On asymmetric induction in allylic alkylation via enantiotopic facial discrimination

Full text of DOI:10.1021/ja960649x

J. Am. Chem. Soc. 1996, 118, 6297-6298
6297
Scheme 1. Asymmetric Induction via Enantiotopic Facial
Discrimination
On Asymmetric Induction in Allylic Alkylation via
Enantiotopic Facial Discrimination
Barry M. Trost,* Michael J. Krische, Rumen Radinov, and
Giuseppe Zanoni
Department of Chemistry, Stanford UniVersity
Stanford, California 94305
ReceiVed February 28, 1996
Table 1. Asymmetric Sulfone Formation with Crotyl Carbonatea
Metal catalyzed allylic alkylations differentiate themselves
from most transition metal catalyzed reactions in that many
different phenomena may lead to asymmetric induction.1-3 One
of the lesser explored phenomena involving discrimination in
complexing enantiotopic faces followed by ionization to lead
to asymmetric induction has not been very successful in Pd-
catalyzed processes.4,5 Scheme 1 outlines some of the issues
that complicate such studies. A major obstacle in Pd-catalyzed
reactions stems from the propensity of nucleophiles to attack
at the sterically more accessible terminal position (path b) which
leads to an achiral product.6 While changing the metal to Mo7
or W8 may overcome this issue, the greater scope of Pd-
catalyzed reactions makes them the method of choice when
possible. Nevertheless, within a limited range of allylic
substrates, some promising results with W involving enantio-
topic discrimination have been obtained.9 A second obstacle
is the facility of migration of Pd from one enantiotopic face to
the other (path c) via a η3-η1-η3 mechanism.1,6a Indeed, such
a phenomenon, if fast relative to nucleophilic attack, can lead
to asymmetric induction if the difference in activation energies
for path a and ent-path a is sufficiently large in the presence of
chiral ligands. In such an event, the nucleophilic addition step
determines the asymmetric induction.10 A third obstacle derives
from the fact that the preferred motions with chiral scalemic
ligands during the ionization and alkylation steps must be
opposite. Thus, if the former is a matched event, the latter must
become a mismatched one and vice versa. Such stereochemical
mismatching would tend to favor paths b and c competing with
path a or ent-path a. We wish to report that asymmetric Pd-
temp time isolated ratiob
entry substrate ligand (°C) (h) yield
2:3 %ee 2c config
1
2
3
4
5
6
7d
E-1a
E-1a
E-1a
E-1a
Z-1a
Z-1a
E-1a
4
4
5
6
4
4
4
0
20
0
0
0
1
2
3
2
2
97% 83:17 90 (92)
R
R
S
R
R
R
R
91% 76:24 80
84% 70:30 80
92% 69:31 28
92% 79:21 29
20 0.33 99% 82:18 16
92% 79:21 (92)
0
2
a Reaction performed with 0.25 mol % (dba)3Pd2‚CHCl3, 0.6 mol
% L*, 6% (n-C6H13)4NBr, 1.5 equiv PhSO2Na, 3:1 CH2Cl2:H2O unless
otherwise stated. b Determined by GC or HPLC analysis. c Determined
by chiral HPLC (Chiralpak AD column) using 90:10 heptane-
isopropanol as eluting solvent; values in parentheses were determined
by NMR spectroscopy with a chiral shift reagent. d p-Toluenesulfinate
rather than benzenesulfinate was employed as the nucleophile.
catalyzed allylic alkylations, which differentiate enantiotopic
faces of the substrate, can be synthetically useful with the
modular asymmetric ligands under development in these
laboratories.3
Our initial efforts focused on the concept of rapidly inter-
converting π-allyl intermediates thereby requiring chiral rec-
ognition in the nucleophilic addition step.10a We examined the
reaction of eq 1 because prior studies with a tartrate-derived
(1) Trost, B. M.; Van Vranken, D. L. Chem. ReV. 1996, in press.
(2) For a few examples with the most successful ligands, see: (a) Trost,
B. M.; Murphy, D. J. Organometallics 1985, 4, 1143. (b) Hayashi, T.;
Yamamoto, A.; Hagihara, I.; Ito, Y. Tetrahedron Lett. 1986, 27, 191. (c)
Dawson, G. J.; Frost, C. G.; Williams, J. M. J.; Coote, S. J. Tetrahedron
Lett. 1993, 34, 3149. (d) Springz, J.; Helmchen, G. Tetrahedron Lett. 1993,
34, 1769. (e) von Matt, P.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1993,
32, 566. (f) Trost, B. M.; Peukert, S.; Zambrano, J.; Ziller, J. W. Angew.
Chem., Int. Ed. Engl. 1995, 34, 2386.
(3) Trost, B. M.; Van Vranken, D. L. Angew. Chem., Int. Ed. Engl. 1992,
31, 228. Trost, B. M.; Van Vranken, D. L.; Bingel, C. J. Am. Chem. Soc.
1992, 114, 9327. Trost, B. M.; Li, L.; Guile, S. D. J. Am. Chem. Soc. 1992,
114, 8745. Trost, B. M.; Pulley, S. R. J. Am. Chem. Soc. 1995, 117, 10143.
(4) For a creation of axial chirality, see: Legros, J. Y.; Fiaud, J. C.
Tetrahedron 1994, 50, 465. Gil, R.; Fiaud, J. C. Bull. Soc. Chim. Fr. 1994,
584.
bis-phosphine ligand (DIOP) and chiral 2-(2-diphenylphos-
phino)oxazolines have suggested that such equilibration was
rapid.5,11 Table 1 summarizes the results we obtained for the
reaction of crotylmethyl carbonate (1a, R1 ) OCH3) using our
modular ligands 4,3 5,10a and 62g under phase transfer conditions
(5) Complete or partial equilibration has normally been observed. See:
(a) Hiroi, K.; Makino, K. Chem. Pharm. Bull. 1988, 36, 1744. (b) Hayashi,
T.; Kishi, K.; Yamamoto, H.; Ito, Y. Tetrahedron Lett. 1990, 31, 1743. (c)
Hayashi, T.; Yamamoto, A.; Ito, Y. Tetrahedron Lett. 1988, 29, 669. (d)
Hayashi, T.; Ohno, A.; Lu, S.; Matsumoto, Y.; Uozumi, Y.; Miki, M.;
Yanagi, K. J. Am. Chem. Soc. 1994, 116, 775. (e) Takemoto, T.; Nishikimi,
Y.; Sodeoka, M.; Shibasaki, M. Tetrahedron Lett. 1992, 33, 3572, 3531.
(f) Yamamoto, K.; Deguchi, R.; Ogimura, Y.; Tsuji, J. Chem. Lett. 1984,
1657. (g) Gene´t, J. P.; Grisone, S. Tetrahedron Lett. 1988, 29, 4543.
(6) (a) Godleski, S. A. ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Semmelhack, M. F., Eds.; Pergamon Press: Oxford: Vol. 4,
Chapter 3.3, pp 585-662. (b) Trost, B. M.; Hung, M.-H. J. Am. Chem.
Soc. 1984, 106, 6837.
(see footnote a of Table 1). The absolute configuration was
established by chemical correlation12 with S-2-butylphenyl
sulfone and comparison to the literature.5a
The reaction with ligand 4 gives the highest ee recorded, 92%.
In addition, the regioselectivity favoring 2 improves from 3:1
with other chiral ligands5a,11 to 5:1 with 4. However, inspection
of the table reveals that it does not arise by rapid equilibration
of the π-allyl intermediates with enantiodiscrimination occurring
in the subsequent alkylation step. Notably, the racemic regio-
isomeric substrate 1-buten-3-yl methyl carbonate gives a 94%
yield of an 87:13 mixture of 2:3 in which 2 is racemic in contrast
to the high ee with the achiral E substrate. Thus, in contrast to
other chiral ligands,5 the source of the enantiodiscrimination
(7) Trost, B. M.; Lautens, M. J. Am. Chem. Soc. 1987, 109, 1469;
Tetrahedron 1987, 43, 4817.
(8) Trost, B. M.; Tometzki, G. B.; Hung, M.-H. J. Am. Chem. Soc. 1987,
109, 2176. Trost, B. M.; Hung, M.-H. J. Am. Chem. Soc. 1983, 105, 7757.
(9) Lloyd-Jones, G. C.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1995,
34, 462.
(10) (a) Trost, B. M.; Bunt, R. C. Angew. Chem., Int. Ed. Engl. 1996,
35, 99. (b) Dawson, G. J.; Williams, J. M. J.; Coote, S. J. Tetrahedron
Lett. 1995, 36, 461. (c) Auburn, P. R.; Mackenzie, P. B.; Bosnich, B. J.
Am. Chem. Soc. 1985, 107, 2033.
(11) Eichelmann, H.; Gais, H.-J. Tetrahedron: Asymm. 1995, 6, 643.
(12) cf.: Hiroi, K.; Kitayama, R.; Sato, S. Chem. Pharm. Bull. 1984,
32, 2628.
S0002-7863(96)00649-X CCC: $12.00 © 1996 American Chemical Society

6298 J. Am. Chem. Soc., Vol. 118, No. 26, 1996
Communications to the Editor
Table 2. Asymmetric Induction in Cyclization to N-Heterocylesa
sub- (η3-C3H5PdCl)2 temp time isolated
7b gave good enantioselectivities with the R,R-ligand 4 giving
the (S)-2-vinylpyrrolidine and piperidine, respectively. On the
other hand, starting with the chiral racemic precursors 9a and
9b, only low enantioselectivities have been observed. These
results are quite consistent with the enantioselectivity being
determined by asymmetric recognition in the complexation step.
A completely different picture emerges in going from the
six- to seven-membered ring substrate. With the achiral
substrate 7c, the reaction was very slow, gave low yields even
at ambient temperature, and gave a maximum ee of 51% (entry
5). Furthermore, the major enantiomer produced was R (ent-
8c). On the other hand, the reaction using the chiral racemic
precursor 9c (entry 9) went to completion with 1% catalyst
between -35 and 0 °C in 2.5 h to give both high yield and
high ee of the S enantiomer 8c. This remarkable behavior of
7c and 9c in contrast to the other ring sizes can be rationalized
by the notion of the need to dock the substrate in a chiral
pocket.14 Such a pocket should have steric constraints wherein
a limit will be reached as to what can sterically be accom-
modated. Apparently that limit is reached between 7b and 7c.
On the other hand, fitting the regioisomeric starting material
9c into the pocket orients the sterically bulky portion of the
molecule toward a less sterically demanding region thereby
permitting more facile ionization, although not as rapidly as
either 7a or 7b. The high enantioconvergency observed with
the racemic substrate requires that pseudoracemization (i.e., a
mechanistic pathway which would constitute racemization in
the presence of achiral ligands) be faster than cyclization.
Apparently, the slower rate of formation of a seven-membered
ring compared to a five- or six-membered ring allows the
equilibration to dominate. Interestingly, the formation of the
same enantiomeric product with high ee for 7a, 7b, and 9c
means that the sense of chiral recognition for the two mecha-
nisms for enantiodiscrimination must be the same. Thus, the
product determining intermediate is the kinetically accessible
one from 7a and 7b (but not 7c) but becomes accessible only
by thermodynamic equilibration from 9c (but not 9a nor 9b).
The results reported herein demonstrate the utility of enan-
tiofacial discrimination for asymmetric induction in Pd-catalyzed
reactions. The intramolecular version would have particular
utility in order to control unambiguously the regioselectivity
of the process, a phenomenon that may result from both the
effect of the ligand and the nature of the nucleophile. Com-
bining the concept of a cleavable tether with the current reaction
should extend the utility of this asymmetric process to noncyclic
products. Choice of regioisomeric substrate may be important.
A delicate balance can exist between enantiodiscrimination by
complexation in the ionization event and by alkylation of rapidly
equilibrating diastereomeric intermediates as a function of steric
bulk of the substrate and ring size. By properly choosing the
mechanism for chiral recognition, high ee can be achieved in
formation of five-, six-, and seven-membered heterocyclic rings
in which the stereochemistry of the product is dictated by the
“chiral space” regardless of the mechanism. The products
obtained herein are useful building blocks as in a synthesis of
indolizidines and quinolizidines13b,15 and in ring expansions to
medium-sized rings.16
entry strate
mol %
(°C)
(h) yield %ee (er S:R)
1
2
3
4
5
6
7
8
9
7a
7a
7a
7b
7c
7c
9a
9b
9c
5
5
1
2.5
1
4
2.5
5
-25
6
48
2.5
16
7
1.5
16
16
90
92
97
82
80 (90:10)
81 (90.5:9.5)
91 (95.5:4.5)
88 (94:6)
-50
-45
-30
0
(30)c 51 (24.5:75.5)
20
-78
-50
48
70
83
84
41 (29.5:70.5)
4 (48:52)
6 (53:47)
1
-35 to 0 2.5
92 (96:4)
a All reactions were performed with a ligand to (η3-C3H5PdCl)2 ratio
of 2.5-3 in THF. b Determined by HPLC using a Chiralcel OD column
using 99.5:0.5 heptane-isopropanol containing 0.1% diethylamine as
eluting solvent. c Percent conversion determined by GC.
occurs in the alkene complexation, the first example of such
chiral recognition with this family of ligands. Similar results
were obtained using tiglyl methyl carbonate 1b whereby 2b of
94% ee was obtained.
Attempts to extend this asymmetric induction to amine
nucleophiles were thwarted by an unfavorable regioselectivity.
Resolution of this problem lay in tethering the nucleophile.
Alkylation in the presence of the chiral scalemic ligand 4 only
gives internal alkylation regardless of the tether length for n )
1, 2, or 3 (see eq 2 and Table 2). The enantioselectivities were
established by chiral stationary phase HPLC. The absolute
configurations were determined as shown in eq 3, all of which
Acknowledgment. We thank the National Science Foundation and
the National Institutes of Health, General Medical Sciences, for their
generous support of our programs. Mass spectra were generously
provided by the Mass Spectrometry Facility, University of California-
San Francisco, supported by the NIH Division of Research Resources.
Palladium salts were loaned by Johnson Matthey Alfa Aesar.
Supporting Information Available: Characterization data for 2a
(R ) H, Ar ) Ph), 2b (R ) CH3, Ar ) Ph), 8a (n ) 1), 8b (n ) 2),
and 8c (n ) 3) and typical experimental procedures for an inter- and
intramolecular reaction (3 pages). Ordering information is given on
any current masthead page.
were synthesized independently as either S or R enantiomers
from commercially available S or R amino acids.13
Several aspects revealed by the data are quite interesting. For
the five- and seven-membered rings, lowering the amount of
palladium precursor had a more beneficial effect on ee (entries
2 and 3) than lowering temperature (entries 1 and 2). For five-
and six-membered ring formation, the achiral substrates 7a and
JA960649X
(14) Trost, B. M.; Breit, B.; Organ, M. G. Tetrahedron Lett. 1994, 35,
5817.
(15) Hassner, A.; Maurya, R.; Padwa, A.; Bullock, W. H. J. Org. Chem.
1991, 56, 2775.
(16) Vedejs, E.; Hagen, J. P.; Roach, B. L.; Spear, K. L. J. Org. Chem.
1978, 43, 1185. Vedejs, E.; Arco, M. S.; Powell, D. W.; Renga, J. M.;
Singer, S. P. J. Org. Chem. 1978, 43, 4831.
(13) (a) Cf., for 5 ring: Pettit, G. R.; Singh, S. B.; Herald, D. L.; Lloyd-
Williams, P.; Kantoci, D.; Burkett, D. D.; Barko´czy, J.; Hogan, F.; Wardlaw,
T. R. J. Org. Chem. 1994, 59, 6287. For 6 ring and 7 ring: Pauly, R.;
Sasaki, N. A.; Potier, P. Tetrahedron Lett. 1994, 35, 237. (b) For racemic
8, n ) 1 and 2, see: Edstrom, E. D. J. Am. Chem. Soc. 1991, 113, 6690.