Regioselective catalytic allylic alkylation directed by removable 2-PyMe2Si group

Full text of DOI:10.1021/ja0157346

J. Am. Chem. Soc. 2001, 123, 6957-6958
6957
Table 1. Palladium-Catalyzed Directed Allylic Alkylation of 1
with Various Nucleophiles
Regioselective Catalytic Allylic Alkylation Directed
by Removable 2-PyMe2Si Group
Kenichiro Itami, Tooru Koike, and Jun-ichi Yoshida*
Department of Synthetic Chemistry and
Biological Chemistry, Kyoto UniVersity
Yoshida, Kyoto 606-8501, Japan
ReceiVed February 28, 2001
entry
1
2
catalyst
yield (%) ratio (3/4)
ReVised Manuscript ReceiVed June 11, 2001
1a 1a NaCH(CO2Me)2
[allylPdCl]2
PPh3
[allylPdCl]2
P(C6F5)3
[allylPdCl]2
P(C6F5)3
48
67
72
82
99
51
62
56
48
45
90/10
(3aa/4aa)
94/6
(3aa/4aa)
95/5
(3ab/4ab)
88/12
(3ac/4ac)
100/0
(3bb/4bb)
100/0
(3bc/4bc)
100/0
(3bd/4bd)
100/0
(3be/4be)
0/100
2a
An intramolecular directing group provides a powerful strategy
for enhancing the efficiency of an otherwise sluggish process and
for steering the course of the reaction by taking advantage of
attractive substrate-reagent interactions.1 In most cases, such
reactions are directed by the suitable heteroatoms on the substrate.
However, the greatest flaw of the current directed reaction is the
difficulty in removing or functionalizing such directing groups
after the reaction.2
Recently, we discovered that a 2-pyridyldimethylsilyl (2-PyMe2-
Si) group functions as the removable directing group for various
metal-catalyzed or -mediated processes.3 In ongoing efforts to
exploit this group for the metal-catalyzed reactions, we paid
particular attention to the regioselectivity problem in the pal-
ladium-catalyzed allylic alkylation.4 When unsymmetrically sub-
stituted allylic substrates are used, nucleophiles are preferentially
introduced into the allylic terminus that is sterically less hindered.
In this communication, we report on the palladium-catalyzed
regioselective allylic alkylations5 that are efficiently directed by
the removable 2-PyMe2Si group. Moreover, we observed the
dramatic switch of the regioselectivity by the type of nucleophile
(eq 1).
2a 1a 2a
3a 1a NaCH(CO2Et)2
2b
4a 1a NaC(Me)(CO2Et)2 [allylPdCl]2
2c
P(C6F5)3
[allylPdCl]2
P(C6F5)3
[allylPdCl]2
P(C6F5)3
5a 1b 2b
6a 1b 2c
7a 1b NaCH(CN)(CO2Et) [allylPdCl]2
2d
P(C6F5)3
[allylPdCl]2
P(C6F5)3
8a 1b NaCH(CN)2
2e
9b 1a (CH2dCH)SnBu3 Pd2(dba)3
2f
10b 1a PhSnMe3
2g
LiCl (3 equiv)
Pd2(dba)3
LiCl (3 equiv)
(3af/4af)
0/100
(3ag/4ag)
a Reactions were performed in THF at room temperature. b Reactions
were performed in DMF at room temperature.
P(C6F5)3 gave rise to higher yield and regioselectivity (entry 2).
This unusual inner site-selective allylic alkylation was found to
be a general phenomenon for the stabilized carbon nucleophiles
(entries 3-8). Quite interestingly, the use of organotin compound
as a nucleophile,6 on the other hand, gave rise to the complete
switch of the regioselectivity of the reaction, in which the
nucleophilic attack occurred at the allylic terminus remote from
the silyl group (entries 9 and 10). To the best of our knowledge,
such a regioselectivity switch has never been observed in the
(π-allyl)palladium chemistry.
2-PyMe2Si-substituted allylic acetates (1a and 1b) were easily
prepared by the reaction of 2-PyMe2SiCH2Li3b with R,â-unsatur-
ated aldehydes followed by acetylation. Subjection of 1a and
NaCH(CO2Me)2 (2a) to the action of Pd/PPh3 catalyst led to the
predominant production of 3aa as the result of a nucleophilic
attack at the more substituted carbon of the allylic moiety (Table
1, entry 1). The catalyst/ligand combination of [allylPdCl]2 and
In general, the palladium-catalyzed allylic substitution occurs
at the less substituted carbon of the allylic moiety.4 Indeed, Szabo´
has established in the related allylic substitution of â-silyl allylic
acetates that the nucleophilic attack selectively occurs at the allylic
terminus remote from the silyl group because of the steric and
electronic reasons.7 The allylic acetate 5, in which SiMe2 is
substituted by CH2, underwent the alkylation with good regiose-
lectivity (76/24) favoring the nucleophilic attack at the more
hindered allylic carbon (eq 2). This result, together with the work
(1) For an excellent review, see: Hoveyda, A. H.; Evans, D. A.; Fu, G. C.
Chem. ReV. 1993, 93, 1307.
(2) Examples of a removable directing group in metal-catalyzed reactions,
see: (a) Jun, C. H.; Lee, H.; Hong, J. B. J. Org. Chem. 1997, 62, 1200. (b)
Breit, B. Angew. Chem., Int. Ed. 1998, 37, 525.
(3) (a) Yoshida, J.; Itami, K.; Mitsudo, K.; Suga, S. Tetrahedron Lett. 1999,
40, 3403. (b) Itami, K.; Mitsudo, K.; Yoshida, J. J. Org. Chem. 1999, 64,
8709. (c) Itami, K.; Nokami, T.; Yoshida, J. Org. Lett. 2000, 2, 1299. (d)
Itami, K.; Mitsudo, K.; Kamei, T.; Koike, T.; Nokami, T.; Yoshida, J. J. Am.
Chem. Soc. 2000, 122, 12013. (e) Itami, K.; Kamei, T.; Mitsudo, K.; Nokami,
T.; Yoshida, J. J. Org. Chem. 2001, 66, 3970. (f) Itami, K.; Nokami, T.;
Yoshida, J. J. Am. Chem. Soc. 2001, 123, 5600.
(4) For reviews, see: (a) Godleski, S. A. In ComprehensiVe Organic
Synthesis; Trost, B. M., Fleming, I., Semmelhack, M. F., Eds.; Pergamon:
New York, 1991; Vol.4, Chapter 3.3. (b) Frost, C. G.; Howarth, J.; Williams,
J. M. J. Tetrahedron: Asymmetry 1992, 3, 1089. (c) Tsuji, J. Palladium
Reagents and Catalysts; John Wiley: Chichester, 1995.
(5) For the directed allylic substitution using nonremovable directing group,
see: (a) Didiuk, M. T.; Morken, J. P.; Hoveyda, A. H. J. Am. Chem. Soc.
1995, 117, 7273. (b) Didiuk, M. T.; Morken, J. P.; Hoveyda, A. H. Tetrahedron
1998, 54, 1117. (c) Krafft, M. E.; Fu, Z.; Procter, M. J.; Wilson, A. M.; Dasse,
O. A.; Hirosawa, C. Pure Appl. Chem. 1998, 70, 1083.
of Szabo´, clearly implies that the unusual regioselectivity observed
for 1 (Table 1, entries 1-8) is primarily attributed to the effect
of pyridyl group and that the effect of silyl group is at least
additive with regard to the inner site regioselectivity.
(6) Valle, L. D.; Stille, J. K.; Hegedus, L. S. J. Org. Chem. 1990, 55, 3019.
(7) Macsa´ri, I.; Hupe, E.; Szabo´, K. J. J. Org. Chem. 1999, 64, 9547.
10.1021/ja0157346 CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/23/2001

6958 J. Am. Chem. Soc., Vol. 123, No. 28, 2001
Communications to the Editor
Scheme 1
Figure 1. X-ray crystal structure of 8 (40% probability thermal
ellipsoids).
Whereas soft carbon nucleophiles add to (π-allyl)palladium
intermediate directly on the allyl ligand, nucleophilic attack of
organotin compounds occurs to the metal (transmetalation), and
the subsequent reductive elimination process produces the prod-
ucts.10 In the (π-allyl)palladium complex 8, the chlorine atom
sits on palladium cis to the allylic C1 atom. On the other hand,
the coordination site cis to the allylic C3 atom is occupied by the
intramolecular pyridyl group. We assume that this regioselective
transmetalation and reductive elimination might be responsible
for the complete regioselectivity switch.
To gain mechanistic insight into this unusual regioselectivity,
we began by establishing the nature of (allyl)palladium complex
generated from 1. Thus, the stoichiometric reaction of 1a and
Pd2(dba)3‚CHCl3 was conducted. However, the corresponding
(allyl)palladium complex was decomposed during the isolation,
presumably because of its instability. After several trials, we found
that the addition of LiCl produces an air-stable (π-allyl)palladium
complex 8 in 76% isolated yield (eq 3). Recrystallization from
Finally, the removal of 2-PyMe2Si group was demonstrated as
shown in eq 4. The LiAlH4 reduction of 3 afforded cyclic silyl
ether 9 in good yield with in situ removal of the pyridyl group.
Cyclic ether 9 was then subjected to H2O2 oxidation11 to give the
corresponding triol, which was converted into the corresponding
triacetate 10 in good overall yield.
EtOAc/hexane generated orange crystals. The molecular structure,
characterized by X-ray crystallographic analysis, depicts syn-η3-
coordination at the allylic moiety (Figure 1). Additionally, the
intramolecular pyridyl group occupies one coordination site trans
to the terminal allylic C1 atom with a pseudo square planar
geometry. More importantly, the distortion of allylic moiety was
observed. The Pd-C bond trans to chlorine (Pd-C3 ) 2.183 Å)
is substantially longer than the Pd-C bond trans to nitrogen (Pd-
C1 ) 2.089 Å). The allylic C-C bond trans to chlorine (C2-
C3 ) 1.365 Å) is substantially shorter than the allylic C-C bond
trans to nitrogen (C1-C2 ) 1.413 Å). The stoichiometric reaction
of 8 and 2b afforded 3ab with 90% regioselectivity, which is
close to that observed in the catalytic reaction (Table 1, entry 3).
One explanation for this allylic distortion may be the result of
the difference of trans influence.8 The stronger trans influence
of chlorine compared to the pyridyl group might elongate the
Pd-C3 bond. In the catalytic reactions, difference of the trans
influence in (π-allyl)palladium intermediate should be greater than
that in 8, since the trans influence of phosphine ligand is known
to be stronger than that of chlorine.8 Another plausible explanation
may be the structural reason. Allylpyridyl chelate might be
responsible for the allylic distortion.
In summary, we have established that the palladium-catalyzed
allylic alkylation was effectively directed by the removable
2-PyMe2Si group and that the regioselectivity can be completely
switched by the type of nucleophile. The structural analysis of
the (π-allyl)palladium complex revealed the distortion of allyl
ligand on palladium, which might be the reason for unusual inner
site-selective nucleophilic attack of soft carbon nucleophiles. We
believe that this study should provide the strategic basis for
controlling the regioselectivity in the palladium-catalyzed allylic
alkylation.
Thus, the observed inner site-selective allylic alkylation for
the stabilized carbon nucleophiles is most likely attributed to the
trans influence or chelation-induced distortion of allylic moiety,
in which the soft carbon nucleophile preferentially attacks the
longer and, accordingly, more reactive Pd-C bond (Scheme 1).9
The explanation for the regioselectivity observed with the
organotin nucleophiles should be mechanistically different.
Supporting Information Available: Experimental procedures and
analytical and spectroscopic data; tables of crystallographic data for 8
(PDF). This material is available free of charge via the Internet at
http://pubs.acs.org.
JA0157346
(10) Sheffy, F. K.; Godschalx, J. P.; Stille, J. K. J. Am. Chem. Soc. 1984,
106, 4833. See also, Kurosawa, H.; Ogoshi, S.; Kawasaki, Y.; Murai, S.;
Miyoshi, M.; Ikeda, I. J. Am. Chem. Soc. 1990, 112, 2813.
(11) (a) Tamao, K. In AdVances in Silicon Chemistry, Larson G. L., Ed.;
JAI Press Inc.: Greenwich, CT, 1996; Vol. 3, p1. (b) Fleming, I.
ChemTracts: Org. Chem. 1996, 1. (c) Jones, G. R.; Landais, Y. Tetrahedron
1996, 52, 7599.
(8) Appleton, T. G.; Clark, H. C.; Manzer, L. E. Coord. Chem. ReV. 1973,
10, 335. The general trend for the trans influences is: PR3 > Cl > C5H5N.
(9) For discussion of allylic distortion in asymmetric allylic substitution,
see: von Matt, P.; Lloyd-Jones, G. C.; Minidis, A. B. E.; Pfaltz, A.; Macko,
L.; Neuburger, M.; Zehnder, M.; Ru¨egger, H.; Pregosin, P. S. HelV. Chim.
Acta 1995, 78, 265.