, H. J. Chem. Soc., Chem. Commun. 1998,
2561-2562. (h) Minutolo, F.; Pini, D.; Petri, A.; Salvadori, P. Tetrahe-
dron: Asymm. 1996, 7, 2293-2302.
(14) Schaus, S. E.; Jacobsen, E. N. Tetrahedron Lett. 1996, 37, 7937-
7940.
(15) Furrow, M. E.; Schaus, S. E.; Jacobsen, E. N. J. Org. Chem. 1998,
63, 6776-6777.
(16) Cummins, J.; Waldram, D.; Bone. S.; ChiRex Inc., Dudley, U.K.
Private communication.
(11) Pettman, R. B. Spec. Chem. 1997, 118-120.
(12) Hansen, K. B.; Leighton, J. L.; Jacobsen, E. N. J. Am. Chem. Soc.
1996, 118, 10924-10925.
(13) Konsler, R. G.; Karl, J.; Jacobsen, E. N. J. Am. Chem. Soc. 1998,
120, 10780-10781.

Polymer-Supported Chiral Co(Salen) Complexes
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4149
Figure 4. HKR of 4-hydroxy-1-butene oxide with solid-phase catalyst
5.
Figure 5. Dynamic kinetic resolution of epibromohydrin with polymer-
bound catalyst 5.
under air with 9:1 toluene/acetic acid, was recyclable with no
apparent loss of reactivity or selectivity (cycles 2-5). Combina-
tion of the crude organic-soluble products of the recycle exper-
iments and concentration under reduced pressure provided R-6
in 99% ee and in 41% isolated yield based on racemate used.
Another illustration of practical advantages conferred by the
use of immobilized catalyst 5 was provided through the HKR
of epoxy alcohol 8 (Figure 4). The ring-opened product, triol
9, is a direct precursor to 3-hydroxytetrahydrofuran,17
a
component of the HIV protease inhibitor VX-478.18 While 9
can be accessed in high enantiomeric purity via HKR using
catalyst 1, its isolation from catalyst residue is difficult due to
foaming upon distillation.19 Utilization of polymer-bound
complex 5 provided a straightforward solution to this problem.
Five reaction cycles were performed with a single catalyst batch,
and the filtrates and washings from each run were combined.
The enantioselectivity for this substrate is lower than that
observed for most terminal epoxides; however, high ee product
could be accessed by stopping the reaction at moderate (ca. 40%)
conversion of starting material. The resulting crude product
mixture contained epoxide 8 in 59% ee and triol 9 in 94.4%
ee. Filtration and removal of volatile materials in vacuo yielded
9 in 36% isolated yield, 94.4% ee, and > 98% chemical purity.
The recently reported dynamic hydrolytic kinetic resolution
of epibromohydrin, 10, catalyzed by complex 1, represents an
appealing case for the application of an insoluble catalyst.
Starting material is converted quantitatively to a single, non-
volatile product, so the use of catalyst 5 would allow product
isolation by simple filtration and solvent removal. Results of
the dynamic HKR of racemic 10 by catalyst 5 are shown in
Figure 5. The reaction was repeated with the same catalyst batch
through five cycles. The products obtained from each reaction
were combined and the solvents used for rinsing the beads were
removed in vacuo. Diol 11 was thus isolated in 94% overall
yield, 96% ee, and in 90% chemical purity, the only detectable
impurity being 1,3-dibromo-2-propanol.
Figure 2. Resin-capture synthesis of polystyrene-bound chiral Co-
(salen) complexes.
Very recently complex 1 has been discovered to be an
efficient catalyst for the kinetic resolution of terminal epoxides
by the addition of phenols.20 As numerous epoxide-phenol
Figure 3.
using as low as 0.25 mol % catalyst loadings. The catalyst was
recoverable by simple filtration, and the filtrate and combined
dichloromethane rinsings could be extracted with water to
remove the diol product 7. Use of the polymer-supported catalyst
in this reaction thus avoided the necessity of catalyst separation
by distillation. The catalyst, reoxidized between cycles by rinsing
(17) Tandon, V. K.; van Leusen, A. M.; Wynberg, H. J. Org. Chem.
1983, 48, 2767-2769.
(18) Kim, E. E.; Baker, C. T.; Dwyer, M. D.; Murcko, M. A.; Rao, B.
G.; Tung, R. D.; Navia, M. A. J. Am. Chem. Soc. 1995, 117, 1181-1182.
(19) Hansen, K. B. Ph.D. Thesis, Harvard University, 1998.
(20) Ready, J. M.; Jacobsen, E. N. Submitted for publication.

4150 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
Annis and Jacobsen
Figure 6. A possible application of the enantioselective ring-opening
of epoxides by phenols to the combinatorial synthesis of novel
compounds.
Figure 7.
Figure 9. A possible application of the enantioselective ring-opening
of epibromohydrin with phenols in combinatorial synthesis.
However, the crude reaction mixture could be converted
quantitatively to 15 after catalyst removal by filtration by the
addition of solid KOH. Subsequent filtration and removal of
volatile materials provided the epoxide product in high yield,
>99% purity and excellent enantiomeric excess without further
purification. Further elaboration of the aryl glycidyl ether
provides a straightforward strategy for the synthesis of combi-
natorial libraries of the general structure 16, an example of which
includes the antihypertensive agent Propranolol, 17 (Figure 9).
B. Synthesis and Applications of Silica-Bound Chiral Co-
(Salen) Complexes. We have also explored the use of silica as
an insoluble support for the covalent attachment of chiral Co-
(salen) complexes. With the attributes of inflexibility and
noncompressibility, silica provides a stationary phase amenable
to incorporation in continuous flow reactors, as well as a surface
upon which site isolation of supported catalysts may be more
carefully defined than on a flexible polymer backbone such as
polystyrene. The preparation of ligand 19, suitable for attach-
ment to a silica surface, is shown in Figure 10.
Figure 8.
combinations can be envisioned, with subsequent derivatization
steps providing additional diversity, simplified purification
afforded by the use of polymer-bound catalysts could enable
the facile combinatorial synthesis of libraries of stereochemically
defined ring-opened products (Figure 6).
The bifunctional tether 18, bearing one end reactive toward
silica and the other appropriately functionalized for attachment
to the ligand 3, was prepared in three steps without need for
chromatographic purification starting from 10-undecylenic acid.
Carbodiimide-mediated coupling to 3 afforded 19 as a yellow
oil, which could be purified chromatographically with good
recovery. Attachment of the silyl ether to silica using the
conditions of Pirkle22 provided the silica-bound ligand as a
yellow powder after filtration and rinsing. Addition of a solution
of Co(OAc)2 in methanol/toluene yielded silica-bound Co(salen)
complex 20 as a red-brown powder.
The efficacy of complex 20 in the HKR reaction was tested
using styrene oxide, 21, as a model substrate (Figure 11). The
HKR of 21 using silica-bound catalyst 20 was found to be
efficient and highly enantioselective. Using only 0.7 mol %
catalyst, in 3 h the reaction had proceeded to nearly complete
conversion of the reactive enantiomer of the starting material,
forming the diol product 22 in excellent ee. The krel for this
particular substrate was calculated to be 58 based on the ee of
The addition of phenol, 12, to 1-hexene oxide was examined
as a model system to test the efficacy of immobilized catalyst
5 for phenol additions to epoxides (Figure 7). In the presence
of low loadings of 5, this reaction provided excellent yields of
the ring-opened product 13 in very high enantiomeric excess.
Addition of tris(trifluoromethyl)methanol, a volatile, nonnucleo-
philic protic acid additive, was found to accelerate this reaction
with no reduction in enantioselectivity or yield, presumably by
helping to maintain the catalyst in the CoIII oxidation state. By
carrying out the reactions to low conversion of epoxide (ca.
25%), the ring-opened product was obtainable in very high
ee.21
To further explore the utility of polymer-bound catalysts for
the enantioselective ring-opening of terminal epoxides by
phenols, epibromohydrin was evaluated as a substrate for the
reaction (Figure 8). Though the reaction proceeded with high
efficiency and enantioselectivity, the bromohydrin product 14
was found to undergo ring-closure slowly to the phenyl glycidyl
ether 15 under the conditions of the ring-opening reaction.
(21) Kagan, H. B. in Topics in Stereochemistry; Eliel, E. L., Wilen, S.
H., Eds.; Wiley: New York, 1987; Vol. 14, pp 249-330.
(22) Pirkle, W. H.; Welch, C. J.; Lamm, B. J. Org. Chem. 1992, 57,
3854-3860.

Polymer-Supported Chiral Co(Salen) Complexes
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4151
Figure 12. HKR of 4-hydroxy-1-butene oxide with silica-bound Co-
(salen) complex (R,R)-20.
Figure 13. Continuous-flow apparatus for the HKR of 8 over silica-
bound catalyst 20.
Figure 10.
Figure 14. Continuous-flow HKR of 4-hydroxy-1-butene oxide over
silica-bound Co(salen) complex (R,R)-20.
enantioselectivity comparable to the polystyrene-bound catalyst
variant, affording product in 93% ee at 34% conversion of
epoxide (Figure 12).
Figure 11. HKR of styrene oxide with silica-bound Co(salen) complex
(R,R)-20.
A simple apparatus for the continuous-flow HKR of 8 over
a packed bed of catalyst 20 was constructed using a standard
laboratory syringe pump as a solvent delivery device, an HPLC
injector valve equipped with an injection loop for loading
reagents into the reactant stream, a stainless steel HPLC column
packed with 20 as the catalyst bed, and a receiver flask for the
collection of product. A schematic diagram of the apparatus is
shown in Figure 13.
The results of this continuous-flow HKR of 8 over silica-
bound catalyst 20 are summarized in Figure 14. A solution of
8 in 5.9:1 THF/water was injected into a solvent stream of 5.9:1
THF/water and passed through the column containing 20. The
column was then rinsed with additional THF/water and the
combined eluates were concentrated in vacuo, removing excess
solvent and remaining starting material, to yield the desired triol
9 in good yield and high enantiomeric excess. The column could
the recovered epoxide. The ee of the diol product is elevated
relative to the unreacted epoxide as a result of ring-opening at
the benzylic position of the less reactive enantiomer of epoxide.
This affords the same enantiomer of diol as does reaction at
the terminal position of the more reactive epoxide enantiomer.
C. Application of Silica-Bound Co(Salen) to a Continuous-
Flow System for HKR. Continuous-flow reaction systems
incorporating an immoibilized catalyst can have clear practical
advantages relative to homogeneous systems.23 Epoxy alcohol
8 was selected as a model substrate for evaluating this
methodology in the HKR with catalyst 20. This particular
substrate has the property that all reaction components (substrate,
water, product), with the exception of the catalyst, are miscible
under the reaction conditions.24 The reaction was first examined
in a batchwise manner. The HKR of 8 was found to proceed in
the presence of silica-bound catalyst 20 with reactivity and
(24) While only select epoxide substrates form a single phase in the HKR
and are thus amenable to continuous flow processes, the ring-opening of
epoxides by phenols occurs in a single-phase reaction mixture. Application
of continuous flow processes to this class of asymmetric transformation
holds considerable promise and is currently under investigation.
(23) (a) Kamahori, K.; Ito, K.; Itsuno, S. J. Org. Chem. 1996, 61, 8321-
8324. (b) Itsuno, S.; Ito, K.; Maruyama, T.; Kanda, N.; Hirao, A.; Nakahama,
S. Bull. Chem. Soc. Jpn. 1986, 59, 3329-3331.

4152 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
Annis and Jacobsen
Figure 15. Proposed cooperative mechanism for the hydrolytic kinetic
resolution of terminal epoxides.
Figure 17. Plot of the rate of the HKR of styrene oxide and
epichlorohydrin vs the probability of catalyst interaction on the silica
surface.
for the absolute amount of catalyst present in each experiment,
versus the probability that any catalyst can interact with any
other catalyst is shown in Figure 17.
A remarkable correlation between reaction rate and site
isolation is evident from this analysis. While there is essentially
no observable reaction at lower catalyst concentrations on the
silica surface, increasingly higher rates of reaction are observed
for systems wherein high surface loading allows, and ultimately
favors, the interaction of bound catalysts. These results lend
strong support to a cooperative bimetallic mechanism. It is
interesting to note that, in the case of polystyrene-bound catalyst
5, the flexibility of the support appears to facilitate intercomplex
interactions, as the reactions proceed with rates comparable to
the solution-phase catalyst 1 and to the high-loading silica-bound
catalyst 20.
Figure 16. Plot of epichlorohydrin conversion vs time in the HKR
using various loadings of catalyst on the silica surface.
be regenerated by elution of a small volume of acetic acid/
toluene; a second cycle of the HKR of 8 yielded product 9 with
ee and yield comparable to the first cycle. Given the straight-
forward preparation of the immobilized catalyst 20, and the
effectiveness and experimental ease of its use in the synthesis
of enantioenriched products, such catalysts could hold promise
for the practical preparation of chiral building blocks in a
continuous-flow manner.
Conclusions
D. Mechanistic Investigation of the HKR of Terminal
Epoxides Using Silica-Bound Chiral Co(Salen) Complexes.
Preliminary studies into the mechanism of the hydrolytic kinetic
resolution of terminal epoxides using Co(salen) complex 1
indicate that the reaction rate exhibits a second-order dependence
on catalyst concentration.8a By analogy to the mechanism of
the Cr(salen)-catalyzed enantioselective ring-opening of meso
epoxides by hydrazoic acid,12 the kinetic behavior exhibited by
the Co(salen) catalyst in the HKR suggests a cooperative
mechanism for epoxide ring-opening such as that illustrated in
Figure 15.
A direct means of examining the degree of catalyst cooper-
ativity is to assess the effect of site-isolation of the catalyst on
an inflexible solid support.4 Batches of catalyst 20 were prepared
with various loadings of Co(salen) on the silica surface, and
the rates of the HKR of styrene oxide and epichlorohydrin were
determined using these different catalyst batches. The enantio-
selectivity of the reaction was not affected by variation of the
catalyst loading. Kinetic profiles for epichlorohydrin conversion
as a function of different batches of 20 are shown in Figure 16.
Analysis of the kinetic data reveals clearly that there is a
minimum loading of catalyst on the silica surface required for
the reaction to proceed.
This work demonstrates the synthetic applicability and
mechanistic utility of polymer-supported chiral Co(salen)
complexes. The benefits usually postulated to be associated with
immobilization of soluble catalystssfacilitated product purifica-
tion, catalyst recycling, and adaptation to continuous flow
methodologyshave all been realized in the solid-supported
chiral Co(salen) complexes 5 and 20. Use of these catalyst
systems proved particularly advantageous in the HKR of
substrates such as epichlorohydrin 6 and epoxyalcohol 8 where
purification had proven problematic with homogeneous catalysts.
However, the immobilized salen complexes are likely to hold
more general utility with possible large scale applications using
immobilized catalyst beds and small scale applications to library
synthesis. These are the subject of ongoing investigation.
Experimental Section
General. Tetrahydrofuran, toluene, and TBME were distilled from
sodium/benzophenone ketyl prior to use. Commercially available
starting materials were used as received or distilled over calcium
hydride. Analytical thin-layer chromatography was performed on EM
Reagent 0.25 mm silica gel 60-F plates. Flash chromatography was
performed using EM silica gel 60 (230-400 mesh).
Gas chromatographic (GC) analyses were conducted on Hewlett-
Packard HP5890 Series II gas chromatographs equipped with the
following commercially available capillary columns: HP-5 (30 m ×
0.25 mm × 0.25 µm film; Hewlett-Packard), Cyclodex-B (30 m ×
0.25 mm × 0.25 µm film; J & W Scientific), and Chiraldex γ-TA (20
m × 0.25 mm × 0.25 µm film; Advanced Separation Technologies,
Inc.).
The probability that any catalyst on the silica surface can
interact with any other catalyst was calculated according to the
method of Collman4a from the loading level of Co(salen) on
the silica surface (established by elemental analysis), an estima-
tion made by molecular modeling of the length of the tether
between the silica surface and the salen ligand, and the surface
area of the silica particulate used in the synthesis. A plot of the
rates of the HKR of epichlorohydrin and styrene oxide, corrected
Infrared spectra were recorded on a Mattson Galaxy Series FTIR
3000 spectrometer. 1H NMR spectra were recorded on a Bruker DMX-
500, AM-500, AM-400, or AM-300 spectrometer. Chemical shifts are

Polymer-Supported Chiral Co(Salen) Complexes
J. Am. Chem. Soc., Vol. 121, No. 17, 1999 4153
reported in ppm downfield from tetramethylsilane as an internal
standard. 13C NMR spectra were recorded on a Bruker AM-500 (125
MHz) or AM-400 (100 MHz) spectrometer with broadband proton
decoupling. Chemical shifts are reported in ppm downfield from
tetramethylsilane with the solvent reference as the internal standard
(deuteriochloroform: δ 77.0 ppm). Fast atom bombardment (FAB) mass
spectra were obtained on a JEOL AX-505 or SX-102 high-resolution
magnetic sector mass spectrometer by the Harvard University Mass
Spectrometry Laboratory. HPLC analysis was performed on a Hewlett-
Packard Series 1050 quaternary pump HPLC system equipped with an
HP 1050 diode array detector.
Cl2, 9:1 toluene/HOAc, CH2Cl2, MeOH, and CH2Cl2 and then dried in
vacuo to yield the product as very dark red beads. The IR spectrum
(KBr pellet) contained strong absorbances at 1755 and 1630 cm-1
.
Elemental analysis indicated 0.94% Co, corresponding to a final loading
of 0.16 mmol/g of Co.
HKR of Epichlorohydrin 6 with Polystyrene-Bound Co(Salen)
Complex 5. To complex 5 (160 mg, 26 µmol) was added CH2Cl2 (0.78
mL), 6 (0.78 mL, 10 mmol), and H2O (0.126 mL, 7.0 mmol). The
reaction was gently stirred for 3 h. An aliquot taken for GC analysis
indicated 6 to be present in >99% ee (CD γ-ΤΑ, 40 °C, tR ) 8.5, 10.6
min); 92.4% ee of 7 (Cyclodex B, 75 °C, tR ) 7.4, 8.0 min, as the
acetonide prepared using 1% w/v camphorsulfonic acid in dimethoxy-
propane). The conversion was calculated from measurement of the ee
of the starting material and product, having established that no
racemization occurs under these reaction conditions, to be 52%,
indicating a krel of 133.26 The reaction was filtered, and the polymer
beads were rinsed with 10 mL of CH2Cl2. The combined filtrates were
washed with water (3 × 2 mL), dried over MgSO4, filtered, and
concentrated by rotary evaporation, carefully controlling the pressure
at 400 mbar, to yield 0.38 g of 6 (41%) with ee >99%. The polymer
beads were rinsed sequentially with MeOH, CH2Cl2, 9:1 toluene/HOAc,
CH2Cl2, MeOH, and CH2Cl2 and dried in vacuo prior to recycle.
HKR of 4-Hydroxy-1-butene Oxide 8 with Polystyrene-Bound
Co(Salen) Complex 5. To a solution of 827 (50 mg, 0.57 mmol) in 50
µL of THF was added 5 (18 mg, 2.8 µmol) and H2O (7.2 µL, 0.40
mmol) and the reaction was gently stirred at room temperature for 2.5
h. An aliquot taken for GC analysis indicated 8 to be present in 73%
ee (γ-ΤΑ, 75 °C, tR ) 9.0, 11.9 min). The ee of product 9 was
established by first converting the triol to 3-hydroxytetrahydrofuran
by concentrating the aliquot in vacuo to remove remaining 8, adding
catalytic p-toluenesulfonic acid, and heating the sample at 100 °C for
1 h. The ee of the cyclized product was determined to be 94% (Cyclodex
B, 90 °C, tR ) 6.6, 6.8 min, as the acetate prepared using acetyl chloride/
pyridine in CH2Cl2), corresponding to a conversion of 44%, and krel of
71. The reaction was filtered and the polymer beads rinsed with MeOH
into a tared flask. The beads were then rinsed sequentially with MeOH,
CH2Cl2, 9:1 toluene/HOAc, CH2Cl2, MeOH, and CH2Cl2 and dried in
vacuo prior to recycle. After five recycles, the combined filtrates were
concentrated in vacuo to yield 9 as a colorless oil (108 mg, 36% yield,
94.4% ee). The ee of 8 in the combined filtrate was determined to be
59%, corresponding to an overall conversion of 38%, and krel of 63.
Dynamic HKR of Epibromohydrin 10 with Polystyrene-Bound
Co(Salen) Complex 5. To complex 5 (16 mg, 2.6 µmol) was added
THF (43 µL), 10 (43 µL, 0.50 mmol), and H2O (13.5 µL, 0.75 mmol).
The reaction was gently stirred at room temperature for 24 h. An aliquot
taken for GC analysis indicated complete consumption of 10 and 11
to be present in 96% ee (Cyclodex B, 75 °C, tR ) 15.6, 17.0 min, as
the acetonide prepared using 1% w/v camphorsulfonic acid in dimeth-
oxypropane). The reaction was filtered and the polymer beads rinsed
with CH2Cl2 into a tared flask. The beads were then rinsed sequentially
with MeOH, CH2Cl2, 9:1 toluene/HOAc, CH2Cl2, MeOH, and CH2Cl2
and dried in vacuo prior to recycle. After five recycles, the combined
filtrates were concentrated in vacuo to yield 11 as a colorless oil (352
Synthesis of (R,R)-3. (a) 3-tert-Butyl-2,5-dihydroxybenzaldehyde:
A tetrabutylammonium fluoride solution (1.0 M in THF, 30.5 mL, 30.5
mmol, Aldrich) was added dropwise to a solution of 3-tert-butyl-2-
hydroxy-5-triisopropylsiloxybenzaldehyde25 (9.0 g, 25.4 mmol) in 60
mL of THF at -78 °C and allowed to warm to room temperature over
3 h. The reaction mixture was poured into 100 mL of water and
extracted (100 mL × 3) with diethyl ether. The organic layers were
combined, washed with saturated aqueous ammonium chloride (100
mL × 2), dried over magnesium sulfate, and concentrated. The product
was purified by silica gel chromatography (diethyl ether/hexanes, 1:4)
to give 3.5 g (71% yield) of product as a yellow solid. Mp (open
capillary) 181-183 °C; IR (KBr) 3400 (b), 2968, 2877, 1596, 1645,
1
1589, 1495, 1263, 1154 cm-1; H NMR (500 MHz, CDCl3) δ 10.07
(s, 1H), 7.22 (d, J ) 1.6 Hz, 1H), 7.13 (d, J ) 1.4 Hz, 1H), 1.52 (s,
9H); 13C NMR (500 MHz, CDCl3) δ 196.3, 153.1, 149.6, 138.4, 122.9,
115.4, 101.8, 34.4, 28.9; HRMS (FAB) calcd for C11H14O3Si [M]+
194.0943, found 194.0935.
(b) Dissymmetric salen ligand 3: To a solution of 3,5-di-tert-
butylsalicylaldehyde (2.25 g, 9.6 mmol) and 3-tert-butyl-2,5-dihy-
droxybenzaldehyde (0.62 g, 3.2 mmol) in CH2Cl2 (20 mL) was added
(R)-1,2-diaminocyclohexane (0.73 g, 6.4 mmol). The reaction mixture
was stirred at room temperature for 12 h, then concentrated to yield a
yellow foam. (This product was used without further purification in
the synthesis of 5 by resin-capture.) The mixture of salen ligands could
be separated by silica gel chromatography (gradient elution: diethyl
ether/hexanes, 1:20 to 1:1) to give 3 (1.15 g, 95% theoretical yield) as
a yellow foam. IR (film) 3316 (b), 2954, 2864, 1630, 1598, 1465, 1440
1
cm-1; H NMR (500 MHz, CDCl3) δ 8.35 (s, 1H), 8.12 (s, 1H), 7.42
(d, J ) 2.3, 1H), 7.05 (d, J ) 2.3, 1H), 6.87 (d, J ) 2.3, 1H), 6.44 (d,
J ) 2.3, 1H), 3.29 (m, 2H), 1.89 (m, 4H), 1.71 (m, 4H), 1.52 (s, 9H),
1.42 (s, 9H), 1.32 (s, 9H); 13C NMR (500 MHz, CDCl3) δ 166.0, 165.2,
158.9, 154.6, 147.0, 140.0, 138.6, 136.8, 127.3, 126.2, 118.4, 118.3,
117.8, 114.9, 72.1, 72.0, 35.1, 34.9, 34.1, 33.2, 33.0, 31.7, 31.5, 29.6,
29.4, 24.3; HRMS (FAB) calcd for [C32H46N2O3 + H]+ 507.3587, found
507.3583.
Synthesis of (R,R)-5. Hydroxymethyl polystyrene (Advanced
Chemtech, 2% cross-linked 90 µm beads, 0.8 mmol/g, 0.50 g, 0.4
mmol), 4-nitrophenyl chloroformate (322 mg, 1.6 mmol), and DMAP
(49 mg, 0.4 mmol) were combined, suspended in CH2Cl2 (5 mL), and
shaken at room temperature for 1 h. Filtration and rinsing with
anhydrous CH2Cl2 followed by drying in vacuo yielded colorless beads.
The IR spectrum (KBr pellet) revealed a strong absorbance at 1765
cm-1. To a suspension of this material in anhydrous DMF (5.0 mL)
was added (R,R)-3 (304 mg, 0.59 mmol), DMAP (49 mg, 0.4 mmol),
and DIPEA (0.14 mL, 0.8 mmol). The resulting yellow suspension was
shaken at room temperature for 1.5 h, then filtered and rinsed
sequentially with DMF, CH2Cl2, MeOH, and CH2Cl2 and dried in vacuo
to yield the product as yellow beads. The IR spectrum (KBr pellet)
contained strong absorbances at 1755 and 1630 cm-1. Preparation of
the polystyrene-bound salen by resin capture was performed exactly
as above except the mixture of isomers obtained in the preparation of
3 (outlined above) was used.
1
mg, 94% yield, 96.4% ee) in 90% purity by H NMR, the remainder
being 1,3-dibromo-2-propanol.
Reaction of Phenol and 1-Hexene Oxide Catalyzed by Polystyrene-
Bound Co(Salen) Complex 5. To a solution of phenol (11 mg, 0.11
mmol) in 1-hexene oxide (53 µL, 0.44 mmol) was added 5 (5.5 mg,
0.88 µmol) and F9-tert-butyl alcohol (3.1 µL, 22 µmol). The reaction
was gently stirred at room temperature for 4 h. An aliquot taken for
GC analysis indicated complete consumption of phenol and 13 to be
(26) Reaction conversion was calculated using the equation:21
Cobalt insertion into the polystyrene-bound ligand was accomplished
by adding a solution of Co(OAc)2‚4H2O (199 mg, 0.8 mmol) in 5 mL
1:1 MeOH/toluene to the resin beads (0.50 g) with gentle stirring at
room temperature. The beads turned dark red over a 2-min period. After
1 h the beads were filtered and rinsed sequentially with MeOH, CH2-
eesm/eeprod
conversion )
1 + (eesm/eeprod
)
krel was calculated using the equation:
ln[(1 - conversion) × (1 + eeprod)]
ln[(1 - conversion) × (1 - eeprod)]
krel
)
(25) Finney, N. S.; Pospisil, P. J.; Chang, S.; Palucki, M.; Konsler, R.
G.; Hansen, K. B.; Jacobsen, E. N. Angew. Chem., Int. Ed. Engl. 1997, 36,
1720.
(27) Feringa, B. L.; Lange, B. D. Tetrahedron 1988, 44, 7213-7222.

4154 J. Am. Chem. Soc., Vol. 121, No. 17, 1999
Annis and Jacobsen
present in >99% ee (Cyclodex B, 120 °C, tR ) 63.9, 66.0 min). The
reaction was filtered, the polymer beads were rinsed with CH2Cl2, and
the filtrates were concentrated in vacuo to yield 13 as a colorless oil
(22.3 mg, 98% yield, >99% ee) in 94% purity by 1H NMR, the
remainder being 1,2-hexanediol.
HKR of Styrene Oxide 21 with Silica-Bound Co(Salen) Complex
20. To complex 20 (15 mg, 1.4 µmol) was added tBME (50 µL), 21
(50 µL, 0.44 mmol), and H2O (5.5 µL, 0.31 mmol). The reaction was
gently stirred at room temperature for 3 h. An aliquot taken for GC
analysis indicated 50% conversion, with 21 present in 88.9% ee and
22 in 96.3% ee (Cyclodex B, 75 °C for 25 min, then 10°/min to 120
°C, hold 12 min, tR (21) ) 20.0, 21.2 min and tR (22) ) 34.0, 35.0
min as the acetonide prepared using 1% w/v camphorsulfonic acid in
dimethoxypropane). A krel of 51 was calculated from the reaction
conversion as measured by GC and from the ee of 21.28
HKR of 4-Hydroxy-1-butene Oxide 8 with Silica-Bound Co-
(Salen) Complex 20. To a solution of 8 (10 mg, 0.10 mmol) in 10 µL
of THF was added 20 (5.0 mg, 0.44 µmol) and H2O (2.0 µL, 0.10
mmol) and the reaction was gently stirred at room temperature for 2.5
h. An aliquot taken for GC analysis (as described for the HKR of 8
with catalyst 4) indicated 8 to be present in 48% ee and 9 in 93% ee,
corresponding to a conversion of 34% and krel of 44.
Continuous-Flow HKR of 4-Hydroxy-1-butene Oxide 8. A 2.0
mm i.d. × 25 mm (internal volume ) 79 µL) stainless steel HPLC
guard column equipped with 4.5 µm end frits was dry packed with ca.
30 mg (2.6 µmol) of 20. The catalyst bed was wetted with 5.0 mL of
5.9:1 THF/H2O at 35 µL/min, then 100 µL of a 5.1 M solution of 8 in
THF/H2O (corresponding to 45 mg, 0.51 mmol 8) was eluted through
the column with 350 µL of THF/H2O at 20 µL/h. The column was
then washed with another 350 µL of THF/H2O at 35 µL/min, and
collected eluate was analyzed by GC (as described for the HKR of 8
with catalyst 5) to indicate 8 to be present in 54% ee and 9 in 95% ee,
corresponding to a conversion of 36% and krel of 63. Removal of the
volatiles in vacuo yielded 18.6 mg (34%) of 9. The catalyst bed was
reoxidized by eluting a small volume of 200 µL of 9:1 toluene/HOAc
with anhydrous THF at 35 µL/min, then rinsing the column with 1.0
mL of THF/H2O at 35 µL/min prior to recycle.
Investigation of the Effect of Catalyst Site Isolation on the HKR
Reaction. Catalyst 20 was prepared at loading levels of 16, 27, 52,
and 87 µmol/g, as established by carbon analysis, by varying the ratio
of 15 to silica used in the synthesis. The probability that any catalyst
can interact with any other catalyst was determined according to the
method of Collman.29 The tether length r2/2 was calculated using the
molecular modeling program Chem3D30 to be 22.8 Å, with a base radius
r1/2 of 3.0 Å. The HKR of styrene oxide and epichlorohydrin were
conducted with these catalyst batches as described above. Aliquots were
periodically taken for GC analysis of reaction conversion and selectivity.
Rates of reaction were calculated from plots of ln[conversion] vs time.
Reaction of Phenol and Epibromohydrin Catalyzed by Polysty-
rene-Bound Co(Salen) Complex 5. To a solution of phenol (20 mg,
0.21 mmol) in epibromohydrin (73 µL, 0.85 mmol) was added 5 (13.3
mg, 2.1 µmol) and F9-tert-butyl alcohol (5.9 µL, 42 µmol). The reaction
was gently stirred at room temperature for 2 h. An aliquot taken for
GC analysis indicated complete consumption of phenol. HPLC analysis
indicated 14 to be present in 97.1% ee (Chiralcel OD, 9:1 hexanes/2-
propanol, tR ) 11.0, 19.6 min). The reaction was filtered, and the
polymer beads were rinsed with ether into a tared flask. The beads
were then rinsed sequentially with MeOH, and CH2Cl2 9:1 toluene/
HOAc, CH2Cl2, MeOH, CH2Cl2, and dried in vacuo prior to recycle.
After five recycles, powdered KOH (100 mg) was added to the
combined ether filtrates (ca. 50 mL). The reaction was stirred at room
temperature at which time GC analysis indicated complete conversion
of 14 to phenyl glycidyl ether 15. Filtration and concentration yielded
142 mg (90%) of 15 in >99% purity by 1H NMR and 96.6% ee
(Chiralcel OD, 9:1 hexanes/2-propanol, tR ) 8.0, 13.3 min).
Synthesis of (R,R)-19. Carboxylic acid 18: To a solution of
1-undecylenic acid (0.50 g, 2.7 mmol) in 1.5 of mL CH2Cl2 was added
N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA, 1.0 g, 5.0
mmol) with stirring at room temperature. The reaction gently warmed,
and was stirred 1 h then concentrated at the rotary evaporator to yield
a slightly yellow oil. The oil dissolved in 1.5 mL of CH2Cl2 and
ethoxydimethylsilane (0.75 mL, 5.4 mmol) and H2PtCl6 (ca. 10 mg in
50 µL of 2-propanol) were added with stirring. The reaction was heated
to reflux for 1 h then concentrated at the rotary evaporator to yield a
slightly yellow oil. To 400 mg (ca. 1.1 mmol) of this product was added
5.5 mL of EtOH and 0.39 mL of DIPEA. The solution was stirred at
room temperature for 1 h to selectively deprotect the trimethylsilyl ester
in the presence of the ethoxysilane, then was concentrated in vacuo to
yield 18 as a slightly yellow oil which was used without further
purification. To this was added 11 mL of CH2Cl2, DMAP (6 mg, 49
µmol), DIPEA (0.39 mL, 2.2 mmol), diisopropylcarbodiimide (DIC,
0.19 mL, 1.2 mmol), and (R,R)-3 (139 mg, 0.27 mmol). The reaction
was stirred at room temperature for 10 h, then diluted with 20 mL of
CH2Cl2 and extracted with aqueous NH4Cl. The organic phase was dried
over MgSO4, filtered, and concentrated to yield a yellow oil which
was purified by flash chromatography over silica gel (1:4 ethyl acetate/
hexanes) to yield 19 as a yellow film (118 mg, 55% based on 3). Rf )
0.50 (10% EtOAc/hexanes); 1H NMR (CDCl3) δ 8.30 (s, 1H), 8.23 (s,
1H), 7.32 (d, J ) 2.5 Hz, 1H), 6.98 (d, J ) 2.5 Hz, 1H), 6.93 (d, J )
2.5 Hz, 1H), 6.76 (d, J ) 2.5 Hz, 1H), 3.65 (q, J ) 6.5 Hz, 2H), 3.33
(m, 2H), 2.49 (m, 2H), 2.0-1.1 (m, 60 H), 0.10 (s, 6H). 13C NMR
(CDCl3) 171.2, 165.9, 164.6, 158.3, 158.0, 141.5, 140.0, 138.7, 136.5,
127.0, 126.0, 122.7, 121.3, 118.3, 117.8, 72.6, 72.2, 35.0, 34.9, 34.8,
34.1, 33.3, 33.2, 33.1, 31.5, 29.2-29.7, 24.3, 16.4, -2.0. Exact mass
(FAB, NBA + NaI) calcd for [C47H76N2O5Si + Na]+ 799.5422, found
799.5452.
Acknowledgment. This research was supported in part by
the NCI through the Institute of Chemistry and Cell Biology
(ICCB) at Harvard University by the NIH (GM-43214), and
by graduate fellowships to D.A.A. from Eli Lilly. The authors
also thank Dr. Stefan Peukert for valuable experimental
contributions.
JA984410N
(28) Values for krel were calculated using the equation:21
Synthesis of (R,R)-20. To a solution of 19 (74 mg, 95 µmol) in 1.5
mL of 2:1 CH2Cl2/DMF was added 320 mg of 10 µm spherical silica
(YMC Inc.) having a surface area of 350 m2/g. The slurry was stirred
and concentrated in vacuo to a yellow powder, which was then heated
at 100 °C at 1 Torr with gentle stirring for 20 h. The product was
filtered and rinsed sequentially with CH2Cl2, MeOH, and CH2Cl2 and
dried in vacuo to yield a bright yellow powder. Carbon analysis
indicated a loading of 4.67% C, corresponding to 87 µmol of ligand/g.
To 50 mg of the silica-bound ligand was added a solution of Co(OAc)2‚
4H2O (50 mg, 0.20 mmol) in 0.83 mL of 1:1 MeOH/toluene. The
reaction was stirred at room temperature for 0.5 h, filtered, rinsed
sequentially with MeOH, CH2Cl2, 9:1 toluene/HOAc, CH2Cl2, MeOH,
and CH2Cl2, and dried in vacuo to yield 20 as a green powder.
ln[(1 - conversion) × (1 + eesm)]
krel
)
ln[(1 - conversion) × (1 + eesm)]
(29) Assuming a random distribution of catalyst moieties on the silica
surface, the probability that any one can interact with any other is given by
N2X2 N3X3 N4X4
P ) NX -
+
-
+ ...
2!
3!
4!
2
where N is the total number of surface-bound species, and X ) π(r2
-
2
r1 )/A, where r1/2 is the van der Waals radius of the base of the molecule
at the surface, r2/2 is the length of the molecule, and A is the total area of
the surface.4a
(30) CambridgeSoft Corporation, 100 CambridgePark Drive, Cambridge,
MA 02140.