Modification of chiral dimethyl tartrate through transesterification: Immobilization on POSS and enantioselectivity reversal in sharpless asymmetric epoxidation

Article abstract of DOI:10.1002/chir.20814

Modification of dimethyl tartrate has been investigated through transesterification with aminoalcohols to provide reactive functionalities for the covalent bonding of chiral tartrate to polyhedral oligomeric silsesquioxanes. The transesterification of dimethyl tartrate has been widely studied using different catalytic systems and reaction conditions. Through the proper selection of both the catalytic system and the reaction conditions, it is possible to achieve monosubstituted or bis-substituted tartrate derivatives as sole products. All the intermediate chiral tartrate-derived ligands were successfully used in the homogeneous enantioselective epoxidation of allylic alcohols providing moderate enantiomeric excess over the products. Attached amine groups have been used to support the modified tartrate ligands on to a haloaryl-functionalized silsesquioxane moiety. This final chiral tartrate ligand displays reverse enantioselectivity in the asymmetric epoxidation of allylic alcohols with regard to the starting dimethyl tartrate ligand, both molecules having the same chiral sign. However, the POSS-containing ligand can be easily recovered in almost quantitative yield and reused in asymmetric epoxidation reactions. In addition, recovered silsesquioxane-pendant ligand, though displaying decreasing catalytic activity in recycling epoxidation tests, showed very stable enantioselective behavior.

Full text of DOI:10.1002/chir.20814

CHIRALITY 22:675–683 (2010)
Modification of Chiral Dimethyl Tartrate Through
Transesterification: Immobilization on POSS and
Enantioselectivity Reversal in Sharpless Asymmetric Epoxidation
1
1
1
2
RAFAEL A. GARCIA, RAFAEL VAN GRIEKEN, JOSE IGLESIAS, DAVID C. SHERRINGTON, AND COLIN L. GIBSON2
´
´
*
1Department of Chemical and Environmental Technology, Universidad Rey Juan Carlos, Madrid E28933, Spain
2Department of Pure and Applied Chemistry, Strathclyde University, Glasgow G1 1XL, United Kingdom
ABSTRACT
Modification of dimethyl tartrate has been investigated through trans-
esterification with aminoalcohols to provide reactive functionalities for the covalent
bonding of chiral tartrate to polyhedral oligomeric silsesquioxanes. The transesterifica-
tion of dimethyl tartrate has been widely studied using different catalytic systems and
reaction conditions. Through the proper selection of both the catalytic system and the
reaction conditions, it is possible to achieve monosubstituted or bis-substituted tartrate
derivatives as sole products. All the intermediate chiral tartrate-derived ligands were
successfully used in the homogeneous enantioselective epoxidation of allylic alcohols
providing moderate enantiomeric excess over the products. Attached amine groups
have been used to support the modified tartrate ligands on to a haloaryl-functionalized
silsesquioxane moiety. This final chiral tartrate ligand displays reverse enantioselectivity
in the asymmetric epoxidation of allylic alcohols with regard to the starting dimethyl tar-
trate ligand, both molecules having the same chiral sign. However, the POSS-containing
ligand can be easily recovered in almost quantitative yield and reused in asymmetric
epoxidation reactions. In addition, recovered silsesquioxane-pendant ligand, though dis-
playing decreasing catalytic activity in recycling epoxidation tests, showed very stable
C
VV
enantioselective behavior. Chirality 22:675–683, 2010.
2009 Wiley-Liss, Inc.
KEY WORDS: tartrate ligands; transesterification; asymmetric epoxidation;
enantioselectivity reversal; POSS
INTRODUCTION
tion of the catalytic species leads, in most of the cases, to a
great decrease in the catalytic activity because mass trans-
fer becomes a limiting stage of the reaction process. An
interesting way to prepare reusable catalysts without intro-
ducing mass transfer restrictions consists of using a car-
rier for the catalytic active species which helps the recov-
ering of the catalytic complex. One of the possibilities
deals with the handling of polyhedral oligomeric silses-
quioxanes (POSS) which easily dissolve or precipitate
depending on the organic solvent, allowing their recover-
ing. Silsesquioxanes are organosilicon compounds usually
employed as models for the study of the behaviour of cer-
tain species immobilized on to the surface of silica-based
supports. Those compounds have been applied to the
immobilization of both metal species10-12 or organic com-
pounds,13,14 giving an approximate idea of the catalytic
behaviour of the grafted species on to the surface of solid
supports. However, operating with these species simplifies
the working procedures since, through the proper choice
The development of the titanium-tartrate system as a
chiral catalyst for the asymmetric epoxidation of allylic
alcohols (the sharpless asymmetric epoxidation) has been
one of the major breakthrough in enantioselective synthe-
sis,1–3 as revealed by the large number of publications
related to this topic (the topic ‘Sharpless asymmetric epox-
idation’ gives 581 entries in a search performed on Sci-
Finder Scholar). Thus, the asymmetric epoxidation of al-
lylic alcohols has become a valuable tool for the synthesis
of enantiopure epoxy-alcohols, which are versatile chiral
building blocks.4,5 However, one of the main drawbacks of
this process is the complexity of the work-up procedure
for carrying out the oxidation and quenching the reaction.
Thus, not only are large amounts of solvents necessary to
purify the products but also the reuse or recovery of the
catalyst becomes impossible. Different attempts, more or
less successful, have been carried out to overcome these
problems. Some of these deal with the possibility of
supporting the chiral ligands on different solids, both
organic or inorganic, to obtain heterogeneous analogs
to the Sharpless homogeneous catalyst.6–9 In this way
the quenching of the reaction is largely simplified, since
the filtration of the catalysts is enough to stop the reaction
*Correspondence to: Rafael A. Garcia, Department of Chemical and Envi-
ronmental Technology, Universidad Rey Juan Carlos, C/ Tulipa´n s/n.
Mo´stoles, Madrid E28933, Spain. E-mail: rafael.garcia@urjc.es
Received for publication 11 May 2009; Accepted 21 October 2009
DOI: 10.1002/chir.20814
Published online 10 December 2009 in Wiley InterScience
and also the catalyst is reusable. However, the immobiliza- (www.interscience.wiley.com).
C
VV
2009 Wiley-Liss, Inc.

´
GARCIA ET AL.
676
in parts per million (ppm), in reference to the residual pro-
ton signals from the deuterated solvents. FTIR analysis
were acquired on a Mattson Infinity series FTIR spectrom-
eter using the KBr buffer technique. Elemental analyzes
were performed on a Elementar Vario EL III. Thin layer
chromatography (TLC) was carried out using precoated
sheets (Aldrich silica gel) and visualizing the products by
developing with phosphomolybdic acid/ethanol or ammo-
nium molybdate and cerric sulfate in H2SO4/H2O.15 Prod-
uct purification was carried out, unless otherwise stated,
on a semipreparative scale Varian Prepstar HPLC fitted
with a normal phase Dynamax Microsorb 100–8 Si column
(250 mm length, 41.4 mm I.D.) using n-hexane:i-propanol
mixtures as solvent.
Scheme 1. Enantioselective synthesis of mono 3 and di-amine 4
esters from dimethyl tartrate 1.
Protection of Starting Materials
of the solvent, silsesquioxanes can disolve to form homo-
geneous catalytic systems but they are easily recovered by
selective precipitation in other solvents, such as THF.
tert-Butyl 2-hydroxyethyl(methyl)carbamate (5). To
a solution of 2 (10 g, 0.133 mol) in THF (50 ml) at 08C,
The work described here presents the modification of was added a solution of di-tert-butyl dicarbonate (32 g,
chiral tartrate molecules, starting from dimethyl L-(1)-tar- 0.146 mol) in THF (10 ml) dropwise. The reaction mixture
trate 1, suitable for covalent bonding to properly function- was stirred overnight at room temperature and then con-
alized silsesquioxanes. This approach is based on the centrated in vacuo. The residue was purified by semipre-
modification of the ester groups, which minimizes the parative HPLC to give the title compound (21.5 g, 0.123
1
mol, 92%). H NMR (400 MHz, CDCl3): d 5 1.36 (s, 9H;
structural change with respect to catalyst enantioselectiv-
ity (Scheme 1). The enantiopure tartrate-derived ligands ꢀꢀC(CH3)3), 2.82 (s, 3H; ꢀꢀNꢀꢀCH3), 3.25 (m, 2H;
were used, together with titanium isopropoxide, in the ꢀꢀNꢀꢀCH2ꢀꢀ); 3.61 ppm (m, 2H; ꢀꢀCH2ꢀꢀOH). 13C NMR
(100 MHz, CDCl3):
d
5
28.2 (ꢀꢀC(CH3)3), 34.9
asymmetric epoxidation of various allylic alcohols, using
tert-butyl hydroperoxide (TBHP) as oxidant agent, provid- (ꢀꢀNꢀꢀCH3), 51.3 (ꢀꢀCH2ꢀꢀOH), 60.8 (ꢀꢀNꢀꢀCH2ꢀꢀ), 79.5
ing good asymmetric induction. Silsesquioxane-pendant (ꢀꢀC(CH3)3), 156.7 ppm (ꢀꢀNꢀꢀCO2ꢀꢀtBu). IR tmax (neat):
tartrate ligands showed reversal enantioselective with 1682 (C¼¼O), 2975 (CꢀꢀH), 3435 cm21 (OꢀꢀH). Elemental
regards nonmodified tartrate in the epoxidation of cin- analysis calcd (%) for C8H17NO3: C 54.84, H 9.78, N, 7.99;
namyl alcohol, being this result attributed to the bulky found: C 54.81, H 9.77, N 8.03.
size of the silicon substituent. Recycling tests with POSS-
Dimethyl (2R,3R,5R 0,6R 0) 5,6-dimethoxy-5,6-di-
functionalized tartrate ligands showed constant enantiose-
methyl-1,4-dioxane-2,3-dicarboxylate (6). To a solu-
lectivity though the activity decreases in reutilization runs.
tion of 1 (10.0 g, 55.6 mmol), trimethylorthoformate
(TMOF, 17.8 g, 166.9 mmol) and 2,3-butanodione (6.0 g,
67.6 mmol) in methanol, was added camphorsulfonic acid
(CSA, 1.3 g, 54.8 mmol). The mixture was then heated
under reflux and stirring was continued overnight. The
reaction was quenched by slow addition of NaHCO3 (10 g,
119 mmol) and reflux was maintained for two more hours.
The resultant suspension was filtered and concentrated to
dryness, giving a brown solid. The residue was then puri-
fied by recrystallization from n-hexane/ethyl acetate to
give the desired product as a white solid (14.3 g, 49,0
mmol, 88.1%). m.p. 105.28C. 1H NMR (400 MHz, CDCl3): d
5 1.34 (s, 6H; ꢀꢀCꢀꢀCH3), 3.30 (m, 6H; ꢀꢀOꢀꢀCH3), 3.75
(s, 6H; ꢀꢀCO2CH3), 4.51 ppm (m, 2H; ꢀꢀOꢀꢀCHꢀꢀ). 13C
NMR (100 MHz, CDCl3): d 5 17.4 (ꢀꢀCꢀꢀCH3), 48.5
(ꢀꢀOꢀꢀCH3), 52.5 (ꢀꢀCO2CH3), 68.7 (ꢀꢀOꢀꢀCHꢀꢀ), 99.0
(CꢀꢀCH3), 168.1 ppm (-CO2CH3). IR tmax (KBr): 1030,
1141, 1204, 1756 (C¼¼O), 2956 (CꢀꢀH) cm21. Elemental
analysis calcd (%) for C12H20O8: C 49.31, H 6.90; found: C
49.47, H 6.94.
EXPERIMENTAL
Materials and General Procedures
Dimethyl-L-tartrate (DMT, Acros, 199%) and N-methyl
ethanolamine (NMEA, Aldrich, 99%) were distilled under
an inert atmosphere before being used. N-butyl tin hydrox-
ide oxide (Aldrich, 97%) was used as received. Titanium
chloride was used and stored in dry box. TBHP anhidrous
solution in dichloromethane was prepared from acqueous
solution (TBHP, Aldrich, 70%) by extraction with dichloro-
methane followed by azeotropic distillation in a dean-stark
for solvents heavier than water. The obtained solution was
characterized by iodometric titration and stored at low
˚
temperature (148C) in presence of activated 3A molecular
sieves.
All nonaqueous reactions were carried out under an
inert atmosphere (usually nitrogen or argon) using stand-
ard Schlenk techniques, avoiding at all times the presence
of traces of water in the starting materials. Solvents were
distilled before use as follows: CHCl3 from P2O5; THF
Dimethyl (4R,5R ) 2-phenyl-1,3-dioxolane-4,5-
from Na/benzophenone; benzene and toluene from Na. dicarboxylate (7). A solution of 1 (10.0 g, 55.6 mmol)
Melting points were determined using a Mettler Toledo was mixed with benzaldehyde dimethyl acetal (9.4 g, 61.1
DSC822e. NMR spectra were recorded on a Varian Mer- mmol) in benzene (50 ml). To this mixture was added p-
cury 400 MHz spectrometer. Chemical shifts are reported toluenesulfonic acid (0.05 g, 0.3 mmol) in a 100 ml round
Chirality DOI 10.1002/chir

ENANATIOSELECTIVITY REVERSAL IN SHARPLESS ASYMMETRIC EPOXIDATION
677
bottom flask connected to a Dean-Stark apparatus for sol- 1749, 2959 cm21
. Elemental analysis calcd (%) for
vents lighter than water. The mixture was heated under C19H33NO10: C 52.40, H 7.64, N 3.22; found: C 51.96, H
reflux for 12 h during which time the solvent was with- 7.52, N 3.35.
drawn from the Dean-Stark trap in order to displace the
equilibrium. The reaction was allowed to cool and then
quenched with K2CO3 (4.2 g, 30.0 mmol). The resultant
suspension was filtered and concentrated in vacuo to give
a yellowish solid. The residue was recrystallized from
Bis{2-[(tert-butoxycarbonyl)(methyl)amino]ethyl}
(2R,3R)-5,6-dimethoxy-5,6-dimethyl-1,4-dioxane-
2,3-dicarboxylate (10). This compound was obtained
as a colorless oil (2.00 g, 3.47 mmol, 99%) starting from 6.
1H NMR (400 MHz, CDCl3): d 5 1.30 (m, 6H; ꢀꢀCꢀꢀCH3),
CH2Cl2/n-hexane to give compound 7 (13.1 g, 49 mmol,
1
89%). m.p. 73.28C. H NMR (400 MHz, CDCl3): d 5 3.84
1.42 (s, 18H; ꢀꢀC(CH3)3), 2.91 (s, 6H; CH3ꢀꢀNꢀꢀ), 3.43 (m,
(s, 3H; ꢀꢀCO2CH3), 3.89 (s, 3H; ꢀꢀCO2CH3), 4.92 (d, 2H, J
5 4.0Hz; ꢀꢀOꢀꢀCHꢀꢀ), 6.15 (s, 1H; ꢀꢀCHꢀꢀPh); 7.47 ppm
(5H; HAr). 13C NMR (100 MHz, CDCl3): d 5 52.9
(ꢀꢀCO2CH3), 76.7 (ꢀꢀOꢀꢀCHꢀꢀ), 106.6 (CHꢀꢀPh), 127.0,
128.2 and 129,9 (HCAr), 135.0 (CAr), 169.8 (ꢀꢀCO2CH3). IR
tmax (KBr): 1108, 1244, 1435, 1754 (C¼¼O), 2958 (CꢀꢀH)
cm21. Elemental analysis calcd (%) for C13H14O6: C 58.65,
H 5.30; found: C 58.40, H 5.37.
4H; ꢀꢀNꢀꢀCH2ꢀꢀ), 4.22 (m, 4H; ꢀꢀCH2ꢀꢀOꢀꢀ), 4.47 ppm
(2H; ꢀꢀCHꢀꢀOꢀꢀ). 13C NMR (100 MHz, CDCl3): d 5 17.7
(ꢀꢀCꢀꢀCH3), 28.6 (ꢀꢀC(CH3)3), 35.8 (ꢀꢀNꢀꢀCH3), 48.6
(ꢀꢀNꢀꢀCH2ꢀꢀ), 52.8 (ꢀꢀOꢀꢀCH3), 64.4 (ꢀꢀCH2ꢀꢀOꢀꢀ),
68.6 (ꢀꢀCHꢀꢀOꢀꢀ), 80.1 (ꢀꢀC(CH3)3), 99.4 (ꢀꢀCꢀꢀCH3),
155.8 (ꢀꢀNꢀꢀCO2tBu), 167.9 ppm (ꢀꢀCO2ꢀꢀCH2ꢀꢀ). Ele-
mental analysis calcd (%) for C26H46N2O12: C 53.97, H 8.01,
N 4.84; found: C 54.12, H 7.88, N 4.65.
Preparation of transesterification catalyst: (2-
Methyl-boc-amino)ethyl orthotitanate (8). To a solu-
tion of TiCl4 (0.10 g, 0.53 mmol) in CHCl3 (10 ml), was
added 5 (0.37 g, 2.1 mmol) dropwise to give a yellowish
solution. The resultant mixture was then stirred for 30 min
and then triethylamine (0.21 g, 2.1 mmol) was added via
syringe, giving a colorless suspension. The reaction was
then stirred for an additional hour before being used as
catalyst for transesterification reactions without further
purification.
4-{2-[(tert-Butoxycarbonyl)(methyl)amino]ethyl}-5-
methyl
(4R,5R)-2-phenyl-1,3-dioxolane-4,5-dicar-
boxylate (11). This compound was obtained as a color-
less oil (0.72 g, 1.76 mmol, 46.8%) starting from 7. 1H
NMR (400 MHz, CDCl3): d 5 1.41 (m, 9H; ꢀꢀC(CH3)3),
2.86 (s, 3H; CH3ꢀꢀNꢀꢀ), 3.28 (s, 2H; ꢀꢀNꢀꢀCH2ꢀꢀ), 3.88
(m, 3H; ꢀꢀCO2CH3), 4.24 (m, 2H; ꢀꢀCH2ꢀꢀOꢀꢀ), 4.86 (m,
2H; ꢀꢀCHꢀꢀOꢀꢀ), 6.07 (s, 1H; ꢀꢀCHꢀꢀPh), 7.42 ppm (m,
5H; HAr). 13C NMR (100 MHz, CDCl3): d 5 28.6
(ꢀꢀC(CH3)3), 35.3 (ꢀꢀNꢀꢀCH3), 47.4 (ꢀꢀNꢀꢀCH2ꢀꢀ), 52.8
(ꢀꢀCO2CH3), 63.7 (ꢀꢀCH2ꢀꢀOꢀꢀ), 77.7 (ꢀꢀCHꢀꢀOꢀꢀ), 80.1
Transesterification Products
Transesterification reactions were carried out by mixing (ꢀꢀC(CH3)3), 98.9 (ꢀꢀCHꢀꢀPh), 126.7, 127.6 and 129.2
acetal proctected DMT, 6 or 7 (3.5 mmol scale), and com- (HCAr), 135.2 (CAr), 155.2 (ꢀꢀNꢀꢀCO2tBu), 167.9 ppm
pound 5 (3.5 mmol for monosubstitution reactions, 7.7 (ꢀꢀCO2R). Elemental analysis calcd (%) for C20H27NO8: C
mmol for disubstitution reactions) in dry solvents (200 58.67, H 6.65, N 3.42; found: C 58.55, H 6.63, N 3.57.
ml), typically benzene. To the resultant mixtures were
Bis{2-[(tert-butoxycarbonyl)(methyl)amino]ethyl}
(4R,5R)-2-phenyl-1,3-dioxolane-4,5-dicarboxylate
(12). This compound was obtained as a colorless oil
added the transesterification catalysts: compound 8 (0.53
mmol) and butylstannonic acid (1.7 mmol) for the mono-
subtitution and disubtitution reactions respectively. The
so-prepared suspensions were then heated at reflux for 1
to 3 days, then filtered off through a column of Florisil to
remove the organometallic species and concentrated in
vacuo. The residues were purified by semi-preparative
scale HPLC and the fractions were collected for the main
products. In each case the resultant fractions were concen-
trated in vacuo to give the following products:
1
(1.91 g, 3.46 mmol, 92%) starting from 7. H NMR (400
MHz, CDCl3): d 5 1.43 (s, 18H; ꢀꢀC(CH3)3), 2.91 (s, 6H;
CH3ꢀꢀNꢀꢀ), 3.46 (m, 4H; ꢀꢀNꢀꢀCH2ꢀꢀ), 4.30 (m, 4H;
ꢀꢀCH2ꢀꢀOꢀꢀ), 4.92 (d, 2H; J517.9Hz, ꢀꢀCHꢀꢀOꢀꢀ), 6.09
(s, 1H; ꢀꢀCHꢀꢀPh), 7.44 ppm (m, 5H; HAr). 13C NMR
(100 MHz, CDCl3):
d
5
28.7 (ꢀꢀC(CH3)3), 35.8
(ꢀꢀNꢀꢀCH3), 47.8 (ꢀꢀNꢀꢀCH2ꢀꢀ), 64.4 (ꢀꢀCH2ꢀꢀOꢀꢀ),
77.7 (ꢀꢀCHꢀꢀOꢀꢀ), 80.3 (ꢀꢀC(CH3)3), 107.0 (ꢀꢀCHꢀꢀPh),
127.4, 128.5 and 130.1 (HCAr), 135.6 (CAr), 155.3
(ꢀꢀNꢀꢀCO2tBu), 169.3 ppm (ꢀꢀCO2CH3). Elemental analy-
sis calcd (%) for C27H40N2O10: C 58.68, H 7.30, N 5.07;
found: C 58.32, H 7.78, N 4.98.
2{2-[(tert-Butoxycarbonyl)(methyl)amino]ethyl}-3-
methyl (2R,3R) 5,6-dimethoxy 5,6-dimethyl-1,4-
dioxane-2,3-dicarboxylate (9). Compound obtained
as a colorless oil (1.43 g, 3.2 mmol, 93%) starting from 6.
1H NMR (400 MHz, CDCl3): d 5 1.33 (s, 6H; ꢀꢀCꢀꢀCH3),
1.44 (s, 9H; ꢀꢀC(CH3)3), 2.90 (s, 3H; CH3ꢀꢀNꢀꢀ), 3.30 (s,
Bis{2-[(benzyl)(methyl)amino]ethyl} (2R,3R)-2,3-
6H; ꢀꢀOꢀꢀCH3), 3.45 (m, 2H; ꢀꢀNꢀꢀCH2ꢀꢀ), 3.74 (s, 3H; dihydroxybutanedioate (18). This compound was
ꢀꢀCO2CH3), 4.24 (m, 2H; ꢀꢀCH2ꢀꢀOꢀꢀ), 4.50 ppm (m, 2H; obtained as a yellowish oil (2.06 g, 3.46 mmol, 92%)
ꢀꢀCHꢀꢀOꢀꢀ). 13C NMR (100 MHz, CDCl3): d 5 17.5 starting from 7 and subsequent deprotection with Pd/C
(ꢀꢀCꢀꢀCH3), 28.7 (ꢀꢀC(CH3)3), 35.5 (ꢀꢀNꢀꢀCH3), 47.3 in ethanol. The product was purified by flash chroma-
(ꢀꢀNꢀꢀCH2ꢀꢀ), 48.4 (ꢀꢀOꢀꢀCH3), 52.4 (ꢀꢀCO2CH3), 64.2 tography on silica with hexane:diethyl ether 50:50
(ꢀꢀCH2ꢀꢀOꢀꢀ), 68.7 (ꢀꢀCHꢀꢀOꢀꢀ), 80.0 (ꢀꢀC(CH3)3), vol. 13C NMR (100 MHz, CDCl3): d 5 39.5 (ꢀꢀNꢀꢀCH3),
99.2 (ꢀꢀCꢀꢀCH3), 154.9 (ꢀꢀNꢀꢀCO2tBu), 167.9 ppm 54.8 (ꢀꢀNꢀꢀCH2ꢀꢀ), 60.5 (ꢀꢀNꢀꢀCH2ꢀꢀPh), 62.1
(ꢀꢀCO2ꢀꢀR). IR tmax (neat): 1043, 1153, 1394, 1459, 1700, (ꢀꢀCH2ꢀꢀOꢀꢀ), 74.1 (ꢀꢀCHꢀꢀOꢀꢀ), 126.7, 128.1 snf 128.6
Chirality DOI 10.1002/chir

´
GARCIA ET AL.
678
(HCAr), 135.6 (CAr), 137.7 (ꢀꢀCAr), 169.0 ppm ꢀꢀCO2CH3), 4.28 (m, 2H; ꢀꢀCH2ꢀꢀOꢀꢀ), 4.53 ppm (m, 2H;
(ꢀꢀCO2CH2ꢀꢀ). Elemental analysis calcd (%) for ꢀꢀCHꢀꢀOH). 13C NMR (100 MHz, DMSO-D3): d 5 32.1
C24H32N2O6: C 64.85, H 7.26, N 6.30; found: C 64.98, H (ꢀꢀNꢀꢀCH3), 46.2 (ꢀꢀNꢀꢀCH2ꢀꢀ), 51.2 (ꢀꢀCO2CH3), 59.1
7.27, N 6.32.
(ꢀꢀCH2ꢀꢀOꢀꢀ), 71.5 (ꢀꢀCHꢀꢀOH), 170.7 ppm (ꢀꢀCO2R).
IR tmax (KBr): 1158, 1253, 1399, 1459, 1694, 1754, 2975,
3470 cm21. Elemental analysis calcd (%) for C8H15NO6: C
43.44, H 6.83, N 6.33; found: C 43.67, H 6.92, N 6.52.
Cleavage of Transesterification Products
After transesterification reactions and with the purpose
of using modified tartrates as chiral ligands in the asym-
metric epoxidation of cinnamyl alcohol, the protecting
groups were removed as follows:
(R,R) bis[2-(Methylamino)ethyl] tartrate (4). Start-
ing from 14 (0.634 g; 1.36 mmol) using deprotection method
ii) gave compound 4 (0.312, 1.18 mmol, 87%) as a solid from
CH2Cl2/EtOH. The same compound was produced starting
from 10 using procedure iii). m.p.5 133.738C. 1H NMR (400
MHz, DMSO-D3): d 5 2.61 (s, 6H; CH3ꢀꢀNꢀꢀ), 3.22 (m, 4H;
ꢀꢀNꢀꢀCH2ꢀꢀ), 4.31 (m, 4H; ꢀꢀCH2ꢀꢀOꢀꢀ), 4.63 ppm (d, 2H;
J 5 0.8Hz, ꢀꢀCHꢀꢀOH). dC ppm (DMSO-D6, 100 MHz): 32.4
(ꢀꢀNꢀꢀCH3), 46.3 (ꢀꢀNꢀꢀCH2ꢀꢀ), 59.7 (ꢀꢀCH2ꢀꢀOꢀꢀ), 72.0
(ꢀꢀCHꢀꢀOH), 170.4 ppm (ꢀꢀCO2CH3ꢀꢀ). Elemental analysis
calcd (%) for C10H20N2O6: C 45.45, H 7.63, N 10.60; found: C
45.31, H 7.70, N 10.67.
i) Benzyl acetal protected transesters were treated with
catalytic amounts of Pd/C (10%) in methanol under H2
atmosphere for at least 24 h. The reactions were carried
out until no substrate was detected by TLC and then fil-
tered and concentrated in vacuo.
ii) The 2,3-butanedione and BOC protecting groups were
removed by acid treatment with TFA in CH2Cl2. The
reactions were carried out in ultrasonic bath until com-
pletion and then concentrated in vacuo. The resultant
products were purified by crystallization from MeOH/
CH2Cl2 at 2308C.
Silsesquioxane Derived Molecules
(N-Methyl, methylphenylethyl-POSS)-aminoethanol,
15. A solution of chloromethyl)phenylethyl-POSS (1-[2-
[(chloromethyl)phenyl]ethyl]-3,5,7,9,11,13,15-heptacyclo-
pentylpentacyclo[9.5.1.13,9.15,15.17,13]octasiloxane) (1 g,
1 mmol) in dichloromethane was treated with 2 (1,5 eq)
and pyridine (0,2 eq). The resultant solution was heated to
reflux and stirred for 24 h. After the reaction was com-
plete, the solvent was removed under reduced pressure
and the crude reaction was suspended in a small quantity
of dichloromethane. The product was then recovered by
1-{2-[(tert-Butyoxycarbonyl)(methyl)amino]ethyl}
4-methyl (2R,3R)-2,3-dihydroxybutanedioate (13). Start-
ing from 11 (1.0g, 2.44 mmol) and using deprotection pro-
cedure i), the title product was obtained as a light yellow oil
(0.78 g, 2.43 mmol, 99%). 1H NMR (400 MHz, DMSO-D3): d
5 1.38 (s, 9H; ꢀꢀC(CH3)3), 2.80 (s, 3H; CH3ꢀꢀNꢀꢀ), 3.41
(m, 2H; ꢀꢀNꢀꢀCH2ꢀꢀ), 3.64 (s, 3H; ꢀꢀCO2CH3), 4.15 (2xt,
2H; J54.8Hz, J511.8Hz, J517.4Hz, ꢀꢀCH2ꢀꢀOꢀꢀ), 4.39 ppm
(m, 2H; ꢀꢀCHꢀꢀOH). 13C NMR (100 MHz, DMSO-D3): d 5
27.6 (ꢀꢀC(CH3)3); 33.8 (ꢀꢀNꢀꢀCH3); 46.7 (ꢀꢀNꢀꢀCH2ꢀꢀ);
51.2 (ꢀꢀCO2CH3); 62.0 (ꢀꢀCH2ꢀꢀOꢀꢀ); 72.1 (ꢀꢀCHꢀꢀOH);
78.3 (ꢀꢀC(CH3)3); 153.8 (ꢀꢀNꢀꢀCO2tBu); 170.7 ppm
(ꢀꢀCO2R). Elemental analysis calcd (%) for C13H23NO8: C
48.59, H 7.21, N 4.36; found: C 48.77, H 7.23, N 4.42.
1
precipitation in acetonitrile in nearly quantitative yield. H
NMR (400 MHz, CDCl3): d 5 0.93 (m; 2H; ꢀꢀSiꢀꢀCH2ꢀꢀ),
0.96 (m, 7H; ꢀꢀSiꢀꢀCH), 1.47, 1.57 and 1.73 (m, 56H;
ꢀꢀ(CH2)ꢀꢀ), 2.33 (s, 3H; CH3ꢀꢀNꢀꢀ), 2.65 (m; 2H;
ꢀꢀNꢀꢀCH2ꢀꢀ), 2.71 (m; 2H; ꢀꢀCH2ꢀꢀPhꢀꢀ), 3.80 (s, 3H;
ꢀꢀOꢀꢀCH3), 3.83 (m, 2H; ꢀꢀCH2ꢀꢀOH), 3.87 (m, 2H;
ArꢀꢀCH2ꢀꢀN-), 7.24 and 7.75 ppm (m, 8H; ꢀꢀCHArꢀꢀ). 13C
NMR (100 MHz, CDCl3): d 5 22.4, 27.2 (CCp), 41.1
(CH3ꢀꢀNꢀꢀ), 49.9 (SiꢀꢀCHꢀꢀ), 51.5 (ꢀꢀOꢀꢀCH3), 57.1
Bis{2[(tert-butyoxycarbonyl)(methyl)amino]ethyl}
(2R,3R)-2,3-dihydroxybutanedioate (14). Starting
from 12 (1.0 g, 1.80 mmol), the title compound was
obtained using deprotection method i) (0.82 g, 1.76 mmol,
(ꢀꢀCH2ꢀꢀOH),
57.8
(ꢀꢀPhꢀꢀCH2ꢀꢀN),
61.4
1
97.5%). H NMR (400 MHz, DMSO-D3): d 5 1.39 (s, 18H;
(ꢀꢀNꢀꢀCH2ꢀꢀ), 126.2, 129.1, 130.2 and 140.4 ppm (ꢀꢀCAr).
IR tmax (neat): 504, 1115, 1450, 2860, 2950, 3431 cm21. Ele-
mental analysis calcd (%) for C47H81NO13Si8: C 51.66, H
7.47, N 1.28; found: C 51.51, H 7.52, N 1.25.
ꢀꢀC(CH3)3), 2.84 (s, 6H; CH3ꢀꢀNꢀꢀ), 3.47 (m, 4H;
ꢀꢀNꢀꢀCH2ꢀꢀ), 4.25 (s, 4H; ꢀꢀCH2ꢀꢀOꢀꢀ), 4.50 ppm (s,
2H; ꢀꢀCHꢀꢀOH). 13C NMR (100 MHz, DMSO-D3): d 5
28.2 (ꢀꢀC(CH3)3), 35.1 (ꢀꢀNꢀꢀCH3), 47.4 (ꢀꢀNꢀꢀCH2ꢀꢀ),
62.9 (ꢀꢀCH2ꢀꢀOꢀꢀ), 72.0 (ꢀꢀCHꢀꢀOH), 79.5 (ꢀꢀC(CH3)3),
155.5 (ꢀꢀNꢀꢀCO2tBu), 170.4 ppm (ꢀꢀCO2CH3). IR tmax
(neat): 1203, 1433, 1463, 1679, 1759, 3030, 3345 cm21. Ele-
mental analysis calcd (%) for C20H36N2O10: C 51.71, H
7.81, N 6.03; found: C 51.85, H 7.74, N 5.97.
16. A solution of 15 (0.750 g, in benzene was treated
with 7 (1.0 eq) in the presence of butylstannonic acid (0,05
eq). The resultant suspension was then refluxed for
3 days, using a Dean-Stark apparatus to displace the equi-
librium, filtered off through a column of Florisil and the
clean solution concentrated in vacuo to give a yellow solid.
1-Methyl 4-[2-(methylamino)ethyl] (2R,3R)-2,3- The crude product was suspended in 5 ml of dichlorome-
dihydroxybutanedioate (3). Starting from 13 (0.583 g; thane and 50 ml of acetonitrile were added to precipitate
1.81 mmol), using deprotection method ii) gave compound the silsesquioxane products. The title compound was then
3 (0.280g, 1.27 mmol, 70%) as needle-shaped crystals after purified by flash chromatography on silica using n-hex-
crystallization from CH2Cl2/EtOH. The same compound ane:diethyl ether to give a white product (0,725 g, 0,53
was produced starting from 9 using procedure iii). m.p. 5 mmol, 76%). 1H NMR (400 MHz, CDCl3): d 5 0.91 (m, 2H;
123,328C. 1H NMR (400 MHz, DMSO-D3): d 5 2.60 (s, ꢀꢀSiꢀꢀCH2ꢀꢀ), 1.00 (m, 7H; ꢀꢀSiꢀꢀCH), 1.42, 1.50 and 1.72
3H; CH3ꢀꢀNꢀꢀ), 3.22 (s, 2H; ꢀꢀNꢀꢀCH2ꢀꢀ), 3.66 (s, 3H; (m, 56H; ꢀꢀ(CH2)ꢀꢀ), 2.41 (s, 3H; CH3ꢀꢀNꢀꢀ), 2.65 (m; 2H;
Chirality DOI 10.1002/chir

ENANATIOSELECTIVITY REVERSAL IN SHARPLESS ASYMMETRIC EPOXIDATION
679
ered and purified by semipreparative HPLC. The isolated
products were analyzed either by using a GC, fitted with
a chiral capillary column (Chiraldex G-TA; 40 m x
0.25 mm) and a FID detector, or by HPLC using n-hepta-
ne:isopropanol in a chiral column ((S,S)-Whelk-01; 25 cm
3 2.5 mm) fitted with a UV diode-array detector.
RESULTS AND DISCUSSION
Design and Synthesis of Mono (3) and
di-amino (4) Ester
Synthesis optimization. For anchoring purposes on
an appropriately functionalized silsesquioxane, N-methyl-
ethanolamine 2, was chosen as the amine functionality for
carrying out a transesterification reaction. To avoid unde-
sired reactions, the hydroxylamine 2 was BOC protected
to afford the carbamate 5 as a colorless oil in 95 % yield.
The starting material dimethyl L-tartrate 1, was first pro-
tected as the bisacetal by using 2,3-butanedione, accord-
ingly to the methodology developed by Dixon and
coworkers,16 yielding a white solid of (2,3,5,6)-dimethoxy-
5,6-dimethyl-1-dioxane-dimethyl tartrate 6 in 90–95% yield
after purification by recrystallization, or by using
dimethyl–benzyl acetal, following the Seebach proce-
dure,17 yielding a white solid of 2,3-O-benzylidene-di-
methyl tartrate 7, in 90–95% yield after recrystallization
(Scheme 2). In both cases, the chirality of the starting tar-
trate was maintained. The use of different protecting group
strategies was justified because of the different de-protec-
tion procedures to be employed. Thus, while the dime-
thoxy-1-dioxane protecting group in 6 may be removed by
acid hydrolysis with trifluoroacetic acid (TFA), the benzyli-
dine acetal in 7 may be removed by hydrogenolysis.
Scheme 2. Protection of hydroxyl groups from L-(1)-dimethyl tartrate
1 using either 2,3-butanedione or dimethyl-benzaldehyde as reagents.
ꢀꢀNꢀꢀCH2ꢀꢀ), 2.68 (m; 2H; ꢀꢀCH2ꢀꢀPhꢀꢀ), 3.68 (m, 2H;
ꢀꢀCH2ꢀꢀOꢀꢀ), 3.76 (s, 3H; ꢀꢀOꢀꢀCH3), 3.85 (m, 2H;
ArꢀꢀCH2ꢀꢀNꢀꢀ), 5.01 (m, 2H; ꢀꢀCHꢀꢀOꢀꢀ), 6.17 (s,
1H; ꢀꢀCHꢀꢀPh), 7.24, 7.43, 7.75 ppm (m, 18H; HꢀꢀAr). 13C
NMR (100 MHz, CDCl3): d 5 22.4 and 27.2 (CCp), 40.9
(CH3ꢀꢀNꢀꢀ), 49.9 (SiꢀꢀCHꢀꢀ), 51.2 (ꢀꢀOꢀꢀCH3), 58.4
(ꢀꢀPhꢀꢀCH2ꢀꢀN), 61.9 (ꢀꢀCH2ꢀꢀOꢀꢀ), 62.6 (ꢀꢀNꢀꢀCH2-),
80.1 (-CH-O-), 127.4, 128.5, 130.2, 135.5 and 142.1 (-CAr), 169.5
ppm (-CO2-).. IR tmax (neat): 501, 1112, 1450, 1746, 2868, 2950
cm21. Elemental analysis calcd (%) for C59H91NO18Si8: C
53.40, H 6.91, N 1.06; found: C 53.47, H 6.74, N 0.98.
17. The cleavage of the benzilidene acetal group in 16
(0.4 g, 0.36 mmol) was carried out using the above depro-
tection method i) giving 17 as a white solid after washing
1
with acetonitrile (0,417 g, 0.32 mmol, 89%). H NMR (400
MHz, CDCl3): d 5 0.93 (m, 2H; -Si-CH2-), 0.97 (m, 7H; -Si-
CH), 1.45, 1.51 and 1.73 (m, 56H; -(CH2)-), 2.39 (s, 3H;
CH3-N-), 2.67 (m; 2H; -N-CH2-), 2.71 (m; 2H; -CH2-Ph-),
3.74 (m, 2H; -CH2-O-), 3.79 (s, 3H; -O-CH3), 3.83 (m, 2H;
Ar-CH2-N-), 4.58 (m, 2H; -CH-OH), 7.21 and 7.43 ppm (m,
8H; H-Ar). 13C NMR (100 MHz, CDCl3): d 5 22.5 and
27.2 (CCp), 41.2 (CH3-N-), 50.0 (Si-CH-), 51.4 (-O-CH3), 58.0
(-Ph-CH2-N), 62.3 (-CH2-O-), 62.5 (-N-CH2-), 73.0 (-CH-OH),
127.4, 128.5, 130.5 and 141.9 (-CAr), 170.9 ppm (-CO2-). IR
tmax (neat): 501, 1112, 1746, 2868, 2950, 3411 cm21. Ele-
mental analysis calcd (%) for C52H87NO18Si8: C 50.41, H
7.08, N 1.13; found: C 50.29, H 7.04, N 1.19.
The next step in the tartrate modification protocol con-
sisted of the transesterification of the protected starting
materials 6 and 7. In this step both the mono (9 and 11)
and disubstituted (10 and 12) compounds can be obtained
(Scheme 3). For the synthesis of monosubstituted com-
pounds 9 and 11, the acetals 6 and 7 were reacted with
BOC ethanolamine 5 using a titanium alkoxide, following a
procedure similar to that described by Dixon et al.18 N-
Methyl-tert-butoxycarbonylethanolamine
5 was treated
General Procedure for the Asymmetric
Epoxidation of Allylic Alcohols
with titanium tetrachloride in the presence of triethylamine
as catalyst to give the corresponding titanium alkoxide, 8,
(Scheme 4) to be used as a transesterification catalyst. The
synthesis of 8 was targetted bearing in mind that transes-
terification reactions catalyzed by titanium alkoxides take
place by the alkoxide binding to the metallic center with
transfer to the ester group19 (Scheme 4).
For comparison purposes the chiral ligands prepared
accordingly to the above mentioned procedures have
been used in the asymmetric epoxidation of several al-
lylic alcohols in presence of titanium isopropoxide as the
metallic source and tert-butyl hydroperoxide as the oxi-
˚
dant. In a typical assay 1.0 g of 4A molecular sieves were
suspended under an inert atmosphere in 50 ml of dry
CH2Cl2 before cooling the resultant suspension down to
2208C. The next step consisted of the addition of 42 mg
of freshly distilled Ti(OiPr)4 (0.15 mmol), an equimolar
ammount of the chiral ligand (0.15 mmol) and 2.15 ml of
an anhydrous solution of 120 mmol TBHP in dry CH2Cl2.
The resultant suspension was then stirred for 1 h before
adding the substrate (30 mmol) dropwise over 1 h with a
syringe pump. The reaction was then stirred for another
Scheme 3. Transesterification of protected dimethyl tartrate 6 and 7
additional hour. The resultant epoxides were then recov- with 5 to give the mono and disubstituted compounds, 9–12.
Chirality DOI 10.1002/chir

´
GARCIA ET AL.
680
As the above methodology is clearly effective in achiev-
ing selective monosubstitution, an alternative procedure
was employed to synthesize the disubstituted products 10
and 12. A strongly acidic catalyst, n-butyl-stannonic acid,
has been described as an effective catalyst for the transes-
terification reaction of diesters.20 Bearing in mind the
effect of the solvent in the monosubstitution transesterifi-
cations, benzene was used as the reaction solvent. These
results are detailed in Table 2.
Increasing the molar ratio of 5:7 led to higher yields of
12 as the main product (Table 2, entries 1, 2 and 3). Addi-
tionally, further improvements in the yield of diesters 10
and 12 were obtained by varying the amount of catalyst.
The effect of this parameter was more pronounced than
altering the amine/DMT molar ratio. Here, the bis-substi-
tuted product 10 or 12 was the sole product when
the Ti:substrate ratio is raised to 0.5:1 (Table, 2, entries 5
and 6).
The next step in the procedure was the removal of the
protecting groups in 9–12, for which two different
approaches were used (Scheme 5). In the first route each
group was removed separately. Thus, the benzylidene ace-
tal protecting group in modified tartrates 11 and 12 was
first removed by hydrogenation, using the method
described by Kocienski.21 This procedure gave the corre-
sponding diols 13 and 14 in 95% yield after purification
by preparative HPLC. Subsequent cleavage of the BOC
protecting group was then carried out by acid hydrolysis
with CF3COOH. This gave the deprotected amine modi-
fied tartrates 3 and 4 in 70–85% yield after crystallization.
Alternatively, a method similar to that developed by Dixon
et al.18 was used to remove the protecting groups from tar-
trates 9 and 10. Thus, BOC and dimetoxy acetal deprotec-
tion was attempted using CF3COOH at rt. However, using
this procedure only the BOC protecting group was
removed. It was found that a second treatment increasing
the temperature and reaction time up to 24 h yielded the
unprotected tartrate 3 and 4.
Scheme 4. Preparation of titanium alkoxide 8 catalytic complex based
in protected aminoalcohol 5 and transesterification mechanism.
Table 1 sumarizes the results obtained for the transes-
terification of 6 and 7 with 5 using 8 as catalyst. The
investigated titanium alkoxide to tartrate ratios seem to
confirm the high catalytic activity of the titanium alkoxide
8, since better yields for the transesterification products
are obtained as the titanium content increases (Table 1,
entries 1–4). At this point it is noteworthy that the mono-
substituted compound 9 is produced as the unique prod-
uct with Ti/tartrate ratios of 0.05–0.15 (Table 1, entries 1–
3). The influence of the solvent polarity on the product dis-
tribution has also been investigated (Table 1, entries 5 and
6). Thus, similar results were found when THF or chloro-
form were used, while the use of a non-polar solvent such
as benzene led to a remarkable increase in the yield of the
monoester 9, without detecting any of the diester 10 (Ta-
ble 1, entry 6). These results suggest a strong influence of
the polarity of the solvent on the reaction outcome, as less
polar solvents provide higher yields of transesterified
product 9.
Comparison between the two differently protected start-
ing tartrates revealed that 7 was more reactive than 6 as
evidenced from the shorter transesterification reaction
times to achieve similar yields of 9 and 11 1 12, respec-
tively (Table 1, entries 6–7). Monitoring the reaction
media during the transtesterification reactions did not
reveal the presence of diester 10 when using 6 as starting
material, unlike the protected dimethyl tartrate 7, which
gave diester 12 after few hours. This difference in behavior
could be probably explained by the smaller number of
coordinating OR groups in 7, making the alkoxide more
available.
Synthesis of silsesquioxane-pendant tartrate
ligands. To evaluate the behavior of the tartrate-derived
compounds as chiral ligands, a step forward in terms of
the synthesis of silsesquioxane immobilized ligands
(Scheme 6) has been carried out. Initially the haloaryl-
functionalized silsesquioxane was reacted with 2 to give
TABLE 1. Transesterification results using catalyst 8
Entrya
Compound
Solvent
Catalyst loadb
Yield monoester (%)
Yield diester (%)
1
2
3
4
6
6
6
6
6
6
7
7
CHCl3
CHCl3
CHCl3
CHCl3
THF
Benzene
Benzene
Toluene
0.05
0.10
0.15
0.20
0.15
0.15
0.15
0.15
9 (16)
9 (14)
9 (26)
9 (59)
9 (25)
9 (93)
11 (47)
11 (34)
10 (–)
10 (–)
10 (–)
10 (15)
10 (–)
10 (–)
12 (27)
12 (11)
5
6
7c
8c
aReactions were performed for 72 h at the solvent boiling point temperature, using an equimolar amount of starting materials.
bTitanium alkoxide:substrate molar ratio.
cAssays carried out for 24 h.
Chirality DOI 10.1002/chir

ENANATIOSELECTIVITY REVERSAL IN SHARPLESS ASYMMETRIC EPOXIDATION
681
TABLE 2. Transesterification results using catalyst n-butyl stannonic acid
Entry
Compound
5 / DMT molar ratioa
Catalyst loadb
Yield monoester (%)
Yield diester (%)
1
2
3
4
5
6
7
7
7
7
7
6
1.0
1.5
2.2
2.2
2.2
2.2
0.10
0.10
0.10
0.20
0.50
0.50
11 (49)
11 (43)
11 (14)
11 (6)
11 (–)
9 (–)
12 (11)
12 (41)
12 (72)
12 (85)
12 (92)
10 (99)
aAmine:Tartrate molar ratio.
bTi:Substrate molar ratio.
product 15. The resultant product 15 displays, as did the suggest only moderate efficiency in the formation of the tar-
protected compound 5, the ability to react only with the trate–titanium complex, the essential active catalytic species
protected tartrate through the hydroxyl group. Thus, 15 for the sharpless epoxidation catalyst.
was reacted with the benzilidene acetal protected dimethyl
On the other hand, it is particularly noteworthy that the
tartrate 7 to give the transester 16. Protected dimethyltar- tartrate-derived chiral ligand supported on silsesquioxane
trate 7 was chosen as the starting material for this study 17 shows a similar activity to that of the Boc-protected tar-
because, unlike the butanedione derivative 6, which trate analog, although the most interesting result for this
requires strong acid treatment for its cleavage, the benzili- reaction lies in the reversed enantioselectivity showed by
dene acetal protecting group can be easily removed using this ligand. The measured enantiomeric excess is in the
mild hydrogenation conditions. In this way, the integrity of same range than using ligand 13, but the achieved chiral
the rest of the molecule was ensured by the cleavage of induction is completely reversed, yielding an excess of the
16 with Pd/C to give 17. All the steps of this sequence opposite enantiomer to the major one achieved with the
have led to similar results to that achieved for the analog rest of the L-(1)-dimethyl tartrate derivatives. Enantiore-
product 11, although butyl stannonic acid was used as cat- version has been previously observed with several ligands
alyst for the transesterification reaction because of its for different reasons.22–28 For instance, the molar ratio
readily availability and its capability to mainly produce between components22 or changing the reaction solvent25
symmetric bis-substituted tartrates through transesterifica- can cause the reversal of the enantioselectivity of certain
tion. The crude reaction product was then submitted to hy- catalysts. Immobilising on polymer supports has also been
drogenation in presence of Pd/C as catalyst for the cleav- described to produce reverse enantioselectivity.26,27
age of the benzylidene acetal, giving product 17 as a Actually, Janda and coworkers27 found enantioreversal
white solid.
induction in the epoxidation of 2-hexen-1-ol when using
poly(ethylene glycol) transesterified chiral tartrates. These
authors have shown that the Sharpless’ catalyst enantiose-
lection can be reversed depending on the size of the ester
substituents at the tartrate ligands, so that if the substitu-
ent is bigger than 750 a.m.u. the sign of the optical rota-
tion of the product is reversed. Bearing in mind the silses-
quioxane fragment attached to the tartrate ligand is larger
than 1,100 a.m.u., a similar enantioreversal behavior could
also arise with this ligand. With regard to the reusability
of the silsesquioxane-pendant tartrate chiral ligand, the
Catalytic Tests
The enantioselective epoxidation of different allylic alco-
hols with TBHP was performed at 2208C in presence of ei-
ther 3, 4, 13, 14 or 17 as chiral ligands and using Ti(O-
iPr)4 as titanium source. Ligands 13 and 14, containing
the BOC protecting groups, were studied for comparison
purposes, to contrast the enantio- and catalytic activity of
ligands 3 and 4, with free amino functionalities. These
tests also allowed checking the different modifications car-
ried out during consecutive protection, transesterification
and deprotection reactions did not cause ligand razemiza-
tion. The results have been summarized in table 3.
The best activity and enantioselectivity were found for
the BOC protected ligands 13 and 14 (Table 3, entries 3
and 4). Using ligand 14 as the chiral catalyst with titanium
tetraisopropoxide gave the epoxy alcohol in up to 70% ee. In
contrast, using tartrates with a free amino group i.e. 3 and
4 (entries 1 and 2) led to the generation of lower epoxide
yields and enantioselectivities, which could be caused by
the formation of different titanium-tartrate derivative com-
plexes when secondary amino groups, free from protective
groups and accordingly with labile hydrogen, are present
within the chiral ligand. In general, these results were
repeated for the different substrates tested, independent of
Scheme 5. Different routes for the cleavage of modified chiral tartrate
protecting groups. i) H2, Pd/C 10%, MeOH. ii) TFA/H2O, 5 min, r.t. iii)
the structure of the oxidized allylic alcohol. These results TFA, 408C, 24 h.
Chirality DOI 10.1002/chir

´
GARCIA ET AL.
682
Scheme 6. Immobilization of tartrate-derived ligands on halo-phenyl functionalized silsesquioxane materials. R = Ciclopentane. i) 2 (1.5 eq), Pyri-
dine (0.2 eq), reflux in dichloromethane for 24 h. ii) 7 (1 eq), BuSnOOH (0.05 eq), reflux in Benzene for 24 h iii) H2, Pd/C in AcEt, r.t. overnight.
same was recovered from the reaction media, after epoxi- other substrates, finding the same enantioselective rever-
dation of cinnamyl alcohol, by precipitation with THF, sion observed for cinnamyl alcohol.
washing with dichloromethane and drying before being
Finally, in order to determine whether this is the cause
used in a second assay (Table 3, entry 6). Results indicate of enantioreversion or just the presence of the benzyl
the enantioselectivity of the complex is preserved well dur- group in the aminoalcohol used for transesterifying the
ing the recycling test, leading to the same enantiomeric chiral tartrate ligand, a new compound was prepared. In
excess in the final glycidol product. On the other hand, this case, N-benzyl N-methyl amino ethanol was used for
the catalytic activity is greatly decreased for the reutiliza- the transesterification of dimethyl tartrate starting from 7
tion test, since less than half the initial epoxide yield is and carrying out the transesterification reaction in pres-
achieved. This loss of catalytic activity could be related to ence of butylstannonic acid (Scheme 7). The resultant
the inactivation of a fraction of the chiral complex during product, obtained after acetal cleavage, 18 was used as
the recycling test. Preliminary results on the epoxidation chiral ligand in the asymmetric epoxidation of cinnamyl
of different allylic alcohols indicate a similar behavior for alcohol inducing the usual chiral configuration on to the
TABLE 3. Asymmetric epoxidation of different allylic alcohols using tartrate-derived ligands
Entrya
Substrate
Ligand
Epoxide yield (%)b
Enantiomeric excess (%)c
1
2
3
4
5
6
7
8
9
3
4
49
64
56
72
51
27
66
57
61
17
28
50
70
240.
241.
65
47
57
13
14
17
17
18
13
14
d
e
10
11
13
14
25
37
44
57
12
13
13
14
29
48
45
60
14
15
13
14
9
17
48
52
aReactions were performed for 1 h 2208C, using TBHP as oxidant.
bIsolated Yield.
cEnantiomeric excess achieved over the (S,S) or (S) enantiomer determined by chiral HPLC or GC.
dReversal enantioselectivity is achieved for the silsesquioxane supported chiral ligand.
eRecycling test.
Chirality DOI 10.1002/chir

ENANATIOSELECTIVITY REVERSAL IN SHARPLESS ASYMMETRIC EPOXIDATION
683
10. Riollet V, Quadrelli EA, Cope´ret C, Basset JM, Andersen RA, Ko¨hler
K, Bo¨ttcher RM, Herdtweck E. Grafting of [Mn(CH2tBu)2(tmeda)] on
silica and comparison with its reaction with a silsesquioxane. Chem
Eur J 2005;11:7358–7365.
11. Mintcheva N, Tanabe M, Osakada K. Preparation and structure of
new phenylplatinum complexes containing silsesquioxane as a mono-
dentate or bidentate ligand. Organometallics 2006;25:3776–3783.
12. Pe´rez Y, Pe´rez Quintanilla D, Fajardo M, Sierra I, Del Hierro I. Immo-
bilization of titanium chiral alkoxides on SBA-15 and modelling the
active sites of heterogeneous catalyst using titanium silsesquioxane
complexes. J Mol Catal A 2007;271:227–237.
Scheme 7. Tartrate derivative incorporating aromatic rings analogue
to 17 without silsesquioxane pendant groups.
13. Kawasaki T, Ishikawa K, Sekibata H, Sato I, Soai K. Enantioselective
synthesis induced by chiral organic-inorganic hybrid silsesquioxane
in conjunction with asymmetric autocatalysis. Tetrahedron Lett 2004;
45:7939–7941.
resultant phenyl glycidol (Table 3, entry 7). In this way,
the enantioselectivity reversal observed for the POSS-func-
tionalized material should be ascribed to the size of the
silsesquioxane fragment more than to the presence of the
aromatic ring.
14. Ionescu G, Van Der Vlugt JI, Abbenhuis HCL, Vogt D. Synthesis and
applications of chiral phosphite ligands derived from incompletely
condensed silsesquioxane backbones. Tetrahedron Asymmetry 2005;
16:3970–3975.
In summary, straightforward strategy for modifying tar-
trates has been developed in order to attach the resultant
chiral ligands onto properly-functionalized polyhedral oli-
gomeric silsesquioxanes. The reaction conditions for
transesterification were optimized to achieve either the
monosubstituted or bis-substituted tartrate derivative as
the sole product. The chiral tartrate-derived ligands so
obtained were used in the asymmetric epoxidation of cin-
namyl alcohol achieving up to 70% ee in the resulting ep-
oxy-alcohol. C2-Symmetrical diesters 4 and 14 gave
higher enantiomeric excesses than the asymmetric mono-
substituted tartrate derivatives 3 and 13. Finally, the
amino groups anchoring the tartrate-derived ligands to a
silsesquioxane fragment result in a enantioselectivity re-
versal of the chiral ligand. Further studies on the applica-
tion of this heterogeneization strategy to the anchoring of
tartrate-derived ligands on the surface of silica supports
are being performed.
15. Gao Y, Hanson RM, Klunder JM, Ko Soo Y, Masamune H, Sharpless
KB. Catalytic asymmetric epoxidation and kinetic resolution: modified
procedures including in situ derivatization.
1987;109:5765–5780.
J Am Chem Soc
16. Barlow JS, Dixon DJ, Foster AC, Ley SV, Reynolds DJ. New building
blocks for efficient and highly diastereoselective polyol production –
synthesis and utility of (R0,R0,S,S) and (S0,S0,R,R)-2,3-butane diacetal
protected butane tetrol derivatives. J Chem Soc Perkin Trans 1
1999;12:1627–1630.
17. Seebach D. Modern synthetic methods. Frankfurt/Main: Wiley;
1980. p 113–171.
18. Dixon DJ, Foster AC, Ley SV, Reynolds DJ. Preparation of desym-
metrised meso-tartrate derivatives
– synthesis and utility of
(R0,R0,S,R)-2,3-butane diacetal procected butane tetrol. J Chem Soc
Perkin Trans 1 1999;12:1631–1633.
19. Uri A, Tuulmets A, Palm V. Transesterification of esters of carboxylic
acids by titanium-alcoholates 2. Transesterification of esters of aro-
matic acids. Org Rectivity (Tartu) 1983;20:122–131.
20. Furla´n RLE, Mata EG, Mascaretti OA. Butylstannonic acid catalyzed
transesterification of carboxylic esters. Tetrahedron Lett 1998;39:
2257–2260.
REFERENCES
21. Kociensky PJ. Protecting groups. Sttutgart: Thieme; 1994. p 96–
102.
1. Katsuki T, Sharpless KB. The first practical method for asymmetric
epoxidation. J Am Chem Soc 1980;102:5974–5976.
22. Shimizu I, Matsumoto Y, Shoji K, Ono T, Satake A, Yamamoto A.
Enantioselective elimination reaction of a 6,6-membered bicyclic al-
lylic carbonate. Importance of chirality reversal depending on the pal-
ladium-chiral phosphine ratio. Tetrahedron Lett 1996;37:7115–7118.
2. Martin VS, Woodard SS, Katsuki T, Yamada Y, Ikeda M, Sharpless
KB. Kinetic resolution of racemic allylic alcohols by enantioselective
epoxidation. A route to substances of absolute enantiomeric purity?
J Am Chem Soc 1981;103:6237–6240.
23. Altava B, Berruguete MI, Fraile JM, Garcia JI, Luis SV, Mayoral JA,
Vicent MJ. How important is the inner matrix of supported enantio-
meric catalysts? Reversal of topicity with two polystyrene backbones.
Angew Chem Int Ed 2000;39:1503–1506.
3. Hanson RM, Sharpless KB. Procedure for the catalytic asymmetric
epoxidation of allylic alcohols in the presence of molecular sieves.
J Org Chem 1986;51:1922–1925.
24. Trost BM, Fettes A, Shireman BT. Direct catalytic asymmetric aldol
additions of methyl ynones. Spontaenous reversal in the sense of
enantioinduction. J Am Chem Soc 2004;126:2660–2661.
4. Gao Y, Sharpless KB. Asymmetric synthesis of both enantiomers of
Tomoxetine and Fluoxetine. Selective reduction of 2,3-Epoxycinnamyl
alcohol with Red-Al. J Org Chem 1988;53:4081–4084.
25. Arseniyadis S, Valleix A, Wagner A, Miokowski C. Kinetic resolution
of amines: a highly enantioselective and chemoselective acetylating
agent with a unique solvent-induced reversal of stereoselectivity.
Angew Chem Int Ed 2004;43:3314–3317.
5. Katsuki T, Martin VS. Organic reactions. New York: Wiley; 1996.
p 1–299.
6. Canali L, Karjalainen JK, Sherrington DC, Hormi O. Efficient polymer-
supported sharpless alkene epoxidation catalyst. Chem Commun
1997;123–124.
26. Altava B, Berruguete MI, Garcia-Verdugo E, Luis SV, Vicent MJ.
Functional monolytic resins for the development of enantioselective
versatile catalytic minireactors with long-term stability: TADDOL sup-
ported systems. Green Chem 2006;8:717–726.
7. Meunier D, Piechaczyk A, De Mallmann A, Basset JM. Silica sup-
ported tantalum catalysts for asymmetric epoxidations of allylic alco-
hols. Angew Chem Int Ed 1999;38:3540–3542.
27. Reed NN, Dickerson TJ, Boldt GE, Janda KD. Enantioreversal in the
Sharpless asymmetric epoxidation reaction controlled by the molecu-
lar weight of a covalently appended achiral polymer. J Org Chem
2005;70:1728–1731.
8. Xiang S, Zhang Y, Xin Q, Li C. Asymmetric epoxidation of allylic alco-
hol on organic-inorganic hybrid chiral catalysts grafted onto the sur-
face of silica and in the mesopores of MCM-41. Angew Chem Int Ed
Engl 2002;41:821–824.
28. Lutz F, Igarashi T, Kawasaki T, Soai K. Small amounts of chiral b-
amino alcohols reverse the enantioselectivity of chiral catalysts in co-
operative asymmetric autocatalysis. J Am Chem Soc 2005;127:12206–
12207.
9. Zhao W, Hao ZP, Hu C, Li J, Xu X. The epoxidation of allylic alcohol
on Ti-complex/MCM-48 catalysts. Microporous Mesoporous Mater
2008;112:133–137.
Chirality DOI 10.1002/chir