Mechanistic implications of nickel-catalyzed reductive coupling of aldehydes and chiral 1,6-enynes

Article abstract of DOI:10.1021/ol052719n

A study of nickel-catalyzed reductive coupling reactions of aldehydes and chiral 1,6-enynes has provided evidence for three distinct mechanistic pathways that govern regioselectivity in this transformation. In the absence of a phosphine additive, high regioselectivity and high diastereoselectivity are obtained as a direct result of coordination of both the alkyne and the olefin to the metal center during the C-C bond-forming step.

Full text of DOI:10.1021/ol052719n

ORGANIC
LETTERS
2006
Vol. 8, No. 3
455-458
Mechanistic Implications of
Nickel-Catalyzed Reductive Coupling of
Aldehydes and Chiral 1,6-Enynes
Ryan M. Moslin and Timothy F. Jamison*
Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
tfj@mit.edu
Received November 9, 2005
ABSTRACT
A study of nickel-catalyzed reductive coupling reactions of aldehydes and chiral 1,6-enynes has provided evidence for three distinct mechanistic
pathways that govern regioselectivity in this transformation. In the absence of phosphine additive, high regioselectivity and high
a
diastereoselectivity are obtained as a direct result of coordination of both the alkyne and the olefin to the metal center during the C
−C
bond-forming step.
Nickel-catalyzed reductive coupling of alkynes and aldehydes
has emerged as an efficient and selective method for the
synthesis of allylic alcohols.1 Enantioselective coupling
reactions have been achieved through the use of chiral
monodentate phosphines,2 and several different classes of
alkynes afford coupling products in excellent regioselectiv-
ity.3
the result of two distinct mechanistic pathways: one involv-
ing an oxametallacycle and one involving alkyne hydro-
metalation.
Scheme 1
Recently, we reported a highly regioselective reductive
coupling of 1,6-enynes and aldehydes (Scheme 1) in which
the sense of regioselectivity was completely reversed upon
the addition of catalytic amounts of tricyclopentylphosphine
(PCyp3).3e We proposed that this remarkable inversion was
(1) Review: Montgomery, J. Angew. Chem., Int. Ed. 2004, 43, 3890-
3908.
(2) (a) Miller, K. M.; Huang, W.-S.; Jamison, T. F. J. Am. Chem. Soc.
2003, 125, 3442-3443. (b) Colby, E. A.; Jamison, T. F. J. Org. Chem.
2003, 68, 156-166.
(3) (a) Oblinger, E.; Montgomery, J. J. Am. Chem. Soc. 1997, 119, 9065-
9066. (b) Huang, W.-S.; Chan, J.; Jamison, T. F. Org. Lett. 2000, 2, 4221-
4223. (c) Mahandru, G. M.; Liu, G.; Montgomery, J. J. Am. Chem. Soc.
2004, 126, 3698-3699. (d) Miller, K. M.; Luanphaisarnnont, T.; Molinaro,
C.; Jamison, T. F. J. Am. Chem. Soc. 2004, 126, 4130-4131. (e) Miller,
K. M.; Jamison, T. F. J. Am. Chem. Soc. 2004, 126, 15342-15343.
Herein we report that further investigation of coupling
reactions involving chiral 1,6-enynes has afforded evidence
to support an alternate theory, one in which both pathways
proceed via a common oxametallacycle, but regioselectivity
10.1021/ol052719n CCC: $33.50
© 2006 American Chemical Society
Published on Web 01/04/2006

is determined by stereospecific ligand substitution at the
nickel center.
The three proposed mechanistic pathways for the reductive
coupling of 1,6-enynes and aldehydes under varying condi-
tions are illustrated in Scheme 2.
by the aldehyde, ultimately leading to regioisomer B by way
of 4. Thus, despite not being bound during the C-C bond
formation, the olefin nevertheless determines regioselectivity.
Finally, we propose that when a smaller ligand such as tri-
n-butylphosphine (PBu3) is employed (type III), two equiva-
lents of phosphine are bound to the metal center displacing
both the olefin tether and L to give 5. In this case,
regioselectivity would not be determined by the olefin, and
a nonselective displacement of either phosphine leads to a
mixture of 6 and 7, which in turn affords a mixture of
regioisomers A and B.
Scheme 2
To test these mechanistic hypotheses and the overriding
assumption of a planar, three-coordinate nickel complex, we
evaluated the effect of a stereogenic center in the olefin
tether. We hypothesized that in the absence of a phosphine
(type I), coordination of the olefin to the metal center should
enhance diastereoselection, while conditions employing
achiral phosphines (types II and III) should lead to lower
diastereoselectivity since the olefin would be dissociated
during the C-C bond-forming step.
Thus, chiral 1,6-enynes 8 and 9 were synthesized and
coupled with isobutyraldehyde under three distinct sets of
catalytic conditions: (I) Ni(cod)2 with no additives; (II) Ni-
(cod)2 + PCyp3; and (III) Ni(cod)2 + PBu3 (Table 1).
Table 1. Coupling Reactions of Chiral 1,6-Enynes
reaction
All pathways involve an approximately planar, three-
coordinate, d8 nickel complex (1) that would be expected to
undergo ligand substitution with retention of stereochemistry
through an associative pathway.4 In all cases, C-C bond
formation is believed to occur through an oxanickellacyclo-
pentene.2,3 Also, in all cases, the third ligand (L) is assumed
to be an olefin5 and, as it is not part of a bidentate chelate,
is considered to be the most weakly bound ligand. Therefore,
in all cases L is the ligand that is displaced from the metal
center in substitution reactions of 1.
entry
enynea
conditionsb products A/Bc
dr Ad dr Bd
1
2
3
4
5
6
8
I
10A, B >95:5
<5:95
95:5
45:55
50:50 45:55
(R ) Et)
II
III
I
II
III
55:45
9
11A, B >95:5 >95:5
<5:95
(R ) t-Bu)
42:58
45:55 42:58
51:49
a Racemic 8 and 9 were employed in this series of reactions. b I: Ni(cod)2
(10 mol %), Et3B (200 mol %). II: reaction conditions I + PCyp3 (20 mol
%). III: reaction conditions I + PBu3 (20 mol %). c Based on isolated yields.
d Determined by H NMR.
1
In the absence of a phosphine ligand (Scheme 2, type I),
ligand substitution places the aldehyde cis to the carbon distal
to the alkene, C(A), and cis to the bound olefin, giving 2.
C-C bond formation occurs at C(A) while the olefin tether
is coordinated to the nickel, resulting in exclusive formation
of regioisomer A. In type II reactions, tricyclopentylphos-
phine (PCyp3) substitutes for L to give 3. PCyp3 is
coordinated more strongly to the metal center than the
tethered alkene, which is thus displaced stereospecifically
As predicted, under type I reaction conditions (no phos-
phine) both enynes gave exclusively regioisomer A (Table
1, entries 1 and 4). In addition, both allylic alcohols were
formed in excellent diastereoselectivity, indicating a strong
influence of the stereogenic center in the tether, despite being
separated from the site of C-C bond formation by five atoms
(1,6-induction).
Conversely, under type II reaction conditions, regioisomer
B is formed exclusively, but diastereoselection is negligible
(entries 2 and 5). Type III reaction conditions are neither
regioselective nor diastereoselective (entries 3 and 6).
(4) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Mill Valley, CA, 1987; pp 241-244.
(5) Ethylene, 1,5-cyclooctadiene, or another equivalent of 8 or 9 are all
possible.
456
Org. Lett., Vol. 8, No. 3, 2006

Taken together, these experiments strongly support the
notion that, in the absence of phosphine (type I), the alkene
is coordinated to Ni during the C-C bond-forming step and
that, in the presence of phosphine (type II or III), the alkene
is not coordinated to Ni during the C-C bond-forming step.
The high levels of diastereoselectivity afforded by enynes
8 and 9 in the absence of phosphine (Table 1, entries 1 and
4) prompted us to investigate coupling reactions of these
chiral enynes further. To determine the sense of induction
in the formation of regioisomer A, enantiomerically enriched
enyne 8 was prepared (Scheme 3). 1-Penten-3-ol was
Scheme 4
Scheme 3
reduced but nevertheless high diastereoselectivity (Scheme
5). The sense of induction, determined to be R using the
Scheme 5
resolved using a Sharpless asymmetric epoxidation,6,7 and
Williamson ether synthesis using the (S) enantiomer afforded
enyne 8.
Nickel-catalyzed coupling of (S)-8 and i-PrCHO in the
absence of a phosphine (type I reaction conditions) afforded
10A in >95:5 regioselectivity and 95:5 diastereoselectivity.
Conversion to the corresponding acetate followed by ozo-
nolysis afforded ketone (+)-12. The sign of the specific
rotation of this compound was opposite that of (-)-12
prepared from commercially available (S)-2-hydroxy-3-
methylbutyric acid,8 thus establishing the allylic alcohol
configuration in 10A as R.
To test the possibility that the oxygen atom in the tether
plays a key role in the mode of diastereoinduction (e.g.,
chelation of this oxygen and that of the aldehyde by an
organoboron species), we synthesized a 1,6-enyne (13) in
which the oxygen was replaced with a methylene group
(Scheme 4).9 A highly diastereoselective Myers alkylation,
Swern oxidation, and Wittig olefination afforded 13 in a
straightforward manner.
same sequence of operations shown in Scheme 3, was also
the same as that observed with enynes 8 and 9. Thus, an
oxygen atom and a CH2 group at this position in the tether
have similar (but measurably different) effects in type I
coupling reactions.
At this stage, we wanted to test a critical aspect of the
type II and III mechanisms, that the phosphine is bound to
Ni during the C-C bond-forming step. We reasoned that
since the influence of the chiral center in the tether in these
cases is minimal, any diastereoselectivity induced by a chiral
phosphine could be attributed to the phosphine alone, a result
that would be consistent with phosphine being bound to Ni
as the C-C bond is formed.
To this end, we subjected enyne 8 and isobutyraldehyde
to reductive coupling conditions in the presence of an achiral
or chiral ferrocenyl-containing phosphine (Table 2).2b,10
Nearly equimolar amounts of regioisomers A and B were
obtained in all cases, suggesting that the reaction occurs via
a type III mechanistic pathway (cf. Scheme 2). Both the R
and S phosphine ligands afforded modest diastereoinduction.
These results demonstrate that the enyne stereocenter exerts
little to no influence on the diastereoselectivity and clearly
indicate that phosphine is bound to nickel during the C-C
bond-forming step.
Under type I coupling conditions, enyne 13 gave results
similar to those obtained with the enynes possessing an
ethereal tether between the alkene and the alkyne. Nickel-
catalyzed reductive coupling of 13 and i-PrCHO afforded
allylic alcohol 17 in very high regioselectivity and in slightly
(6) Hill, M. L.; Raphael, R. A. Tetrahedron 1990, 46, 4487-4594.
(7) Kagan, H. B, Stereochemistry; George Thieme: Stuttgart, 1977; Vol.
4, p 224.
(8) Bach, J.; Berenguer, R.; Farra`s, J.; Garcia, J.; Meseguer, J.; Vilarrasa,
J. Tetrahedron: Asymmetry 1995, 6, 2683-2686.
(9) Myers, A. G.; Yang, B. H.; Chen, H.; McKinstry, L.; Kopecky, D.
J.; Gleason, J. L. J. Am. Chem. Soc. 1997, 119, 6496-6511.
(10) Miller, K. M.; Jamison, T. F. Org. Lett. 2005, 7, 3077-3080.
Org. Lett., Vol. 8, No. 3, 2006
457

influenced not only by the presence or absence of a
phosphine additive, but also by the nature of the phosphine
chosen. These results support the hypothesis that nickel-
catalyzed reductive coupling reactions of alkynes and alde-
hydes proceed through an approximately planar, three-
coordinate d8 nickel complex. Finally, the chelation-
controlled, highly diastereoselective transformations in the
absence of phosphine suggest synthetic applications, and the
mechanistic insight gained through this investigation should
facilitate the development of other selective, nickel-catalyzed
transformations.
Table 2. Coupling Reactions of Chiral, Enantiomerically
Enriched 1,6-Enynes with Ferrocenyl-Containing Phosphines
ligand
A/Ba
dr 10A (R/S)b
dr 10Bb,c
Acknowledgment. This work was supported by the
National Institute of General Medical Science (GM-063755).
We thank Dr. Karen M. Miller for assistance in the
preparation of this manuscript and thank Dr. Li Li (MIT
Department of Chemistry Instrumentation Facility) for
obtaining mass spectrometry data. The MIT DCIF is sup-
ported in part by the NSF (CHE-9809061 and DBI-9729592)
and the NIH (1S10RR13886-01).
(R)-18
(S)-18
FcPPh2
48:52
55:45
54:46
30:70
66:34
56:44
28:72
68:32
48:52
a Based on isolated yields. b Configuration of allylic alcohol stereogenic
center. c Relative stereochemistry not determined.
The exact mode of diastereoinduction is unknown. How-
ever, one possibility is that one of the conformations of 1 is
highly favored and reacts with the aldehyde in a highly
stereoselective manner.
In summary, subtle changes to the reaction conditions of
reductive coupling reactions of aldehydes and chiral 1,6-
enynes have a significant impact on both regioselectivity and
diastereoselectivity. At least three distinct mechanistic
pathways are possible in these reactions, and selectivity is
Supporting Information Available: Full experimental
procedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL052719N
458
Org. Lett., Vol. 8, No. 3, 2006