Mechanistic study of nickel-catalyzed ynal reductive cyclizations through kinetic analysis

Article abstract of DOI:10.1021/ja200867d

The mechanism of nickel-catalyzed, silane-mediated reductive cyclization of ynals has been evaluated. The cyclizations are first-order in [Ni] and [ynal] and zeroth-order in [silane]. These results, in combination with the lack of rapid silane consumption

Full text of DOI:10.1021/ja200867d

COMMUNICATION
pubs.acs.org/JACS
Mechanistic Study of Nickel-Catalyzed Ynal Reductive Cyclizations
through Kinetic Analysis
Ryan D. Baxter and John Montgomery*
Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109-1055, United States
S Supporting Information
b
analyses, including regiochemical, stereochemical, and crossover-
2
d,5
ABSTRACT: The mechanism of nickel-catalyzed, silane-
mediated reductive cyclization of ynals has been evaluated.
The cyclizations are first-order in [Ni] and [ynal] and
zeroth-order in [silane]. These results, in combination with
the lack of rapid silane consumption upon reaction initia-
tion, are inconsistent with mechanisms involving reaction
initiation by oxidative addition of Ni(0) to the silane. Silane
consumption occurs only when both the alkyne and alde-
hyde are present. Mechanisms involving rate-determining
oxidative cyclization to a metallacycle followed by rapid
reaction with the silane are consistent with the data
obtained.
labeling probes.
Despite the insights provided by the above
studies, no quantitative rate studies have been reported, and
considerable ambiguity remains in the operative pathway of
silane-mediated reductive couplings. The ambiguity arises
from the facility of alternate pathways involving the genera-
tion of nickel hydride intermediates that precede CÀC bond
formation.
The facility of SiÀH oxidative addition to Ni(0), likely
involved in Ni(0)-mediated hydrosilylation pathways for either
6
carbonyl or alkyne substrates, has often been proposed in
7
various classes of nickel-catalyzed reductive couplings. The
oxidative addition of silane to a Ni(0) phosphine complex could
afford silyl nickel hydride species 4 that then undergoes inser-
tions of the aldehyde and alkyne (Scheme 2). Different possibi-
lities for the timing of the sequential alkyne and aldehyde
insertions through hydrometalation and/or silylmetalation path-
he nickel-catalyzed reductive and alkylative coupling of two
π-components has been widely studied, with processes
8
ways are depicted with intermediates 5a, 5b, and 5c. Alter-
T
natively, complexation of the aldehyde or alkyne to Ni(0)
could facilitate addition of the silane to nickel prior to CÀC
bond formation (Scheme 2, intermediate 6 or 7). An addition
process of this type would then be followed by insertion of the
remaining π-component. The metallacycle mechanism where
silane σ-bond metathesis occurs as the last step (Scheme 1)
differs fundamentally from mechanisms that involve silane
addition at an early stage that preceeds CÀC bond formation
using aldehydes, ketones, enones, imines, alkynes, dienes, and
1
allenes as the most commonly employed π-components. A
variety of reducing agents, including silanes, organozincs, orga-
noboranes, organoaluminums, and alkenyl zirconium reagents,
have been employed. Of these procedures, the reductive coupling2
of aldehydes and alkynes is one of the most extensively studied.
Numerous variants have appeared, and it is clear that the intra- or
intermolecular nature of the process, the reducing agent in-
volved, and the ligand structure employed must all be carefully
matched depending on the specific application desired. The
formation of a nickel metallacycle intermediate 1 derived from
oxidative cyclization of Ni(0) with the aldehyde and alkyne is a
central feature of the most commonly proposed mechanisms
9
(Scheme 2). The rational design of improved procedures,
more effective catalysts, and enantioselective processes re-
quires an understanding of the key mechanistic steps, in
particular the central involvement of metallacycle or nickel
hydride intermediates.
Complexities in distinguishing these mechanistic pathways
(
Schemes 1 and 2) have persisted throughout the development of
(
Scheme 1). Cleavage of the NiÀO bond of this intermediate by
1,2,7,10
numerous classes of nickel-catalyzed reductive couplings.
To
the reducing agent then leads to product, typically with either
silanes or boranes in reductive variants (H-atom installation,
product 2) or with organozincs in alkylative variants (alkyl group
installation, product 3).
address the long-standing questions of reaction mechanism and the
role of silane in nickel-catalyzed, silane-mediated reductive cou-
plings, this Communication describes a detailed rate analysis of a
representative intramolecular coupling and the direct observation of
silane depletion using various catalystÀsubstrate combinations. The
involvement of metallacyclic intermediates (Scheme 1) is consistent
with the data obtained, and evidence against mechanisms that
involve initial addition of a nickel species to the silane
The generation of a dimeric form of metallacycle 1 was
documented by Ogoshi using benzaldehyde and 2-butyne as
3
substrates. This species was observed to undergo conversion to
an alkylative cyclization product 3 upon exposure to dimethyl-
zinc, although the rate of the organozinc addition was much
slower than the overall rate of a typical catalytic process. For the
(
Scheme 2) is provided.
2c
A rate study of the cyclization of ynal 8 with catalytic
Et B-based procedure, a detailed computational study by Houk
3
Ni(COD) /PCy was carried out under synthetically relevant
2
3
and Jamison found that a metallacycle-based mechanism, fol4-
lowed by NiÀO bond cleavage by the borane, was operative.
A number of additional studies have provided mechanistic
insights by examining various product-oriented mechanistic
Received: January 28, 2011
Published: March 28, 2011
r 2011 American Chemical Society
5728
dx.doi.org/10.1021/ja200867d
|J. Am. Chem. Soc. 2011, 133, 5728–5731

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. Nickel-Catalyzed AldehydeÀAlkyne Couplings
Scheme 2. Mechanisms Initiated by Oxidative Addition to
Et SiH
3
Figure 1. Determination of kinetic order of reaction components. (Top
left) Linear dependence of initial reaction rate on [8]. (Top right) Linear
dependence of initial reaction rate on [Ni(COD) ]. (Bottom) Zeroth-
2
order dependence of initial reaction rate on [Et SiH]. Reported values
3
are given as the average of three or more experiments, with error bars
representing the standard deviation between individual runs.
Scheme 3. Kinetic Isotope Effects and Crossover
Experiments
conditions using the method of initial rates in the production of 9
(eq 1).
crossover, in contrast to the PBu -promoted process origin-
ally reported.
3
In situ IR monitoring was used to measure reaction progressions
in real-time. Due to the fast rate of intramolecular coupling,
kinetic analyses were performed at À25 °C. A clear first-order
dependence of reaction rate on [ynal 8] was observed across all
concentrations studied (Figure 1). The reaction rate was also
2d,12
The zeroth-order dependence on Et SiH in ynal reductive
3
cyclizations and lack of kinetic isotope effect are inconsistent with
rate-determining oxidative addition of Ni(0) to the SiÀH bond
(
Scheme 2). For a mechanism involving the production of nickel
shown to have a linear dependence on [Ni(COD) ] between 10
2
hydride 4 to be consistent with the kinetic order studies, a fast
and 25 mol % catalyst loading, yielding catalyst turnover fre-
reaction between a PCy adduct of Ni(COD) with Et SiH must
À1
3
2
3
quencies of 13.8À32.3 h . These results could be consistent
occur, followed by a slow subsequent insertion step, in order for
the first-order dependence on [Ni], first-order dependence on
with either mechanism described above (Schemes 1 and 2).
However, no significant change in reaction rate was observed
[
ynal], and zeroth-order dependence on [Et SiH] to be ob-
3
with varying [Et SiH] as silane stoichiometry was varied from 1.0
3
served. To examine this possibility, 1.0 equiv of the catalyst
derived from Ni(COD) and PCy was added to a solution of
to 4.0 equiv (Figure 1).
2
3
To further distinguish between mechanisms, kinetic isotope
effect studies and crossover experiments were undertaken to gain
insight into the SiÀH bond cleavage process (Scheme 3).
Et SiH with in situ IR monitoring, and no consumption or
3
À1
change of the silicon hydride stretch (2100 cm ) was observed
on the time scale and temperature of productive catalytic
reactions (Figure 2a). As noted above (Scheme 2, intermediate
6 or 7), rate-determining complexation of either aldehyde or
alkyne to Ni(0) could be followed by rapid consumption of
Reductive cyclization of 8 in the presence of excess Et SiH
3
and Et SiD yielded no significant kinetic isotope effect (k /
3
H
kD = 1.00), consistent with post-rate-limiting participation
1
1
13
of the silane reducing agent. In crossover experiments
silane, and thus be consistent with the rate data. To examine
employing excess Et SiD and Pr SiH, ynal 8 was converted
this possibility, a solution of 1.0 equiv of the same catalyst was
3
3
to products 9 and 10 with approximately 10% crossover.
injected to a THF solution of Et SiH and hydrocinnamaldehyde
3
Thus, the primary mechanistic pathway proceeds without
at À25 °C, and no consumption of silane was observed as judged
5
729
dx.doi.org/10.1021/ja200867d |J. Am. Chem. Soc. 2011, 133, 5728–5731

Journal of the American Chemical Society
COMMUNICATION
’
ASSOCIATED CONTENT
Supporting Information.
S
Full experimental details,
b
NMR spectra, and in situ IR kinetics data. This material is
available free of charge via the Internet at http://pubs.acs.org.
’
AUTHOR INFORMATION
Corresponding Author
jmontg@umich.edu
’
ACKNOWLEDGMENT
The authors acknowledge receipt of NIH grant GM57014
(
including GM57014-13S2) in support of this work. R.D.B.
acknowledges receipt of a U.S. DOE GAANN Fellowship
P200A060119). We are especially grateful to Profs. Anne J.
(
McNeil and Melanie S. Sanford and their research groups for
helpful discussions. J.M. thanks the faculty of the Institute of
Chemical Research of Catalonia (ICIQ) for hospitality during a
sabbatical stay.
Figure 2. Experiments tracking silane depletion, with monitoring of
À1
silane IR stretch at 2100 cm . (a, top left) Et
3
SiH (1.0 equiv) at À25
(1.0 equiv) and PCy (2.0 equiv).
b, top right) Et SiH (1.0 equiv) and hydrocinnamaldehyde (1.0
°
(
C, then a mixture of Ni(COD)
2
3
3
equiv) at À25 °C, then a mixture of Ni(COD) (1.0 equiv) and PCy
’ REFERENCES
2
3
(
(
2.0 equiv). (c, bottom left) Et
3
SiH (1.0 equiv) and phenyl propyne
(1.0 equiv) and
(1) (a) Montgomery, J. Angew. Chem., Int. Ed. 2004, 43, 3890. (b)
1.0 equiv) at À25 °C, then a mixture of Ni(COD)
2
Montgomery, J.; Sormunen, G. J. Top. Curr. Chem. 2007, 279, 1.(c)
Malik, H. A.; Baxter, R. D.; Montgomery, J. “. In Catalysis Without
Precious Metals; Bullock, R. M., Ed.; Wiley-VCH: Weinheim, Germany,
PCy (2.0 equiv). (d, bottom right) Et SiH (1.0 equiv) and ynal 8 (1.0
3
3
equiv) at À25 °C, then a mixture of Ni(COD)
2 3
(1.0 equiv) and PCy
(
2.0 equiv).
2010; p 181. (d) Moslin, R. M.; Miller-Moslin, K.; Jamison, T. F. Chem.
Commun. 2007, 4441. (e) Ikeda, S. Angew. Chem. Int. Ed. 2003, 42, 5120.
For related processes with other metals, see:(f) Skukcas, E.; Ngai, M.-Y.;
Komanduri, V.; Krische, M. J. Acc. Chem. Res. 2007, 40, 1394. (g)
Jeganmohan, M.; Cheng, C. H. Chem. Eur. J. 2008, 14, 10876. (h)
Reichard, H. A.; McLaughlin, M.; Chen, M. Z.; Micalizio, G. C. Eur. J.
Org. Chem. 2010, 391.
1
4
by in situ IR analysis (Figure 2b). Similary, injection of 1.0
equiv of the same catalyst to a solution of phenyl propyne and
Et SiH at À25 °C resulted in no silane depletion (Figure 2c).
3
However, adding 1.0 equiv of the same catalyst to a solution of
ynal 8 and Et SiH at À25 °C led to productive formation of
3
(2) (a) Oblinger, E.; Montgomery, J. J. Am. Chem. Soc. 1997,
product 9 with steady depletion of silane (Figure 2d).
119, 9065. (b) Tang, X.-Q.; Montgomery, J. J. Am. Chem. Soc. 1999,
121, 6098. (c) Huang, W. S.; Chan, J.; Jamison, T. F. Org. Lett. 2000,
2, 4221. (d) Mahandru, G. M.; Liu, G.; Montgomery, J. J. Am. Chem. Soc.
2004, 126, 3698. (e) Yang, Y.; Zhu, S. F.; Zhou, C. Y.; Zhou, Q. L. J. Am.
These experiments, coupled with the rate analysis and lack of
silane kinetic isotope effect, are consistent with the metallacycle
mechanism (Scheme 1) and provide strong evidence against
mechanisms that involve addition of a Ni(0) species to silane
Scheme 2). Additionally, kinetically viable mechanisms that
involve catalyst modification via the conversion of Ni(0) to a
reactive Ni(II) hydride upon reaction initiation would involve
rapid silane consumption at the level of nickel catalyst loading.
However, rapid silane consumption does not precede the begin-
ning of ynal consumption or product formation. The interme-
diacy of metallacycle 1 followed by rapid reaction with Et SiH, is
Chem. Soc. 2008, 130, 14052.
(
(3) Ogoshi, S.; Arai, T.; Ohashi, M.; Kurosawa, H. Chem. Commun.
2
008, 1347.
(4) McCarren, P. R.; Liu, P.; Cheong, P. H.-Y.; Jamison, T. F.; Houk,
K. N. J. Am. Chem. Soc. 2009, 131, 6654.
(5) (a) Miller, K. M.; Jamison, T. F. J. Am. Chem. Soc. 2004,
15
126, 15342. (b) Moslin, R. A.; Miller, K. M.; Jamison, T. F. Tetrahedron
2
006, 62, 7598. (c) Malik, H. A.; Sormunen, G. J.; Montgomery, J. J. Am.
3
Chem. Soc. 2010, 132, 6304.
6) (a) Chaulagain, M. R.; Mahandru, G. M.; Montgomery, J.
consistent with these observations. Metallacycle 1 is likely
formed by oxidative cyclization of a Ni(0)Àynal complex
(
16
Tetrahedron 2006, 62, 7560. (b) Buchan, Z. A.; Bader, S. J.; Montgom-
ery, J. Angew. Chem. Int. Ed. 2009, 48, 4840.
following slow or reversible complexation of substrate.
In summary, the kinetic behavior of a nickel-catalyzed silane-
mediated reductive cyclization of ynals has been evaluated. This
report represents the first such study across the many classes of
related nickel-catalyzed reductive couplings that have been
reported. The rate studies are consistent with a mechanism
involving rate-determining oxidative cyclization to a metallacyc-
lic intermediate, followed by rapid silane-mediated conversion to
the protected allylic alcohol product. A number of mechanisms
involving oxidative addition of silane to nickel can be ruled out on
the basis of this analysis. Future work will be directed toward fully
elucidating the nature of the steps that follow the rate-determin-
ing oxidative cyclization, as well as evaluating the generality of
these results in intermolecular couplings and with other ligand
systems.
(
7) (a) Tamao, K.; Kobayashi, K.; Ito, Y. J. Am. Chem. Soc. 1989,
1
11, 6478. (b) Sato, Y.; Takimoto, M.; Hayashi, K.; Katsuhara, T.;
Takagi, K.; Mori, M. J. Am. Chem. Soc. 1994, 116, 9771. (c) Sato, Y.;
Takimoto, M.; Mori, M. J. Am. Chem. Soc. 2000, 122, 1624. (d) For a
recent demonstration, see:Zell, T.; Schaub, T.; Radacki, K.; Radius, U.
Dalton Trans. 2011, 40, 1852.
(
8) Hydrosilylation of ynals with Rh hydrosilylation catalysts pro-
vides the opposite regiochemistry for silane addition, producing alke-
nylsilane products. See:Ojima, I.; Tzamarioudaki, M.; Tsai, C. Y. J. Am.
Chem. Soc. 1994, 116, 3643.
(
9) In addition to the possibilities described in the text, alteration of
the precatalyst oxidation state during reaction initiation would provide
additional available mechanistic pathways. For example, the Ni(II)-
catalyzed hydrosilylation of carbonyl compounds proceeds by nickel
hydride generation by σ-bond metathesis of silane with a nickel
5
730
dx.doi.org/10.1021/ja200867d |J. Am. Chem. Soc. 2011, 133, 5728–5731

Journal of the American Chemical Society
COMMUNICATION
alkoxide. See:Tran, B. L.; Pink, M.; Mindiola, D. J. Organometallics 2009,
28, 2234.
(
10) For related classes of silane-mediated nickel-catalyzed reductive
couplings, see ref 7 and the following:(a) Sato, Y.; Sawaki, R.; Mori, M.
Organometallics 2001, 20, 5510–5512. (b) Ng, S. S.; Jamison, T. F. J. Am.
Chem. Soc. 2005, 127, 7320. (c) Saito, N.; Sugimura, Y.; Sato, Y. Org. Lett.
2
010, 12, 3494.
(
11) The Ni(COD) /PCy catalyst system studied in this Commu-
2
3
H D
nication proceeded with an inverse kinetic isotope effect (k /k = 0.75)
for the hydrosilylation of cinnamaldehyde at 45 °C in THF (heating at
this temperature was required to achieve good conversion). For discus-
sions of kinetic isotope effects observed in hydrosilylation reactions, see:
(a) Parks, D. J.; Blackwell, J. M.; Piers, W. E. J. Org. Chem. 2000, 65, 3090.
(b) Nolin, K. A.; Krumper, J. R.; Pluth, M. D.; Bergman, R. G.; Toste,
F. D. J. Am. Chem. Soc. 2007, 129, 14684. (c) Schneider, N.; Finger, M.;
Haferkemper, C.; Bellemin-Laponnaz, S.; Hofman, P.; Gade, L. H.
Chem. Eur. J. 2009, 15, 11515.
(
12) A study of the post-rate-limiting steps and origin of crossover is
in progress and will be reported elsewhere.
13) For studies of carbonyl hydrosilylation processes that describe
(
various mechanisms for silane consumption, see ref 11 and the follow-
ing:Ojima, I.; Kogure, T. Organometallics 1982, 1, 1390.
(
14) The consumption of Et
maldehyde using a Ni(COD) /PCy
but requires heating to 45 °C for several hours.
15) Nickel hydride-catalyzed pathways typically involve electrophi-
lic Ni(II) species. See ref 9 and the following: RajanBabu, T. V. Synlett
009, 853–885.
16) The first-order dependence on ynal illustrates that the Ni(0)
ynal complex is not the sole catalyst resting state.
3
SiH via the hydrosilylation of cinna-
2
3
catalyst system in THF is observed
(
2
(
5
731
dx.doi.org/10.1021/ja200867d |J. Am. Chem. Soc. 2011, 133, 5728–5731