Palladium-catalyzed stereoselective allylic alkylation of lithium enolates

Article abstract of DOI:10.1055/s-2005-921908

The lithium enolates, generated from cyclohexanone, cyclopentanone, and 1-tetralone, react with allyl acetate 1b or carbonate 1c enantioselectively, when catalyzed by (R)- or (S)-BINAP-derived palladium complexes. The presence of lithium chloride is cruci

Full text of DOI:10.1055/s-2005-921908

2
968
LETTER
Palladium-Catalyzed Stereoselective Allylic Alkylation of Lithium Enolates
S
M
tereoselective
A
llylic
a
A
lkylation
n
of LithiumE
f
nolatesred Braun,* Thorsten Meier
Institut für Organische Chemie und Makromolekulare Chemie, Universität Düsseldorf, 40225 Düsseldorf, Germany
Fax +49(211)8115079; E-mail: braunm@uni-duesseldorf.de
Received 30 August 2005
There remain, despite this progress, still substantial limi-
Abstract: The lithium enolates, generated from cyclohexanone, cy-
clopentanone, and 1-tetralone, react with allyl acetate 1b or carbon-
ate 1c enantioselectively, when catalyzed by (R)- or (S)-BINAP-
derived palladium complexes. The presence of lithium chloride is
tations inasmuch as only such enolate precursors, which
contain just one acidic proton, have been used in order to
avoid not only double or multifold alkylation, but also a
crucial to stereoselectivity. Diastereoselective and enantioselective later racemization caused by deprotonation at the newly
allylation occurs between cyclohexanone and carbonate 1d. It is created stereogenic carbon center in the presence of a
shown in the case of the acyclic substrates (Z)-12 and (E)-17 that p-
strong base. That type of ‘structural support’ vanishes in
allyl palladium complexes are attacked by lithium enolates from the
the asymmetric allylic alkylation of unrestricted ketones,
face opposite to the noble metal.
so that the most simple-looking substrates become indeed
Key words: asymmetric synthesis, stereoselectivity, ketones,
nucleophilic additions, allyl complexes
5c
the most difficult ones. Having recently shown that dia-
stereoselective and enantioselective reactions of ketones
with the substrate 1a are feasible, we report in this
communication for the first time the enantioselective
palladium-catalyzed reaction of cyclohexanone and cy-
clopentanone with the unsubstituted allyl acetate 1b and
carbonate 1c as well as the extension of this protocol to the
dimethyl-substituted homologue 1d.
The palladium-catalyzed allylic alkylation of carbon nu-
cleophiles has developed into a versatile, frequently ap-
plied method for the formation of carbon–carbon bonds.1
Enantioselective variants converting racemic starting ma-
terials 1 into non-racemic products 3 have been developed
due to chiral ligands at the transition metal, so that the
approach of the nucleophile is directed predominantly to
one of the diastereotopic termini of the allyl complex 2. A
major limitation is that the carbon nucleophiles are almost
exclusively ‘soft’, stabilized carbanions. Among them,
the large majority are symmetrically substituted nucleo-
philes, typically malonates. As a consequence, the allylic
substitution leads to the formation of just one stereogenic
O
R
*
*
R
4
OLi
H
R
R
+
2
center in the allylic position.
PdLn
5
2
R
R
R
R
R
R
PdLn
–X
Nu
PdLn
–
X
PdLn
Nu
O
R
R
H
1
2
3
H
X
Scheme 1
1a: X = OCOCH3, R = Ph
1b: X = OCOCH3, R = H
1c: X = OCO2CH3, R = H
1d: X = OCO2CH3, R = CH3
6
If, however, synthesis targets the formation of olefinic
ketones 4 with stereogenic centers in the homoallylic
position (R = H) or in both, the allylic and the homoallylic
Scheme 2
position (R ≠ H), the appropriate prochiral nucleophiles
3
are preformed non-stabilized enolates 5, as outlined in Cyclohexanone was chosen as the substrate to optimize
retrosynthetic Scheme 2. Although the chemistry of palla- the reaction conditions (Scheme 3, Table 1). The ketone
dium-catalyzed allylic substitution and that of preformed was always deprotonated by means of lithium diisoprop-
enolates evolved at the same time, there have been only ylamide, and the palladium dibenzylidene acetone com-
4
reluctant attempts to combine both concepts. In recent plex [Pd (dba) ]·CHCl was used as the precursor of the
2
3
3
years, a few enantioselective procedures have been elabo- catalyst. Among various chiral ligands tested, phosphora-
5
6
7
8
rated based on enolates of ketones and a-amino esters.
midite 7 (entry 1) as well as BINAP (8, entries 3, 4, and
6
9
–8) and tolyl-BINAP (9, entry 5) provided substantial
SYNLETT 2005, No. 19, pp 2968–2972
Advanced online publication: 04.11.2005
DOI: 10.1055/s-2005-921908; Art ID: G24905ST
0
1
.1
2
.2
0
0
5
enantioselectivity. It turned out that the reaction was
strongly influenced by the addition of metal salts. The
enhancement of enantioselectivity (entries 4 and 5) at the
©
Georg Thieme Verlag Stuttgart · New York

LETTER
Stereoselective Allylic Alkylation of Lithium Enolates
2969
Table 1 Enantioselective Allylic Alkylation of Cyclohexanonea
Entry
Substrate
1b
Ligand/additive
Temp (°C)
–30
Product
Yield (%)b
ee (%)
1
2
3
4
5
6
7
8
9
7c/–
(R)-4a
20
50
7c/ZnBr
2
d
e
1b
–30
f
1b
(S)-8 /–
–30
(S)-4a
(S)-4a
(S)-4a
(S)-4a
(R)-4a
(S)-4a
40
15
17
53
63
32
70
70
75
90
1b
(S)-8/ZnBr2
(S)-9/ZnBr2
(S)-8/LiCl
(R)-8/LiCl
–20
1b
–20
1b
–78
1c
–78
g
1c
(S)-8 /LiCl
–78
87
80
d
e
1c
(R)-8/–
(S)-9/–
–78
d
e
10
1c
–78
a
b
c
d
e
f
Standard catalyst loading: 0.5 mol% [Pd (dba) ]·CHCl , 2 mol% ligand.
Isolated distilled products.
2
3
3
2
.5 mol% [Pd (dba) ]·CHCl , 20 mol% ligand.
2 3 3
Traces.
Not determined.
Catalyst loading: 2.5 mol% [Pd (dba) ]·CHCl , 10 mol% ligand.
2
3
3
g
Catalyst loading: 0.025 mol% [Pd (dba) ]·CHCl , 0.1 mol% ligand.
2
3
3
expense of chemical yields (entries 2, 4, and 5) by the It turned out that by this protocol double allylation and
additive zinc bromide could be explained as an effect of aldol condensations, both undesired side reactions, were
transmetallation of the lithium enolate. This hypothesis is, widely suppressed. In addition, racemization of the allyla-
however, challenged by the effect of the additive lithium tion product 4a was avoided. Under the optimized condi-
chloride (entries 6–8). This modification of the reaction tions (entry 7) ketone (R)-4a was obtained in 90% ee and
conditions leads to the highest yields and enantioselectiv- 63% isolated yield. Remarkably, the amount of palladium
ities. One could conclude that the disaggregation of the could be as low as 0.05 mol% and 0.1 mol% of the ligand,
1
0
lithium enolate, a well-known effect of the addition of as shown by the reaction that was mediated by (S)-BINAP
lithium chloride, is responsible for both enhanced reactiv- (8) and delivered (S)-4a in 80% ee and 87% isolated yield
ity and enantioselectivity. In addition, the lithium salt (entry 8). The optimized conditions thus elaborated were
might also influence the aggregation of the transition met- then applied to the enantioselective allylic alkylation of
1
1
al catalyst. An additional improvement comes from a cyclopentanone (6b) and a-tetralone (6c) to give the ole-
change of the leaving group in the allylic substrate (entries finic ketones 4b and 4c (Scheme 4). So far, most proce-
7
and 8). Remarkably, no allylation occurs in the absence dures for enantioselective allylic alkylations of ketones
of lithium chloride at –78 °C (entries 9 and 10).
rely on the use of azaenolates with chiral auxiliary
groups.12 Amongst the few catalytic variants, the chelat-
ing of the lithium cation in enolates by chiral ligands
O
1. LiN(i-Pr)2, THF
O
O
2
. [Pd2(dba)3]·CHCl3
13
seems to be restricted to special substrates. The very
or
3. ligand/additive
recently reported asymmetric allylic alkylation through
4. 1b or 1c
14a
14b,c
allyl b-ketoesters or enol carbonates
might offer a
6a
(R)-4a
(S)-4a
promising alternative.
Having demonstrated that enantioselective allylic substi-
tutions are feasible with simple ketone enolates, the un-
substituted allyl acetate 1b or carbonate 1c was replaced
by a 1,3-disubstituted substrate. In asymmetric allylic
alkylations, the diphenyl-substituted acetate 1a plays the
role of a ‘standard electrophile’. As a consequence, many
protocols are restricted to the 1,3-diphenyl substitution
pattern and fail or are at least less efficient when applied
to other substrates.
Ph
O
PAr2
PAr2
P
N
O
H
(
(
R)-8: Ar = Ph
R)-9: Ar = p-Tol
7
Scheme 3
Thus, the protocol for the enantioselective allylation of
lithium enolates was applied to the carbonate 1d that fea-
Synlett 2005, No. 19, 2968–2972 © Thieme Stuttgart · New York

2
970
M. Braun, T. Meier
LETTER
1
. LiN(i-Pr)2, THF
As ketones display pK values around 25, they are border-
a
O
O
line cases,18 so that the stereochemistry in their reaction
with p-allyl palladium complexes is hardly predictable a
priori: for cyclic, in particular cyclohexenyl-derived allyl-
ic substrates, the lithium and potassium enolates of ace-
tone have been shown to attack the allyl palladium
2. [Pd2(dba)3]·CHCl3 (0.5 mol%)
3
4
. LiCl; (S)-8 (2 mol%)
. 1c
76%
6b
(S)-4b (64% ee)
1. LiN(i-Pr)2, THF
O
O
2. [Pd2(dba)3]·CHCl3 (0.5 mol%)
4a,b
complex under inversion.
3
4
. LiCl; (R)-8 (2 mol%)
. 1c
85%
Ph
OLi
LiCl
Pd2(dba)3]·CHCl3 (2.5 mol%)
6c
(R)-4c (66% ee)
+
AcO
MEMO
Z)-12
[
Ph
Scheme 4
14 (10 mol%)
(
13
quant.
OMEM
tures the 1,3-dimethyl substitution, as shown in Scheme 5.
The lithium enolate derived from cyclohexanone smooth-
ly reacts with the carbonate 1d at –78 °C in the presence
of lithium chloride and the palladium catalyst formed with
Ph
Ph
O
O3, Me2S
7%
*
O
O
4
Ph
(R)-16
Ph
(
R)-8. The diastereomer 10 formed predominantly and the
15
OMEM
ratio of 10:11 amounts to 97:3. The major diastereomer
was obtained in 96% ee.15
OLi
LiCl
Pd2(dba)3]·CHCl3 (2.5 mol%)
Ph
AcO
+
[
Ph
O
1. LiN(i-Pr)2, THF
1
4 (10 mol%)
2. [Pd2(dba)3]·CHCl3 (0.5 mol%)
(E)-17
13
quant.
3
4
. LiCl; (R)-8 (2 mol%)
. 1d
OMEM
Ph
35%
Ph
O
6a
O3; Me2S
1%
*
O
O
CH3
O
CH3
O
6
Ph
(S)-16
CH3
+
CH3
Ph
18
PPh2
PPh2
1
9
0 (96% ee)
7
11
3
:
14:
Fe
Scheme 5
Scheme 6
The absolute configurations of the allylation products
a–c were assigned based on the comparison of the sign
of their optical rotations with those described in the liter-
4
As the aggregation of enolates might influence the stereo-
chemical outcome, the allylation protocol described
above was applied to the reaction of the lithium enolate of
acetophenone with the acyclic, unrestricted allyl sub-
strates (Z)-12 and (E)-17, which are identical in the con-
figuration of their stereogenic carbon atoms, but differ
1
3,16,17
ature.
In case of the dimethyl-substituted allylic
product 10, the absolute configuration at the a-carbonyl
center could be elucidated by comparison of the circular
dichroism with that of the known compound 4a. The
enantiomeric excesses were determined, aside from com-
paring the optical rotations, by chiral GC for the products
19
just in the configuration of the double bond. They also
contain a stereogenic center that should not be touched in
the allylic substitution, so that it could serve as a probe in
order to evaluate the stereochemical result.
4
a, 4b and 10 and by HPLC on a chiracel OD-H column
(
compound 4c).
Two different pathways have been discussed in the reac-
tion of a nucleophile with p-allyl palladium complexes: It
either attacks the p-allyl system opposite to the transition
metal, as indicated in Scheme 1, or it is first attached to
the noble metal and approaches the p-system from the
same face. According to multifold evidence, ‘soft’ carbon
nucleophiles follow the first type of stereochemical out-
come whereas ‘hard’ carbon nucleophiles react via a pre-
Thus, both substrates 12 and 17 were treated with the eno-
late 13 and lithium chloride and the reactions were medi-
ated by [Pd (dba) ]·CHCl and the achiral ligand 14 (dppf)
to yield the ketones 15 and 18, respectively (Scheme 6).
The products were obtained in a regioselective and diaste-
reoselective manner, and the isomer ratio surpassed 95:5
in both cases.
2
3
3
1
,2
At a glance, it seems surprising that the carbon–carbon
bond formation in 15 has occurred under inversion at the
carbon atom, marked with an asterisk. This result is, how-
coordination according to the second pathway. A border
between the two types of carbon nucleophiles is set in that
way, that conjugate bases of carbon acids with pK < 25
a
2
0
ever, explained by assuming a p-s-p interconversion,
are termed to be ‘soft’ nucleophiles, whereas ‘hard’ nu-
2
c
shown in Scheme 7, that is clearly indicated by the con-
cleophiles are derived from carbon acids with pK > 25.
a
version of the Z-double bond in 12 into E-configured
Synlett 2005, No. 19, 2968–2972 © Thieme Stuttgart · New York

LETTER
Stereoselective Allylic Alkylation of Lithium Enolates
2971
product 15. Thus, the p-complex 19 forms initially from Acknowledgment
the substrate (Z)-12. It is converted, via the s-intermediate
This work was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie. We would like to thank
Prof. Dr. J. Pietruszka, Forschungszentrum Jülich, for chiral HPLC
2
0 and its rotamer 21, into the p-complex 22, which is
attacked by the enolate 13 from the face opposite to the
transition metal. The overall stereochemical outcome re- measurements. Chiral GC measurements were performed by Bayer
sults from two substitutions, each of them occurring under Industry Services. This work is part of the dissertation of T. Meier,
University of Düsseldorf, 2005.
inversion, and a p-s-p process that converts the Z- into
the E-configured double bond.
References
Ph
Ph
PdLn
(
1) (a) Trost, B. M. Tetrahedron 1977, 33, 2615. (b) Tsuji, J.
Organic Synthesis with Palladium Compounds; Springer:
New York, 1980. (c) Trost, B. M. Acc. Chem. Res. 1980, 13,
PdLn
AcO
p
s
(Z)-12
LnPd
–
MEMO
MEMO
20
385. (d) Trost, B. M. Pure Appl. Chem. 1981, 53, 2357.
19
(
e) Tsuji, J. Pure Appl. Chem. 1982, 54, 197. (f) Godleski,
A. In Comprehensive Organic Synthesis, Vol. 4; Trost, B.
M., Ed.; Pergamon: Oxford, 1991, 585.
2) For reviews, see: (a) Reiser, O. Angew. Chem., Int. Ed. Engl.
rotation
OMEM
(
OMEM
1993, 32, 547; Angew. Chem. 1993, 105, 576. (b)Williams,
p
s
13
Ph
Ph
J. M. J. Synlett 1996, 705. (c) Trost, B. M.; Van Vranken, D.
L. Chem. Rev. 1996, 96, 395. (d) Helmchen, G. J.
Organomet. Chem. 1999, 576, 203. (e) Helmchen, G.;
Pfaltz, A. Acc. Chem. Res. 2000, 33, 336. (f) Trost, B. M.;
Crawley, M. L. Chem. Rev. 2003, 103, 2921. (g) Trost, B.
M. J. Org. Chem. 2004, 69, 5813.
1
5
–
PdLn
PdLn
PdLn
22
21
OMEM
PdLn
AcO
13
Ph
(E)-17
18
(3) Heathcock, C. H. In Modern Synthetic Methods 1992;
Scheffold, R., Ed.; VHCA, VCH: Basel, Weinheim, 1992, 1;
and references therein.
4) (a) Fiaud, J.-C.; Malleron, J.-L. J. Chem. Soc., Chem.
Commun. 1981, 1159. (b) Åkermark, B.; Jutand, A. J.
Organomet. Chem. 1981, 217, C41. (c) Negishi, E.;
Matsushita, H.; Chatterjee, S.; John, R. A. J. Org. Chem.
1982, 47, 3188. (d) Trost, B. M.; Keinan, E. Tetrahedron
Lett. 1980, 21, 2591. (e) Trost, B. M.; Self, C. R. J. Org.
Chem. 1984, 49, 468.
–
PdLn
–
PdLn
23
(
Scheme 7
On the other hand, the formation of ketone 18 from the
lithium enolate 13 and the substrate (E)-17 is plausibly
explained by a double inversion, first, in the formation of
the p-allyl palladium complex 23 and, second, in the
approach of the nucleophilic enolate 13. The reaction fol-
lows this ‘simple’ path, because, in the E-configured sub-
strate 17, there is no ‘need’ for a thermodynamically
controlled Z-to-E interconversion.
(
5) (a) Review: Kazmaier, U. Curr. Org. Chem. 2003, 7, 317.
(
1
b) Trost, B. M.; Schroeder, G. M. J. Am. Chem. Soc. 1999,
21, 6759. (c) Braun, M.; Laicher, F.; Meier, T. Angew.
Chem. Int. Ed. 2000, 39, 3494; Angew. Chem. 2000, 112,
637. (d) You, S.-L.; Hou, X.-L.; Dai, L.-X.; Zhu, X.-Z.
3
Org. Lett. 2001, 3, 149.
(
6) (a) Kazmaier, U.; Zumpe, F. L. Angew. Chem. Int. Ed. 1999,
The configurations of the ketones 15 and 18 were as-
signed after ozonolysis that leads to the enantiomeric keto
aldehydes (R)-16 and (S)-16, respectively, whose absolute
38, 1468; Angew. Chem. 1999, 111, 1572. (b) Weiß, T. D.;
Helmchen, G.; Kazmaier, U. Chem. Commun. 2002, 1270.
7) (a) Bartels, B.; García-Yebra, C.; Helmchen, G. Eur. J. Org.
Chem. 2003, 1097. (b) Peña, D.; Minnaard, A. J.; de Vries,
A. H. M.; de Vries, J. G.; Feringa, B. L. Org. Lett. 2003, 5,
(
2
1
configuration is known, so that the assignment is possi-
ble by comparison of the sign of their optical rotations.
Thus, it has been shown that ketone-derived lithium eno-
lates in the presence of lithium chloride do not undergo a
precoordination to the transition metal in p-allyl palladi-
um complexes. They rather approach the p-allyl system
from the face opposite to the noble metal. Thus, pre-
formed lithium enolates follow the pathway of stabilized
carbanions in the allylic substitution.
4
75.
8) Takaya, H.; Akutagawa, S.; Noyori, R. Org. Synth. 1988, 67,
0.
9) Takaya, H.; Mashima, K.; Koyano, K. J. Org. Chem. 1986,
1, 629.
(
(
2
5
(
10) Seebach, D. Angew. Chem., Int. Ed. Engl. 1988, 27, 1624;
Angew. Chem. 1988, 100, 1685.
(11) Lloyd-Jones, G. C.; Stephen, S. C.; Fairlamb, I. J. S.;
Martorell, A.; Dominguez, B.; Tomlin, P. M.; Murray, M.;
Fernandez, J. M.; Jeffery, J. C.; Riis-Johannessen, T.;
Guerziz, T. Pure Appl. Chem. 2004, 76, 589.
(12) Enders, D. In Asymmetric Synthesis, Part B, Vol. 2;
Morrison, J. D., Ed.; Academic Press: New York, 1984,
Chap. 4.
In summary, it has been demonstrated that monoallylation
reactions of simple cyclic ketones are feasible provided
that the appropriate conditions are chosen. Aside from the
workhorse ligand’ BINAP at the transition metal, the use
of lithium enolates in the presence of lithium chloride are
keys to the success. It becomes more and more evident
that allyl palladium and preformed enolate chemistry are
indeed compatible.
‘
2
2
(13) Imai, M.; Hagihara, A.; Kawasaki, H.; Manabe, K.; Koga, K.
Tetrahedron 2000, 56, 179.
Synlett 2005, No. 19, 2968–2972 © Thieme Stuttgart · New York

2
972
M. Braun, T. Meier
LETTER
(
14) (a) Burger, E. C.; Tunge, J. A. Org. Lett. 2004, 6, 4113.
b) Behenna, D. C.; Stoltz, B. M. J. Am. Chem. Soc. 2004,
(22) Procedure for the Synthesis of (R)-4a.
(
A 100-mL two-necked flask is equipped with a magnetic
stirrer and charged with [Pd (dba) ]·CHCl (25.9 mg; 25
1
1
26, 15044. (c) Trost, B. M.; Xu, J. J. Am. Chem. Soc. 2005,
27, 2846.
2
3
3
mmol), (R)-8 (63.0 mg, 101 mmol), and LiCl (0.49 g, 12
mmol). The flask is closed with a septum, connected to a
(
15) The opposite diastereomer was predominantly obtained in
the rearrangement of allyl b-ketoesters: cf. ref. 14a. The
product 10 was isolated in 35% yield aside from recovered
carbonate 1d (40%). The fact that the latter is non-racemic
indicates an at least partial kinetic resolution.
combined N /vacuum line, evacuated for 4 h at 25 °C and
2
filled with N . A solution of 1c (0.562 g, 4.84 mmol) in
2
anhyd THF (13 mL) is added by syringe. The deep purple
mixture is stirred at 25 °C until the color changes to yellow
(approximately 1 min). A 500-mL two-necked flask is
equipped with a magnetic stirrer, a resistance thermometer,
(16) (a) Enders, D.; Eichenauer, H. Angew. Chem., Int. Ed. Engl.
1976, 15, 549; Angew. Chem. 1976, 88, 579. (b) Meyers, A.
I.; Williams, D. R.; Erickson, G. W.; White, S.; Druelinger,
M. J. Am. Chem. Soc. 1981, 103, 3081. (c) Hiroi, K.; Abe,
J.; Suya, K.; Sato, S.; Koyama, T. J. Org. Chem. 1994, 59,
a connection to the combined N /vacuum line and a septum.
2
The air in the flask is replaced by N , and diisopropylamine
2
(17.0 mL, 120.3 mmol) and THF (80 mL) are injected. After
cooling to –78 °C, a 1.6 M solution of n-BuLi in hexane (75
mL, 120.0 mmol) is added, whereby the temperature is kept
below –70 °C. After stirring at 0 °C for 30 min, it is cooled
again to –78 °C and treated with distilled, degassed
203.
(
17) Yanagisawa, A.; Kikuchi, T.; Kuribayashi, T.; Yamamoto,
H. Tetrahedron 1998, 54, 10253.
18) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456.
19) Braun, M.; Unger, C.; Opdenbusch, K. Eur. J. Org. Chem.
(
(
cyclohexanone (12.5 mL, 120.5 mmol) in 70 mL of THF.
After stirring at 0 °C for 30 min, the solution is stored at
– 78 °C. At –78 °C, 11 mL (5.09 mmol) of the enolate
solution are added to the 100-mL flask by syringe. After
stirring at –78 °C for 40 h, the mixture is poured into 100 mL
of phosphate buffer (pH = 7) and extracted four times with
CH Cl (50 mL each). The combined organic layers are
1
998, 2389.
20) Hayashi, T.; Yamamoto, A.; Hagihara, T. J. Org. Chem.
986, 51, 723.
(
(
1
21) Lassaletta, J.-M.; Fernández, R.; Martín-Zamora, E.; Díez,
E. J. Am. Chem. Soc. 1996, 118, 7002.
2
2
dried with MgSO and evaporated at 40 °C at a pressure that
4
does not fall below 200 mbar. The flask is connected by
glass tubes with two subsequent traps that are cooled to 0 °C
and –196 °C, respectively. When the flask containing the
crude product is heated to 50 °C at 0.07 mbar, pure product
is collected in the 0 °C trap. Yield of (R)-4a: 0.422 g (63%);
t = 24.8 min for (R)-4a; t = 26.6 min for (S)-4a; 90% ee.
R
R
Synlett 2005, No. 19, 2968–2972 © Thieme Stuttgart · New York