Ligand exchange as the first irreversible step in the nickel-catalyzed cross-coupling reaction of grignard reagents

Article abstract of DOI:10.1021/ja807000a

Mechanistic studies of the Ni-catalyzed cross-coupling reaction of Grignard reagents through analysis of kinetic isotope effects and theoretical calculations indicated that the product-to-substrate ligand exchange process is the first irreversible step and affects the turnover efficiency and selectivity of the reaction. On the other hand, the oxidative addition step is the first irreversible step in Pd catalysis. This finding has useful implications for the development of efficient Ni catalysis and also illustrates the importance of the catalyst turnover step that has so far received less attention than subsequent catalytic steps. Copyright

Full text of DOI:10.1021/ja807000a

Published on Web 10/25/2008
Ligand Exchange as the First Irreversible Step in the Nickel-Catalyzed
Cross-Coupling Reaction of Grignard Reagents
Naohiko Yoshikai, Hirokazu Matsuda, and Eiichi Nakamura*
Department of Chemistry, The UniVersity of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan
Received September 4, 2008; E-mail: nakamura@chem.s.u-tokyo.ac.jp
Ni-catalyzed cross-coupling of an aryl or alkenyl halide with a
Scheme 1
Grignard reagent, known as the Kumada-Tamao-Corriu (KTC)
1
reaction, paved the way for later development of the currently more
2
popular and selective Pd catalysis. Oxidative addition, transmeta-
lation, and reductive elimination have been studied widely as
important elementary steps of such reactions, as outlined in Scheme
3
1
. We propose herein that the ligand exchange step preceding the
whole catalytic cycle needs more careful attention to understand
these reactions. Studies of the Ni-catalyzed KTC reaction by
analysis of kinetic isotope effects (KIEs) and theoretical calculations
indicated that the ligand exchange process is the first irreversible
step (FIS) and may affect the turnover efficiency and the selectivity
of the reaction. On the other hand, the oxidative addition step is
the FIS in Pd catalysis. This finding not only has useful implications
about group 10 metal catalysis but also illustrates the importance
of the catalyst turnover step, which has so far received less attention
than the subsequent catalytic steps.
The KIE experiments tell us little about the nature of the C-X
bond cleavage that occurs in the Ni(dppp) (or dppe) catalysis after
the FIS. We addressed this issue through experimental analysis of
Pd catalysis and theoretical studies. Pd(0) is a weaker d-electron
9
donor, and the haloarene-Pd(0) complex may become less stable
than the Ni(0) complexes; that is, the oxidative addition step that
follows the ligand exchange step may become the FIS.
10
The Pd(dppf)-catalyzed coupling of o-bromotoluene with
1
2
13
4
First, we measured the C/ C KIE in the cross-coupling
reaction of 1-halo-2-methylbenzene (o-halotoluene, X ) Cl, Br,
or I, eq 1) or 1-bromocyclooctene (eq 2) with phenyl- or al-
kylmagnesium bromide catalyzed by a Ni- or Pd-bisphosphine
complex. The KIE data provide information on the FIS of the
C
6
5 8
H MgBr or n-C H17MgBr (Figure 2, data a and b) exhibited large
KIE values at C1 (1.031, 1.024) and small values at C6 (1.007,
11
1
.009). The reaction of 1-bromocyclooctene (data c) also showed
unequal KIEs at C1 (1.023) and C2 (1.014). Similar results were
obtained previously for the oxidative addition reaction of (CH
3 2
) -
5
catalytic cycle. If there are any elements of π-complexation
12
CuLi·LiI (data d). With Pd(0) being moderately donative of
d-electrons, similar to Cu(I), we can therefore conclude that the
involved in the FIS, we shall observe a dissymmetric distribution
of KIE over the aromatic carbon atoms.
13
FIS in the Pd catalysis is the C-X bond cleavage step.
The information given by the experiments being rather limited,
we performed DFT calculations on the interactions of model Ni-
and Pd-bisphosphine complexes with o-chloro- and o-bromotoluene,
1
4
respectively (Scheme 2), and we found, not unexpectedly, that
both reactions share the same pathway, i.e., formation of a series
6
The KIE data for Ni(dppp) catalysis of o-chloro- or bromo-
toluene with C
6 5 4 9
H MgBr (Figure 1, data a and b) or n-C H MgBr
(
data c) exhibit small KIE values (1.003-1.009) distributed only
7
12 13
on the less hindered side of the benzene ring. The data strongly
suggest that the FIS of the reaction is the π-complexation of the
Ni catalyst on the (less hindered) π-face of the haloarene. This
interpretation is supported by the calculated equilibrium isotope
Figure 1. C/ C KIE values for Ni-catalyzed KTC reactions. Standard
deviations in the last digit are shown in parentheses. The asterisked carbon
atoms were taken as references for each measurement (KIE ) 1 was
assumed). dppp ) 1,3-bis(diphenylphosphino)propane. dppe ) 1,2-bis-
(
diphenylphosphino)ethane.
8
effect (EIE). In other words, once a Ni/haloarene π-complex forms
through ligand exchange, it does not dissociate and proceeds quickly
to the oxidative addition step in an intramolecular manner. We can
ascribe this observation to the d-electron donative property of the
9
Ni(0) catalytic species.
6
Similarly, the KIE data for the Ni(dppe)-catalyzed coupling of
4 9
n-C H MgBr with 1-bromocyclooctene (data d) exhibited a KIE
1
2
13
Figure 2.
C/ C KIE values for Pd-catalyzed KTC reactions and
of equal magnitude at two olefinic carbon atoms (1.015 at C1, 1.014
at C2). Thus, the π-complexation is also the FIS in the alkenylation
reaction.
organocopper-mediated substitution reaction. See Figure 1 caption for
standard deviations and the asterisked carbon atoms. dppf ) 1,1′-
bis(diphenylphosphino)ferrocene. Data d were taken from ref 12.
1
5258 9 J. AM. CHEM. SOC. 2008, 130, 15258–15259
10.1021/ja807000a CCC: $40.75  2008 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2
pensity of the Ni catalyst to retain π-complexation with the
haloarene substrate is beneficial for some reactions, such as the
20
synthesis of oligoarenes via polycondensation, but not for others,
6
such as the chemoselective cross-coupling of polyhaloarenes. It
is possible that accelerated π-complexation/decomplexation is a
reason for the high activity of the Ni/hydroxyphosphine catalyst
2
1
that we reported recently. We suspect that sluggish catalyst
2
2
turnover may be ubiquitous in catalysis, but its significance is
underappreciated.
2
of η -complexes (CP1-3; as well three others illustrated in Scheme
Acknowledgment. We thank MEXT/JSPS for financial support
(KAKENHI No. 18105004 (S) for E.N., No. 20037012 (Chemistry
of Concerto Catalysis) for N.Y., and the Global COE Program for
Chemistry Innovation) and RCCS at Okazaki for computational
time.
2
S1), isomerization between these η -complexes, and C-X bond
cleavage (CP3 to PD via TS).
15,16
However, the Ni and Pd catalyses
have different energetics (Figure 3). For the π-complexation
between the Ni(0) complex and o-chloro- or bromotoluene, forma-
tion of CP1-3 was highly exothermic (17-19 kcal/mol), but much
less so for the interaction between Pd(0) and o-bromotoluene (6-8
kcal/mol). This result predicts that the Ni(0) species in solution would
form stable π-complexes with haloarenes, but the Pd(0) species would
not. This is supported by the fact that some Ni(0)/arene η -complexes,
including a fluoroarene complex, are isolable.
The activation energy from CP3 to the oxidative addition product
is very small for Ni catalysis (∆E ) 9.4 and 5.1 kcal/mol for
Supporting Information Available: Details of experiments and
computation. This material is available free of charge via the Internet
at http://pubs.acs.org.
2
References
17
(
1) (a) Tamao, K.; Sumitani, K.; Kumada, M. J. Am. Chem. Soc. 1972, 94,
4
374. (b) Corriu, R. J. P.; Masse, J. P. Chem. Commun. 1972, 144.
‡
(2) Metal-catalyzed Cross-coupling Reactions, 2nd ed.; de Meijere, A.;
Diederich, F., Eds.; Wiley-VCH: New York, 2004.
chloro- and bromotoluene, respectively; 7.5 kcal/mol for bromo-
(
3) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and
Applications of Organotransition Metal Chemistry, 2nd ed.; Wiley: New
York, 1987.
1
8
toluene in Pd catalysis ). For both the Ni and Pd reactions, the
oxidative addition transition state (TS) retains the character of the
π-complex CP3. This structural property agrees with the unequal
distribution of KIE values for the Pd(dppf) catalysis (Figure 2,
data a and b). We draw the same conclusion for the model
reactions of the Ni and Pd complexes with 1-bromocyclooctene
(
(
(
4) Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117, 9357.
5) Vo, L. K.; Singleton, D. A. Org. Lett. 2004, 6, 2469.
6) Tamao, K.; Sumitani, K.; Kiso, Y.; Zembayashi, M.; Fujioka, A.; Kodama,
S.; Nakajima, I.; Minato, A.; Kumada, M. Bull. Chem. Soc. Jpn. 1976, 49,
1
9
1
958.
(7) The KIE data for o-iodotoluene were similar yet may be uninformative
because the reaction gave the desired product in only 42% yield and toluene
in 36% yield at 88% conversion.
(8) The experimental data are of similar magnitude to the calculated EIE for
the π-complexation.
(
see Supporting Information, SI).
We also discuss briefly the thermodynamics and kinetics of the
migration of Ni(0) from the product to the substrate (Figure 4.).
The least sterically hindered product π-complex (CP1′) was
calculated to be less stable than CP1 by 1.3 kcal/mol and other
more hindered isomers less stable by 1-4 kcal/mol than the
respective haloarene complexes. We also consider that the ligand
exchange that likely accompanies a high entropy-cost process would
take place with a sizable activation barrier (Figure 4.).
In summary, the KIE data and theoretical study showed that the
early stages of the Ni- and Pd-catalyzed KTC reactions share the
same pathway but show different kinetic profiles. With Ni(0) being
a strong electron donor, the FIS of Ni catalysis is the complexation
of the Ni atom on the π-face of the haloarene; i.e., the Ni catalyst
undergoes oxidative addition immediately after ligand exchange
from the product to the haloarene. We can surmise that the pro-
(
9) Massera, C.; Frenking, G. Organometallics 2003, 22, 2758.
(
10) Hayashi, T.; Konishi, M.; Kobori, Y.; Kumada, M.; Higuchi, T.; Hirotsu,
K. J. Am. Chem. Soc. 1984, 106, 158.
(
11) o-Chlorotoluene was unreactive for Pd catalysis. o-Iodotoluene showed the
same trend in KIE with a smaller magnitude (1.011 at C1, 1.004 at C6).
12) Yoshikai, N.; Nakamura, E. J. Am. Chem. Soc. 2004, 126, 12264.
13) Nakamura, E.; Mori, S. Angew. Chem., Int. Ed. 2000, 39, 3750.
14) DFT calculations were performed by the B3LYP method using 6-31G(d)-
and 6-311+G(d,p)-level basis sets for geometry optimization and single-
point energy calculation (Figure 3), respectively. Reaction of the Ni(0)
complex with o-bromotoluene was also studied. See SI for details.
15) Zenkina, O. V.; Karton, A.; Freeman, D.; Shimon, L. J. W.; Martin, J. M. L.;
van der Boom, M. E. Inorg. Chem. 2008, 47, 5114, and references cited
therein.
(
(
(
(
2
(
(
(
16) The entire reaction pathway involving other possible (but less stable) η -
complexes is given in Scheme S1 in the SI.
17) Bach, I.; P o¨ rschke, K.-R.; Goddard, R.; Kopiske, C.; Kr u¨ ger, C.; Rufinska,
A.; Seevogel, K. Organometallics 1996, 15, 4959.
18) Pd(0) catalysis of chlorotoluene requires higher energy (∆E ) 12.3 kcal/
‡
mol), and experimentally, chloroarenes do not readily participate in the
Pd-catalyzed reactions.
(
19) The experimental data agree with the calculated KIE for the transition state
of C-Br bond cleavage.
Figure 3. Potential energy diagram (kcal/mol, B3LYP/6-311+G(d,p)-SDD-
TZVP//B3LYP/6-31G(d)-LANL2DZ-SVP) for the reaction shown in Scheme
2
. See Supporting Information for details of the basis sets.
(
20) (a) Miyakoshi, R.; Yokoyama, A.; Yokozawa, T. J. Am. Chem. Soc. 2005,
1
27, 17542. (b) Miyakoshi, R.; Shimono, K.; Yokoyama, A.; Yokozawa,
T. J. Am. Chem. Soc. 2006, 128, 16012.
(
(
21) Yoshikai, N.; Mashima, H.; Nakamura, E. J. Am. Chem. Soc. 2005, 127,
1
7978.
22) Ammal, S. C.; Yoshikai, N.; Inada, Y.; Nishibayashi, Y.; Nakamura, E.
J. Am. Chem. Soc. 2005, 127, 9428.
Figure 4.
JA807000A
J. AM. CHEM. SOC. 9 VOL. 130, NO. 46, 2008 15259