A regioselective Ru-catalyzed alkene-alkyne coupling to form (Z,Z)-1,3-dienes

Article abstract of DOI:10.1021/ol8028324

Generally, Ru-catalyzed alkene-alkyne coupling forms 1,4-dienes. A version of this process which generates a 1,3-diene as the Z,Z-isomer preferentially has now been observed. The scope of this new atom-economic process is described herein.

Full text of DOI:10.1021/ol8028324

ORGANIC
LETTERS
2009
Vol. 11, No. 5
1071-1074
A Regioselective Ru-Catalyzed
Alkene-Alkyne Coupling To Form
(Z,Z)-1,3-Dienes
Barry M. Trost* and Alicia Martos-Redruejo
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
bmtrost@leland.stanford.edu
Received December 8, 2008
ABSTRACT
Generally, Ru-catalyzed alkene-alkyne coupling forms 1,4-dienes. A version of this process which generates a 1,3-diene as the Z,Z-isomer
preferentially has now been observed. The scope of this new atom-economic process is described herein.
The development of atom-economic reactions1 constitutes a
major theme of our program. The Ru-catalyzed alkene-alkyne
coupling reaction to form 1,4-dienes has been one of the
most exciting of these methods and has found utility in
simplifying synthetic strategies to complex bioactive targets.2
As shown in Figure 1, path a, the proposed mechanism
involves formation of a ruthenacyclopentene3,4 which then
undergoes a ꢀ-hydrogen insertion followed by a reductive
elimination to give diene products. The geometrical con-
straints imposed by the ruthenacycle would be expected to
favor insertion into the external C-Ha bond (path a) since
the C-H and Ru-C bonds can best become syn-coplanar,
which is optimum.
In accordance with this expectation, up to the present, only
1,4-dienes were observed.5 Nevertheless, endocyclic ꢀ-hy-
drogen insertions within the ruthenacycle (i.e., of C-Hb, path
b) have been reported as feasible.6
In the course of our continuing studies, we have now
observed that this latter path may dominate for certain
(1) (a) Trost, B. M. Science 1991, 254, 1471. (b) Trost, B. M. Angew.
Chem., Int. Ed. Engl. 1995, 34, 259.
(2) (a) Wender, P. A.; Miller, B. L. Org. Synth.: Theory Appl. 1993, 2,
27. (b) Bertz, S. H.; Sommer, T. J. Org. Synth.: Theory Appl. 1993, 2, 67.
(3) Trost, B. M.; Indolese, A. J. Am. Chem. Soc. 1993, 115, 4361
.
Figure 1. Proposed mechanism.
(4) Khand, I. U.; Knox, G. R.; Pauson, P. L.; Watts, W. E.; Foreman,
F. I. J. Chem. Soc., Perkin Trans. 1 1973, 977
.
10.1021/ol8028324 CCC: $40.75
Published on Web 02/04/2009
2009 American Chemical Society

substrates even when path a is still feasible. The fact that
1,3-dienes, important building blocks for subsequent atom-
economic cycloadditions, are the products of this pathway
attaches particular importance to this new observation. In
our study of the effect of substituents at the allylic position
of the alkene partner, we examined the reaction of allyl
acetate with ethyl 2-butynoate (Figure 2). Such a study was
also motivated by the prospect of the Ru catalyst initiating
ionization of the allyl ester in competition with the
alkene-alkyne coupling.7
ing the extraordinary chemoselectivity observed in these Ru
catalyzed reactions. No formation of 1,4-dienes was observed
in any of the crude product mixtures.
Table 1. Optimization of Coupling of Ally Acetate 1 and Ethyl
2-Butynoate 2
entry [2] 1:2 ratio 3 (mol %) T (°C) timea (h) % yield
1
2
3
4
5
6
7
8
9
0.8
0.8
0.8
0.8
0.8
0.8
0.4
0.4
1.6
1.6
1:2
1:1
1:2
1:2
1:2
1:1
1:1
1:1
1:2
1:1
10
10
5
10
5
5
5
5
5
rt
rt
rt
40
40
40
rt
40
rt
2
24
24
24
24
24
24
12
12
12
82
(41)b
(36)b
(18)b
(35)b
(32)b
(34)b
(20)b
84
10
5
rt
80
a Time monitored by NMR after 2, 8, and 12 h. b Starting alkyne (>50%)
remains after reaction time.
Figure 2. Effect of substituents at the allylic position.
It is to be noted that the degree of selectivity for formation
of the Z-disubstituted double bond depends upon the nature
of the allylic oxygen substituent. In all of the examples, the
use of acetoxy gave very high to exclusive formation of the
Z-alkene as did benzyloxy. However, in all others, varying
amounts of Z,E-dienes along with the Z,Z-dienes were
observed, and in one case (entry 8), the Z,E-isomer was the
major one; in the four cases that gave the poorest Z vs E
selectivity (entries 6, 8, 9, and 11), the effect of the steric
demands of the catalyst was explored (Table 3).
Subjecting a 1:2 alkene-alkyne mixture to 10 mol % of
a Ru cationic complex 38 gave an 82% yield of 5a and 5b.
Surprisingly, the spectral data indicated the structure of the
major product to be the Z,Z-isomer of the conjugated diene
5b. The chemical shifts and coupling constants in the proton
NMR clearly reveal the 1,3-diene. The 11.7 Hz coupling
constant for the disubstituted double bond of the major
isomer confirms it is Z, whereas the 16 Hz coupling for the
minor one establishes it as E. Table 1 summarizes our efforts
to optimize the result. Lowering the alkene/alkyne ratio at
0.8 M to 1:1 (entry 2) and catalyst loading to 5% led to
incomplete reaction even after 24 h (entry 7). Raising the
temperature at these lower loadings does not help. On the
other hand, raising the concentration to 1.6 M allows
lowering the catalyst load (entry 9) or lowering both the ratio
of substrates to 1:1 and catalyst load (entry 10) while
maintaining complete conversion and no loss in yield albeit
with a longer reaction time (12 h instead of 2 h).
For convenience, we chose to explore the scope by using
the conditions of entry 1 as summarized in Table 2. Using
ethyl 2-butynoate (2), we established that placing a branch
at the allylic position (entry 2) or changing the allylic ester
to an ether (entry 3) has no effect on the outcome. Additional
allyl substituents were also explored. For example, entry 6
examines an aryl ether in contrast to an alkyl ether and entries
8 and 9 the use of a Boc carbonate. Entry 11 tests the
chemoselectivity wherein reaction occurs exclusively at the
monosubstituted allyl double bond, an observation reinforc-
Using the Cp*Ru(+) complex 259 under the same condi-
tions did enhance the selectivity in all cases, save one; that
of the reaction of alkynoate 7 with allyl carbonate 13 (see
Table 3, entry 2) wherein no reaction occurred. In two of
the examples (entries 3 and 4), the Z,Z product became the
sole isomer formed. The source of the regio- and geometric
selectivities is not obvious nor simple.
With respect to the regioselectivity of addition to the
alkynoate, preference for formation of the new C-C bond
R to the ester was previously noted.10 For the reactions of
these alkene partners, this preference is exclusive in all cases.
With respect to the regioselectivity of the ꢀ-H insertion, this
heretofore unprecedented selectivity to form the 1,3-diene
exclusively is more difficult to rationalize. The reaction of
o-allylphenol was performed to show the chemoselectivity
with a phenolic OH (Table 2, entry 7). This substrate, being
unfunctionalized at the allylic position, gave 1,4-diene
product as expected.11 We previously noted that but-3-en-
2-ol behaved normally.12 As reported herein, its acetate
(9) Steinmetz, B.; Schenk, W. A. Organometallics 1999, 18, 943.
(10) (a) Trost, B. M.; Muller, T. J. J. J. Am. Chem. Soc. 1994, 116,
4985. (b) Trost, B. M.; Muller, T. J. J.; Martinez, J. J. Am. Chem. Soc.
1995, 117, 1888. (c) Trost, B. M.; Calkins, T. L. Tetrahedron Lett. 1995,
36, 6021.
(5) Trost, B. M.; Frederiksen, U.; Rudd, M. T Angew. Chem., Int. Ed.
2005, 44, 6630.
(6) Mitsudo, T.; Zhang, S. M.; Nagao, M.; Watanabe, Y. J. Chem. Soc.,
Chem. Commun. 1991, 598.
(11) That reporting o-allylphenol as a substrate is irrelevant, this
substrate, being unfunctionalized at the allylic position, gave what it should
have.
(7) Trost, B. M.; Fraisse, P.; Ball, Z. T. Angew. Chem., Int. Ed. 2002,
41, 1059.
(8) Trost, B. M.; Older, C. M. Organometallics 2002, 21, 2544.
(12) Trost, B. M.; Toste, D. Tetrahedron Lett. 1999, 40, 7739.
1072
Org. Lett., Vol. 11, No. 5, 2009

-
Table 2. Alkynoate-Alkene Coupling Catalyzed by CpRu(CH3CN)3+PF6 (3)
a The (Z/E) geometric isomer 2.5%. b Ratio (Z,Z)/(Z,E). All reactions were run using a 1:2 alkyne/alkene in acetone at room tempererature with 10% of
3. c The o-allylphenol, unfunctionalized at the allylic position, gave 1,4-diene product as expected.
completely changes this regioselectivity. We did note in our
Pd-catalyzed alkene-alkyne coupling that the presence of
an allylic electronegative oxygen substituent did favor 1,3-
over 1,4-diene formation.13 This work indicates that trend
also applies to the Ru-catalyzed alkene-alkyne coupling, at
least for certain types of oxygen substituents such as those
bearing acyl groups. While it has been previously noted that
an endocyclic ꢀ-H insertion can occur in a ruthenacyclo-
pentene6 even though it looks to be geometrically challenged,
why it would favor formation of a Z alkene is certainly obtuse
(13) Trost, B. M.; Chung, J. I. L. J. Am. Chem. Soc. 1985, 107, 4586.
Org. Lett., Vol. 11, No. 5, 2009
1073

Table 3. Cross-Coupling of Alkynoates and Alkenes by
-
Cp*Ru(CH3CN)3+PF6 (25)
entry alkyne time (h) alkene product isolated yield (%)
1
2
3
4
6
7
2
2
8
11
13
13
14
19
21
22
24
(6:1)b 5
s.m.
60a
8c
8
8
56a
a Only the 2Z,4Z-isomer was observed in this case. b Recovered starting
material was recovered in 24% yield. c Starting alkyne (>70%) remains
after 8 and 12 h reaction time.
In summary, we report a most unsual regio- and geo-
metrical selective Ru catalyzed alkene-alkyne coupling.
These observations clearly suggest that the mechanism of
the alkene-alkyne coupling is more complex than previously
realized. As such, it promises to provide new and unexpected
pathways as the scope of this reactivity mode continues to
be explored.
at best. A possible explanation invoking the involvement of
a bicycloruthenacyclopentene as depicted in A (eq 3) may
account for it.14 By slowing the exo ꢀ-H insertion process
due to the presence of the OR3 group, conversion of the
monocycle to the bicycle A now creates a more geometrically
favorable endocyclic ꢀ-hydrogen insertion. Placing the
R3OCH2 group exocyclic as in A then favors the Z-alkene
as depicted in eq 3. At present, this rationalization is tenuous
at best. The lack of any detectable amount of 1,4-diene
products combined with the preferential formation of the Z,Z-
isomer makes a pathway invoking isomerization of an
initially formed 1,4-diene highly unlikely.
Acknowledgment. We thank the National Science Foun-
dation for generous support of our programs and the
University of Granada for a fellowship to A.M.R. We thank
the generosity of Umicore for a gift of the ruthenium salts
that are used to form the Ru complexes.
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(14) The isomerization of vinylmetal complexes to metallacyclopropenes
has been previously proposed. See: Tanke, R. S.; Crabtree, R. H J. Am.
Chem. Soc. 1990, 112, 7984. D’Alliessi, D. G.; Faller, J. W Organometallics
2002, 21, 1743.
OL8028324
1074
Org. Lett., Vol. 11, No. 5, 2009