Cooperative catalysis in the hydrolytic kinetic resolution of epoxides by chiral [(salen)Co(III)] complexes immobilized on gold colloids

Article abstract of DOI:10.1002/adsc.200800028

Chiral salen ligands were incorporated into self-assembled thiolate monolayers (SAMs) on gold colloids. Treatment of the immobilized ligand with Co(OAc)₂·4H₂O yielded the corresponding [(salen)Co(II)] complex, and aerobic oxidation in the presence of triflic acid afforded the catalytically active [(salen)Co(III)] complex. Functionalized gold colloids with a diameter of 3.4 nm, coated with a mixed monolayer of n-octanethiolates and thiolates with chiral [(salen)Co(III)] end groups were studied as catalysts in the hydrolytic kinetic resolution (HKR) of hexene-1-oxide. Extremely high selectivitiy and significant rate acceleration relative to homogeneous monomeric catalysts were observed. Recovery of the immobilized catalyst was accomplished by simple filtration, and catalyst reoxidation and repeated recycling (seven times) was possible with no loss of reactivity or enantioselectivity.

Full text of DOI:10.1002/adsc.200800028

COMMUNICATIONS
DOI: 10.1002/adsc.200800028
Cooperative Catalysis in the Hydrolytic Kinetic Resolution of
Epoxides by Chiral [(salen)Co
Gold Colloids
A
Thomas Belsera,b and Eric N. Jacobsena,*
a
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, MA 02138, USA
Fax : (+1)-617-496-1880; e-mail: jacobsen@chemistry.harvard.edu
Current address: Lonza Ltd., 3930 Visp, Switzerland
b
Received: January 15, 2008; Published online: March 18, 2008
Dedicated to Andreas Pfaltz on the occasion of his 60th birthday.
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Chiral salen ligands were incorporated
into self-assembled thiolate monolayers (SAMs) on
gold colloids. Treatment of the immobilized ligand
two catalyst molecules together in the rate-limiting
transition state and thereby enhancing catalytic effi-
ciency in the HKR. Successful approaches identified to
date include construction of covalently-linked dimers[6]
and oligomers,[7] self-assembly of as well as grafting
[(salen)metal] complexes onto dendrimers[8] or poly-
mers.[9] Enhanced reactivity relative to monomeric cat-
alysts has been demonstrated in each case, and highly
practical systems have been identified that allow use of
with Co(OAc)2·4H2O yielded the corresponding
A
[(salen)Co(II)] complex, and aerobic oxidation in
the presence of triflic acid afforded the catalytically
active [(salen)Co(III)] complex. Functionalized
T
gold colloids with a diameter of 3.4 nm, coated with
a mixed monolayer of n-octanethiolates and thio-
lates with chiral [(salen)Co
A
very low catalyst
loadings.[7c]
A
studied as catalysts in the hydrolytic kinetic resolu-
tion (HKR) of hexene-1-oxide. Extremely high se-
lectivitiy and significant rate acceleration relative to
homogeneous monomeric catalysts were observed.
Recovery of the immobilized catalyst was accom-
plished by simple filtration, and catalyst reoxidation
and repeated recycling (seven times) was possible
with no loss of reactivity or enantioselectivity.
Despite these important advances, the discovery of
HKR catalysts that display high activity and are easily
recovered and recycled remains an important and
challenging goal. In this context, we became interest-
ed in exploring the immobilization of [(salen)Co(III)]
A
on gold colloids. Alkanethiol-coated gold colloids in
the 2–10 nm size range are stable up to 1508C[10] and
maintain many of the practical properties of organic
molecules, being amenable to chromatographic purifi-
cation, dissolution, precipitation, sublimation, crystal-
lization, etc.[11] Characterization techniques that are
not normally suitable for the study of solids, such as
solution-phase NMR and MS, are applicable to gold
colloids. The synthesis of the gold colloids is straight-
forward, and catalyst immobilization takes place
Keywords: asymmetric catalysis; colloids; immobili-
zation; kinetic resolution; self-assembly
The hydrolytic kinetic resolution (HKR) catalyzed by under mild conditions. A single functionalized gold
[(salen)Co
ACHTREUNG
ful and widely used method for generating enantiopure its surface, and the thiolate chains self-assemble to
terminal epoxides as a direct result of its remarkable minimize steric interactions, leading to a uniform cat-
practical appeal,[2] high selectivity, and broad substrate alyst distribution on the colloid surface. Finally, the
scope.[3] Detailed mechanistic investigations of the catalyst loading may be varied, depending upon incu-
HKR and related epoxide ring-opening reactions using bation time. In this paper, we describe the successful
monomeric (salen)metal catalysts have revealed immobilization of salen ligands on gold surfaces. The
second-order kinetic dependence on catalyst concen- colloid-bound catalysts are shown to catalyze the
tration, and point to a cooperative mechanism of catal- highly enantioselective hydrolytic kinetic resolution
ysis (Figure 1).[4,5] Various strategies have been ex- with significant rate enhancements relative to mono-
plored for overcoming the entropic price of bringing meric [(salen)Co(III)] complexes.
A
Adv. Synth. Catal. 2008, 350, 967 – 971ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
967

Thomas Belser and Eric N. Jacobsen
COMMUNICATIONS
Figure 1. Cooperative bimetallic mechanism of the HKR reaction.
n-Octanethiolate-protected gold colloids were pre-
Incorporation of 3 onto the gold colloids was ac-
pared by reduction of gold salts with NaBH4 in tolu- complished via an exchange reaction[12,14] in dichloro-
ene in the presence of n-octanethiol, as described pre- methane. The resulting functionalized colloids were
viously.[12,13] Salen ligand 3, bearing a long alkyl chain stable and remained soluble unless precipitation was
and a terminal thiol functionality (Scheme 1), was induced by addition of hexanes. Reaction with
prepared from 11-mercaptoundecanoic acid by adap-
tation of standard protocols.
U
G
Scheme 1. Compounds 1–3.
968
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 967 – 971

Cooperative Catalysis in the Hydrolytic Kinetic Resolution of Epoxides
COMMUNICATIONS
Scheme 2. Synthesis of catalyst 4.
cally active [(salen)Co(III)] complex (Scheme 2). The plete kinetic resolution (>99.9% ee of recovered ep-
N
resulting colloid derivative 4 was collected by filtra- oxide) within 5 h (0.01mol% catalyst loading), where-
tion with a solvent-resistant stirred cell and purified as similar results were obtained only after 52 h using
by precipitation from ethanol with hexanes. Charac- complex 2 (Figure 2). Substantially higher loading of
terization was accomplished by transmission electron complexes 1 or 2 (>0.1mol%) was necessary to
microscopy (TEM), elemental analysis (EA) and effect complete resolution within 24 h. Both
1H NMR spectroscopy.
[(salen)Co(III)] complex 2 and the functionalized
A
The TEM studies revealed that the size of the gold gold colloids 4 induced the kinetic resolution of 1-
core did not change significantly during the place-ex- hexene oxide with the same high krel values (krel =
change reaction (before: Dd=3.46 nm, sD =1.16 nm; 180),[16] whereas [(salen)Co
ACHTREUNG
time of 36 h, the ratio of functionalized thiolates 3 to
Upon completion of the HKR reaction, the un-
n-octanethiolates was 3:1, as determined by EA and reacted epoxide was isolated by vacuum transfer. The
1H NMR spectroscopy. The molecular weight and thi- colloids could be readily recovered by filtration of the
olate surface coverage of the functionalized gold col- concentrated reaction mixture using centrifugal filter
loids 4 was estimated based on a structural model of units with a cut-off molecular size of 50 kD. Catalyst
the gold core shape. The model employed approxi- 4 reisolated in this manner was reused successfully.
mates the gold core as a sphere, covered by a skin of Monitoring of the reaction progress over time demon-
hexagonally close-packed gold atoms.[15] According to strated that the enantioselectivity remained constant
this model, colloids 4, with an average diameter of during seven recycling steps. After the sixth recycling
3.41Æ1.10 nm, contain 1221 gold atoms, 191 adsorbed step, the reactivity of the recycled colloids was dimin-
[(salen)Co(III)]-containing thiolates and 64 adsorbed ished to half of that observed in the first run. Howev-
G
n-octanethiolates on the surface. Elemental analysis er, reoxidation with TfOH and air resulted in material
of 4 was in agreement with the values obtained from with identical reactivity as the original batch. After
this model (EAtheor. : Au 56.4%, S 3.4%, Co 2.6%, N separation of the colloids, analysis of the product sol-
1.3%; EAmeas.: Au 56.7%, S 3.7%, Co 2.7%, N 1.1%). utions by 1H NMR spectroscopy revealed that no
The [(salen)Co
ACHTREUNG
functionalized gold colloids 4 were compared in the
HKR of racemic hexene-1-oxide (Figure 2). Function- the first example of cooperative reactivity in catalysts
alized colloids 4 were highly active, achieving com- immobilized on gold colloids. The practical advantag-
Adv. Synth. Catal. 2008, 350, 967 – 971ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
969

Thomas Belser and Eric N. Jacobsen
COMMUNICATIONS
Figure 2. HKR of (Æ)-hexene-1-oxide with complexes 1, 2 and functionalized gold colloids 4.
es of the colloid-bound [(salen)Co(III)] were demon-
strated, with high reactivity, selectivity, and facile cat-
alyst recovery and recycling.
A
72.4, 72.7, 118.1, 118.5, 121.8, 123.1, 126.3, 127.1, 136.4,
138.7, 140.3, 142.0, 158.1, 158.2, 165.1, 166.3, 172.7; IR
(NaCl): n=2930 (sb), 2858 (s), 1760 (s), 1631 (s), 1596 (m),
1440 (s), 1361 (m), 1315 (m), 1204 (m), 1151 (mb), 1096 cmÀ1
(m); MS (FAB): m/z (rel int%)=707 ([M]+, 100), 504 (33),
312 (28), 272 (41), 216 (30).
Experimental Section
Functionalized Gold Colloids 4
Thiol 3
n-Octanethiolate-protected gold colloids were prepared as
previously reported.[13] To a solution of n-octanethiol-pro-
tected gold colloids (1.70 g, ~1.3 mmol adsorbed n-octane-
thiol, 1equiv.) in freshly distilled CH 2Cl2 (400 mL) was
added thiol 3 (3.95 g, 5.60 mmol). The solution was stirred
(600 rpm) for 36 h at 238C, then concentrated to 50 mL
1-Ethyl-3-(3’-dimethyl-aminopropyl)carbodiimide
hydro-
chloride (0.60 g, 3.13 mmol) and N-hydroxybenzotriazole
(0.48 g, 3.13 mmol) were added to a solution of 11-mercap-
toundecanoic acid (0.60 g, 2.75 mmol) in freshly distilled
CH2Cl2 (40 mL) at 08C. After five minutes, {(1R,2R)-2-[(À)-
(3,5-di-tert-butyl-2-hydroxybenzylideneamino)-cyclohexylim-
under reduced pressure and treated with Co(OAc)2·4H2O
N
ino]methyl}-6-tert-butylbenzene-1,4-diol[6]
(1.00 g,
(5.60 g, 22.5 mmol) in methanol (50 mL). The solution was
stirred for one hour at 238C. The flask was opened to the
air and trifluoromethanesulfonic acid (2.00 mL, 22.9 mmol)
was added. The mixture was stirred vigorously open to air
for 2 h. The solvent was removed. The crude product was
dissolved in CH2Cl2 (50 mL) and filtered through Celite to
remove excess cobalt. The solution was concentrated to
20 mL under reduced pressure, diluted with ethanol/CH2Cl2/
acetone=3:1:1 (200 mL) and briefly sonicated to ensure
complete dissolution of by products. The colloids were col-
lected by filtration with a solvent-resistant stirred cell (Milli-
pore, 47 mm disc filter and an ultrafiltration membrane (re-
generated cellulose, PLTK, NMWL: 30000)) and washed
with ethanol/dichloromethane/acetone=3:1:1 (5 · 200 mL).
The black solution remaining in the stir cell was concentrat-
ed to 20 mL, suspended in hexanes (300 mL) and briefly so-
nicated to ensure complete dissolution of by-products. The
mixture was kept at À608C for 48 h to effect complete pre-
1.97 mmol) and triethylamine (0.60 mL, 4.31 mmol) were
added and the reaction mixture was stirred for 6 h at 238C.
The resulting solution was washed with a saturated NH4Cl
solution (50 mL) and water (50 mL). The organic phase was
dried over MgSO4, filtered and concentrated under reduced
pressure. Purification of the crude material by flash chroma-
tography, eluting with 10% diethyl ether in n-pentane, gave
thiol 3 as a foamy yellow glass; yield: 0.75 g (54%); Rf =
1
0.63 (n-pentane/diethyl ether=10:1). H NMR (400.1MHz,
CD2Cl2, 293 K): d=1.31 (s, 9H), 1.36 (m, 9H), 1.44 (s,
11H), 1.47 (s, 9H), 1.52 (m, 4H), 1.65 (quin, J=7.2 Hz,
2H), 1.76 (m, 4H), 1.92 (m, 2H), 2.10 (m, 2H), 2.54 (m,
4H), 3.40 (m, 2H), 6.83 (d, J=2.3 Hz, 1H), 6.99 (d, J=
2.3 Hz, 1H), 7.07 (d, J=2.3 Hz, 1H), 7.37 (d, J=2.3 Hz,
1H), 8.32 (s, 1H), 8.38 (s, 1H); 13C{1H}-NMR (100.6 MHz,
CD2Cl2, 293 K): d=24.5, 24.8, 25.2, 28.6, 29.2, 29.3, 29.5,
29.5, 29.7, 29.7, 31.4, 33.3, 33.3, 34.2, 34.4, 34.4, 35.1, 35.1,
970
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 967 – 971

Cooperative Catalysis in the Hydrolytic Kinetic Resolution of Epoxides
COMMUNICATIONS
cipitation. The colloids were collected on a glass filtration
frit, and washed twice with hexanes (200 mL). The black
material was dried under vacuum for 1hour to give the
functionalized gold nanoparticles 4; yield: 1.53 g.
[5] K. B. Hansen, J. L. Leighton, E. N. Jacobsen, J. Am.
Chem. Soc. 1996, 118, 10924–10925.
[6] R. G. Konsler, J. Karl, E. N. Jacobsen, J. Am. Chem.
Soc. 1998, 120, 10780–10781.
[7] a) J. M. Ready, E. N. Jacobsen, J. Am. Chem. Soc. 2001,
123, 2687–2688; b) J. M. Ready, E. N. Jacobsen, Angew.
Chem. 2002, 114, 1432–1435; Angew. Chem. Int. Ed.
2002, 41, 1374–1377; c) D. E. White, E. N. Jacobsen,
Tetrahedron: Asymmetry 2003, 14, 3633–3638; d) X.
Zheng, C. W. Jones, M. Weck, J. Am. Chem. Soc. 2007,
129, 1105–1112.
[8] R. Breinbauer, E. N. Jacobsen, Angew. Chem. 2000,
112, 3750–3753; Angew. Chem. Int. Ed. 2000, 39, 3604–
3607.
[9] a) D. A. Annis, E. N. Jacobsen, J. Am. Chem. Soc. 1999,
121, 4147–4154; b) X. Zheng, C. W. Jones, M. Weck,
Chem. Eur. J. 2006, 12, 576–583; c) S.-D. Choi, G.-J.
Kim, Catalysis Lett. 2004, 92, 35–40; d) B. M. Ross-
bach, K. Leopold, R. Weberskirch, Angew. Chem. 2006,
118, 1331–1335; Angew. Chem. Int. Ed. 2006, 45, 1309–
1312.
Catalyst Recycling
The reaction solution was concentrated under reduced pres-
sure to remove resolved epoxide. The residue was diluted
with acetone (8 mL) and transferred to a Millipore Centri-
plus Centrifugal Filter device with a cut-off molecular size
of 50 kD. After centrifugation (3000 rpm) at 238C for
40 min, the filtrate containing diol was removed and ana-
1
lyzed by H- and 19F NMR to ascertain the amount of cata-
lyst leaching. The remaining colloid solution was diluted
with 8 mL acetone and centrifugated again. This procedure
was repeated twice in order to remove any remaining prod-
uct.
The concentrated colloidal solution was transferred to a
flask, concentrated to dryness and washed with diethyl
ether/n-pentane=1:1. The colloids were then dried under
vacuum and reused as catalysts.
[10] M. Buettner, T. Belser, P. Oelhafen, J. Phys. Chem. B
2005, 109, 5464–5467.
Resolution of (Æ)-1,2-Epoxyhexane
Caution: HKR of terminal epoxides is an exothermicreac-
tion and care should be taken to ensure adequate heat dissi-
pation.
Distilled water (0.18 mL, 10.0 mmol) and catalyst were
added to (Æ)-1,2-epoxyhexane (2.00 mL, 16.6 mmol) in a
room temperature water bath. Every 20 min, an aliquot was
removed, dissolved in diethyl ether, filtered through a silica
plug and analyzed by GC. The enantiomeric excess of epox-
ide was determined by chiral GC analysis [g-TA, 408C, iso-
thermal, tR (minor)=4.3 min, tR (major)=4.5 min].
[11] M. Bartz, J. Küther, R. Seshadri, W. Tremel, Angew.
Chem. 1998, 110, 2646–2649; Angew. Chem. Int. Ed.
1998, 37, 2466–2468.
[12] a) M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R.
Whyman, J. Chem. Soc. Chem. Commun. 1994, 7, 801–
802; b) M. Brust, J. Fink, D. Bethell, D. J. Schiffrin, C.
Kiely, J. Chem. Soc. Chem. Commun. 1995, 16, 1655–
1656.
[13] T. Belser, M. Stoehr, A. Pfaltz, J. Am. Chem. Soc. 2005,
127, 8720–8731.
[14] a) M. J. Hostetler, S. J. Green, J. J. Stokes, R. W.
Murray, J. Am. Chem. Soc. 1996, 118, 4212–4213;
b) R. S. Ingram, M. J. Hostetler, R. W. Murray, J. Am.
Chem. Soc. 1997, 119, 9175–9178.
[15] a) G. D’Agostino, A. Pinto, S. Mobilio, Phys. Rev. B
1993, 48, 14447–14453; b) H. Sellers, A. Ulman, Y.
Shnidman, J. E. Eilers, J. Am. Chem. Soc. 1993, 115,
9389–9401; c) R. L. Whetten, W. M. Gelbart, J. Phys.
Chem. 1994, 98, 3544–3549; d) D. V. Leff, C. Ohara,
J. R. Heath, W. M. Gelbart, J. Phys. Chem. 1995, 99,
7036–7041.
Acknowledgements
This work was supported by the NIH through GM-43214
and postdoctoral fellowships from the “Swiss National Foun-
dation” and “Novartis Stiftung, vormals Ciba-Geigy-Jubi-
läums-Stiftung” to T.B.
[16] The selectivity factor s or krel in a kinetic resolution is
defined as the relative rate of the two reacting enantio-
mers of the substrate (s=krel =kfast/kslow). For a discus-
sion of fundamental and practical considerations in ki-
netic resolution reactions, see: J. M. Keith, J. F. Larrow,
E. N. Jacobsen, Adv. Synth. Catal. 2001, 343, 5–26.
[17] HKR reactions were carried out using 0.20 equiv. of
H2O relative to racemic epoxide. The krel values were
then determined based on the isolated yield and ee of
pure 1,2-diol, and therefore represent lower limits (see
ref.[3]).
References
[1] M. Tokunaga, J. F. Larrow, F. Kakiuchi, E. N. Jacobsen,
Science 1997, 277, 936–938.
[2] J. M. Keith, J. F. Larrow, E. N. Jacobsen, Adv. Synth.
Catal. 2001, 343, 5–26.
[3] S. E. Schaus, B. D. Brandes, J. F. Larrow, M. Tokunaga,
K. B. Hansen, A. E. Gould, M. E. Furrow, E. N. Jacob-
sen, J. Am. Chem. Soc. 2002, 124, 1307–1315.
[4] L. P. C. Nielsen, C. P. Stevenson, D. G. Blackmond,
E. N. Jacobsen, J. Am. Chem. Soc. 2004, 126, 1360–
1362.
[18] Higher selectivity factors have been obtained with
ligand 1 with other counterions [krel(1-acetate)=310];
see ref.[3]
Adv. Synth. Catal. 2008, 350, 967 – 971ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
971