Enhanced Enanthioselectivity in the Desymmetrization of Meso-Biscarbamates

Full text of DOI:10.1021/jo971746r

J . Org. Chem. 1998, 63, 1339-1341
1339
as with ligand 1c (or 2c) (i.e. 79% ee for 1a versus 78%
ee for 1c), but the yield was still low (40% for 1a versus
97% for 1c).
En h a n ced En a n tioselectivity in th e
Desym m etr iza tion of Meso-Bisca r ba m a tes
Since cyclization requires the nonionized urethane to
be deprotonated, we envisioned that the low yield may
arise from sluggish deprotonation of the pronucleophile
in the case of the new ligands compared to 1c. Adding 1
equiv of triethylamine did increase the yield to 53% but
dramatically increased the ee to 96%! To establish
whether the effect of triethylamine was ligand related,
we reexamined the reaction with ligand 1c. Adding 1
equiv of triethylamine to an otherwise standard protocol
as previously described, oxazolidinone 4 was formed in
84% yield and with an ee of >99%!
To determine the generality of the effect, we studied
the series of desymmetrizations previously explored and
summarized in eqs 1 and 2 and Table 1. The data shows
two trends. First, the ee increased in every case to g94%.
Thus, the standard ligand 1c now provides a convenient
synthetic entry to oxazolidin-2-ones and, by deblocking,
to vicinal cis-hydroxy-amines with excellent enantiose-
lectivity. Second, the yield and ee had some dependence
on the quantity of triethylamine. The yields appear to
improve with lower amounts of base, but there may be a
tradeoff in terms of ee. The lack of a consistent trend
suggests that for optimum results in any specific case,
some variation should be explored. The lower yield with
increasing base may stem from base-catalyzed elimina-
tion competing with cyclization.
Barry M. Trost* and Daniel E. Patterson
Department of Chemistry, Stanford University,
Stanford, California 94305-5080
Received September 17, 1997
The palladium-catalyzed desymmetrization of meso-
2-ene-1,4-diol diesters has proven to give the monosub-
stitution products in high ee (Scheme 1, path a).1 On
the other hand, our “standard” ligand 1 gave the oxazo-
lidin-2-ones from the bis-carbamates (R ) NHTs) in
significantly lower ee’s.1a Since the leaving group is
involved in the enantiodiscriminating step, we believed
the differences derived from a leaving group effect on the
chiral discrimination, which we also observed in compar-
ing acetate to benzoate as the leaving group.1b As a
result, considering the utility of this asymmetric oxazo-
lidin-2-one synthesis, we have been designing new ligands
in an effort to enhance the ee of this cyclization.2 In the
course of these studies, we discovered a very simple
solution to this problem whereby the “standard” ligands
suffice and which has significant mechanistic implica-
tions.
Molecular modeling studies suggested that there may
exist a conformation wherein one of the carbonyl oxygens
of the ligand comes within van der Waals contact with
palladium, suggesting an electronic effect superimposed
on any steric effect that is responsible for the chiral
recognition.3 To strengthen such an interaction, we
synthesized the symmetrical (1a ) and unsymmetrical
(1b) p-methoxy ligands and examined the cyclization of
The significance of the base effect provides insight into
the mechanism of the asymmetric induction. The sim-
plest rationale recognizes that the ionization event, which
determines the ee, may not be the rate-limiting event (see
Scheme 2). In the absence of base, the ionization event
produces either π-allylpalladium intermediate A or B.
However, effective cyclization requires formation of the
zwitterion C or D to give product. Since the base present
is the carboxylate generated upon ionization, which
means a weak base is present in very low concentration,4
the rate of formation of C or D may become the rate-
limiting step or, at least, competitive with the ionization
3d generated normally in situ by reaction of the diol with
2 equiv of p-tosyl isocyanate. The urethane was treated
with 2.5 mol % of palladium catalyst precursor and 7.5
mol % of ligand in THF at 0 °C to room temperature.
Surprisingly, both the yield and the ee were somewhat
low using π-allylpalladium chloride dimer (5) as the
catalyst precursor with ligands 1a or 1b. Switching to
(dba)3Pd2‚CHCl3 (6) saw the ee return to the same level
(1) (a) Trost, B. M.; van Vranken; D. L.; Bingel, C. J . Am. Chem.
Soc. 1992, 114, 9327. (b) Trost, B. M.; Li, L.; Guile, S. D. J . Am. Chem.
Soc. 1992, 114, 8745. (c) Trost, B. M.; Pulley, S. R. J . Am. Chem. Soc.
1995, 117, 10143. (d) Trost, B. M.; Shi, Z. J . Am. Chem. Soc. 1996,
118, 3039. (e) Trost, B. M.; Madsen, R.; Guile, S. D. Tetrahedron Lett.
1997, 38, 1707.
(4) It is possible to envision loss of CO2 from the carbamic anion to
form the anion of tosylamide which then serves as base. However, the
(2) Trost, B. M.; Breit, B.; Peukert, S.; Zambrano, J .; Ziller, J . W.
Angew. Chem., Int. Ed. Engl. 1995, 34, 2386.
(3) Hagelin, H., unpublished results in these laboratories.
large difference in pKa between a carbamic acid and tosylamide
suggests that such decarboxylation would not be spontaneous and
would require protonation.
S0022-3263(97)01746-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/04/1998

1340 J . Org. Chem., Vol. 63, No. 4, 1998
Notes
Sch em e 1
Sch em e 2. Mech a n istic Ra tion a le for Ba se Effect
Ta ble 1
without without
This work has significant broader implications for
palladium-catalyzed desymmetrization of meso sub-
strates. To the extent that the initial ionization is
reversible, a deterioration of the ee will be seen. For
maximum chiral recognition, the kinetically preferred
π-allylpalladium intermediate must be captured by nu-
cleophile faster than capture by the departed leaving
group. These observations also suggest an explanation
for the higher enantioselectivity we observed with meso-
dibenzoates and meso-dicarbonates compared to meso-
diacetates despite being better leaving groups. The
poorer nucleophilicity of benzoate or carbonate compared
to acetate makes their return less competitive with
capture by nucleophile. Consideration of the competition
between return of leaving group and capture of nucleo-
phile becomes another factor in maximizing enantiodis-
crimination. At the same time, the results imply that
this family of ligands exhibits extraordinary levels of
molecular recognition in the ionization step with a broad
array of leaving groups.
with
with
oxazolidin- (C2H5)3N, (C2H5)3N, (C2H5)3N, (C2H5)3N,
entry diol
2-one
% yield
% ee
% yield
% ee
1
2
3
4
3a
7a
7b
9
4
97
78
1 eq. 84
0.5 eq. 79
1 eq. 85
0.5 equiv, 88
1 eq. 70
0.5 eq. 92
1 eq. 60
>99
>99
99
93
99
97
94
89
8a
8b
92
85
66
85
90
81
10
0.5 eq. 89
event. In this circumstance, significant return to starting
material competes with cyclization. Under such condi-
tions, more significant competition between the matched
and mismatched pathways would ensue leading to an
erosion of the ee. If the return of the initial π-allyl
intermediate to starting material could be eliminated, the
intrinsic enantiodiscrimination in the ionization step
could be observed. Addition of external base could affect
the process in two ways. First, it would speed the
conversion of A to C (and B to D) and thereby decrease
the competitive return of A to starting material (and B
to starting material). Second, the first step may involve
deprotonation to E or F fast and reversibly. Ionization
of these anions directly to C or D, which constitutes the
enantiodiscriminating event, then becomes the rate-
limiting event. The net result again becomes one in
which the kinetic chiral recognition is reflected in the
enantiomeric purity of the product. Of course, both
phenomena can be operating synergistically.
As previously indicated, this simplifies access to these
valuable building blocks. For example, many glycosidase
inhibitors can derive from these intermediates. We
previously demonstrated racemic syntheses of mannosta-
tin and allosamizoline which can now be made completely
enantioselective using this chemistry.5
(5) Trost, B. M.; Van Vranken, D. L. J . Am. Chem. Soc. 1993, 115,
444.

Notes
J . Org. Chem., Vol. 63, No. 4, 1998 1341
Ta ble 2. Exp er im en ta l Deta ils for Desym m etr iza tion of Meso Bisca r ba m a tes
starting
material
dba3Pd2/
CHCl3
TosNCO
ligand 1
6.9 mg,
0.01 mmol
6.9 mg,
0.01 mmol
12.5 mg,
0.018 mmol
12.5 mg,
0.018 mmol
12.5 mg,
0.018 mmol
11.4 mg,
0.0165 mmol
11.4 mg,
0.0165 mmol
11.4 mg,
0.0165 mmol
12.5 mg,
0.018 mmol
12.5 mg,
0.018 mmol
12.5 mg,
0.018 mmol
TEA
THF
productb
1c,a 60 mg,
0.122 mmol
1c,a 60 mg,
0.122 mmol
7a , 25 mg,
0.219 mmol
7a , 25 mg,
0.219 mmol
7a , 25 mg,
0.219 mmol
7b, 25.6 mg,
0.2 mmol
7b, 25.6 mg,
0.2 mmol
7b, 25.6 mg,
0.2 mmol
9, 39 mg,
0.219 mmol
9, 39 mg,
3.1 mg,
0.003 mmol
3.1 mg,
0.003 mmol
5.7 mg,
0.0055 mmol
5.7 mg,
0.0055 mmol
5.7 mg,
0.0055 mmol
5.3 mg,
0.005 mmol
5.3 mg,
0.005 mmol
5.3 mg,
0.005 mmol
5.7 mg,
0.0055 mmol
5.7 mg,
0.0055 mmol
5.7 mg,
0.0055 mmol
0.017 mL,
0.45 mL
4, 28.5 mg, 0.102 mmol,
84%, (>99% ee)
0.122 mmol
0.008 mL,
0.06 mmol
0.0305 mL,
0.219 mmol
0.015 mL,
0.109 mmol
none
0.45 mL
0.8 mL
0.8 mL
0.8 mL
0.7 mL
0.7 mL
0.7 mL
0.8 mL
0.8 mL
0.8 mL
4, 27.0 mg, 0.097 mmol,
79%, (>99% ee)
101 mg,
0.513 mmol
101 mg,
0.513 mmol
101 mg,
0.513 mmol
92.9 mg,
0.468 mmol
92.9 mg,
0.468 mmol
92.2 mg,
0.468 mmol
101 mg,
0.513 mmol
101 mg,
0.513 mmol
101 mg,
0.513 mmol
8a , 54.5 mg, 0.185 mmol,
85%, (99% ee)
8a , 56.3 mg, 0.193 mmol,
88%, (93% ee)
8a , 58.9 mg, 0.20 mmol,
92%, (85% ee)
8b, 42.6 mg, 0.14 mmol,
70%, (99% ee)
8b, 56.3 mg, 0.184 mmol,
92%, (97% ee)
8b, 53.5 mg, 0.175 mmol,
87%, (90% ee)
10, 46.8 mg, 0.131 mmol,
60%, (94% ee)
10, 69.2 mg, 0.193 mmol,
89%, (89% ee)
0.028 mL,
0.2 mmol
0.014 mL,
0.1 mmol
none
0.0305 mL,
0.219 mmol
0.015 mL,
0.109 mmol
none
0.219 mmol
9, 39 mg,
0.219 mmol
10, 51.5 mg, 0.104 mmol,
66%, (81% ee)
a
In this case, the starting material was preformed bis-urethane. b All enantiomeric excesses were determined by chiral HPLC: Chiralpak
AD column, 85:15 heptane:2-propanol, 1 mL/min, 254 nm, 4: (+) 23.22 min, (-) 28.42 min. 8a : (+) 24.03 min, (-) 30.98 min. 8b: (-)
26.60 min, (+) 43.72 min. Chiralcel OD column, 85:15 heptane:2-propanol, 1 mL/min, 254 nm, 10: (-) 12.29 min, (+) 14.51 min.
overnight. Solvent was removed in vacuo, and column chroma-
tography on silica (15% EtOAc/pet ether) gave the desired
product as a white solid. 4: mp 121-5 °C (lit.1a mp 131 °C);
[R]25 114.8 (c ) 2.15, CH2Cl2) (lit.1a [R]25 (99% ee) 114 (c )
Exp er im en ta l Section
Gen er a l. All reagents were obtained commerically and were
used without further purification. All reactions were carried out
under an inert atmosphere (N2) unless otherwise indicated.
Tetrahydrofuran was distilled from sodium benzophenone ketyl.
Triethylamine was distilled from calcium hydride. Column
chromatography was carried out on E. Merck silica gel 60 (60-
75 mesh) using the solvent system listed under the individual
experiments. All enantiomeric excesses were determined by
chiral HPLC on a solid stationary phase Chiralpak AD or
Chiralcel OD column using heptane/2-propanol mixtures as
eluents. Melting point were taken in open end capillary tubes
and are uncorrected.
Gen er a l P r oced u r e for Oxa zolid in -2-on e F or m a tion . To
a solution of the meso diol (0.2 mmol) in 0.35 mL of THF was
added tosyl isocyanate (101 mg, 0.47 mmol). This colorless
solution was stirred at room temperature for 15 min and at 60
°C for 30 min. The reaction was allowed to cool to room
temperature, and triethylamine (0.028 mL, 0.2 mmol) was
added. The resulting white slurry was cooled to 0 °C, and an
orange solution of tris(dibenzylidineacetone)dipalladium(0) chlo-
roform complex (5.2 mg, 0.005 mmol) and ligand 1 (11.4 mg,
0.017 mmol) in 0.35 mL of THF was added. The orange reaction
solution was stirred at 0 °C for 2 h and at room-temperature
D
D
2.52, CH2Cl2)). 8a : mp 127-9 °C; (99% ee) 95.5 (c ) 1.38, CH2-
Cl2), (93% ee) 91.4 (c ) 1.85, CH2Cl2), (85% ee) 80.0 (c ) 1.77,
CH2Cl2) (lit.2 [R]25 (97% ee) 87.5 (c ) 2.27, CH2Cl2)). 8b: mp
D
171-2 °C; [R]25 (99% ee) 96.6 (c ) 3.40, CH2Cl2), (97% ee) 94.6
D
(c ) 1.43, CH2Cl2), (90% ee) 95.6 (c ) 1.37, CH2Cl2) (lit.2 [R]25
D
(70.2 (c ) 2.68, CH2Cl2)). 10: mp 164-5 °C (lit.1a mp 160-1
°C); [R]25 (94% ee) 62.0 (c ) 4.44, CH2Cl2), (89% ee) 68.2 (c )
D
1.77, CH2Cl2), (82% ee) 55.1 (c ) 4.44, CH2Cl2) (lit.1a [R]25D (87%)
69.4° (c ) 2.08, CH2Cl2)). The details for each of the individual
runs are summarized in Table 2.
Ack n ow led gm en t. We thank the National Science
Foundation and the National Institutes of Health for
their generous support of our programs. Mass spectra
were provided by the Mass Spectrometry Facility at the
University of CaliforniasSan Francisco supported by
the NIH Division of Research Resources.
J O971746R