An extremely active catalyst for the Negishi cross-coupling reaction

Article abstract of DOI:10.1021/ja0474493

A new catalyst system for the Pd-catalyzed cross-coupling of organozinc reagents with aryl halides (Negishi coupling) has been developed. This system permits efficient preparation of hindered biaryls (triand tetra-ortho- substituted), functions effectively at low levels of catalyst, and tolerates a wide range of functional groups and heterocyclic substrates. A systematic study of ligand structure was performed and was correlated with catalyst activity.

Full text of DOI:10.1021/ja0474493

Published on Web 09/15/2004
An Extremely Active Catalyst for the Negishi Cross-Coupling
Reaction
Jacqueline E. Milne and Stephen L. Buchwald*
Contribution from the Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
Received April 30, 2004; E-mail: sbuchwal@mit.edu
Abstract: A new catalyst system for the Pd-catalyzed cross-coupling of organozinc reagents with aryl halides
(Negishi coupling) has been developed. This system permits efficient preparation of hindered biaryls (tri-
and tetra-ortho-substituted), functions effectively at low levels of catalyst, and tolerates a wide range of
functional groups and heterocyclic substrates. A systematic study of ligand structure was performed and
was correlated with catalyst activity.
these reagents, difficulties remain.5 Problems include the
frequent need for recrystallization of the arylboronic acid prior
Introduction
The biaryl motif is found in numerous classes of natural
products and pharmaceutically active compounds. Over the past
30 years, Pd-catalyzed cross-coupling reactions have revolution-
ized the way organic synthesis is practiced. The most widely
studied of these cross-coupling reactions involve the formation
of C(sp2)-C(sp2) bonds by the union of aryl or vinyl halides
and organometallic reagents (ArM; M ) B, Sn, Si, Zn, Mg).1
Among the possible nucleophilic partners, organotin (Stille) and
organoboron (Suzuki-Miyaura) reagents have been studied
most widely; organozinc (Negishi) and organomagnesium
(Kumada) reagents have been employed to a lesser extent.
Studies have shown that the choice of supporting ligand on
Pd is crucial in many transformations.1b,2 Pioneering work by
Dai and Fu afforded the first general protocol for performing
Negishi cross-coupling reactions of unactivated and deactivated
aryl chlorides, in which the electron-rich complex Pd[P(t-Bu)3]2
was used as the precatalyst.3 Using a standard set of conditions
(2% Pd[P(t-Bu)3]2, THF/NMP, 100 °C), the synthesis of quite
hindered biaryls was accomplished in excellent yield.
Recently, we reported that the use of commercially available
1 (SPhos) provided a catalyst with unparalleled reactivity in
Suzuki-Miyaura cross-coupling processes. It allowed for the
efficient construction of hindered biaryls, couplings with
exceptionally high turnover numbers, couplings at room tem-
perature and could be employed with a broad substrate scope.4
Despite the widespread use of boronic acids for Suzuki-
Miyaura cross-couplings, and the advantages associated with
to use, their tendency to form varying amounts of boroxines,
and their propensity to undergo competitive protodeboronation
under the reaction conditions used for cross-coupling, as well
as problems involving the preparation and application of
sterically hindered boronic acids. In addition, the presence of
either electron-donating or -withdrawing groups may reduce the
stability of the arylboronic acid. For example, our efforts to
synthesize the electron-rich [2-(N,N-dimethylamino)-6-meth-
oxyphenyl]boronic acid by halogen-lithium exchange of the
corresponding aryl bromide, followed by treatment with B(Oi-
Pr)3, resulted in deboronation upon attempted conversion of the
initially formed intermediate to the boronic acid. A potential
solution in such cases is to employ an alternative nucleophilic
partner, such as an arylzinc reagent.6 However, previous attempts
in our group to effect Pd-catalyzed Negishi cross-coupling
reactions using our biaryldialkylphosphine ligands gave little
or no yield of the desired products.7 Since these early efforts,
a new generation of biphenyl-based monophosphines has been
developed which are ortho,ortho′-disubstituted on the lower
(non-phosphine-containing) ring. Due to our recent advances
in both C-N8 and C-C9 bond forming transformations employ-
ing these new ligands, we decided to reinvestigate the Negishi
reaction. As a result of this study into the construction of biaryls,
we report herein a new, highly active and general catalyst system
derived from ligand 2 for the Negishi cross-coupling reaction.
(5) For examples highlighting problems with the use of boronic acids, see:
(a) Handy, S. T.; Zhang, Y.; Bregman, H. J. Org. Chem. 2004, 69, 2362-
2366. (b) Chaumeil, H.; Signorella, S.; Le Drian, C. Tetrahedron 2000,
56, 9655-9662. (c) Watanabe, T.; Miyaura, N.; Suzuki, A. Synlett 1992,
207-210.
(1) Recent reviews on palladium-catalyzed cross-coupling processes: (a)
Hassan, J.; Sevignon, M.; Gozzi, C.; Schulz, E.; Lemaire, M. Chem. ReV.
2002, 102, 1359-1469. (b) Littkle, A. F.; Fu, G. C. Angew. Chem., Int.
Ed. 2002, 41, 4176-4211. (c) Stanforth, S. P. Tetrahedron 1998, 54, 263-
303. (d) Metal-Catalyzed Cross-Coupling Reactions; Diederich, F., Stang,
P. J., Eds.; Wiley-VCH: New York, 1998. (e) Tamao, K.; Miyaura, N.
Top. Curr. Chem. 2002, 219, 1-9. (f) Lessene, G. Aust. J. Chem. 2004, 57,
107 and references therein.
(2) van Leeuwen, P. W. N. M.; Kamer, P. C. J.; Reek, J. N. H.; Dierkes, P.
Chem. ReV. 2000, 100, 2741-2769.
(3) Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2001, 123, 2719-2724.
(4) Walker, S. D.; Barder, T. E.; Martinelli, J. R.; Buchwald, S. L. Angew.
Chem., Int. Ed. 2004, 43, 1871-1876.
(6) Organozinc Reagents, A Practical Approach; Knochel, P., Jones, P., Eds.;
Oxford University Press: New York, 1999.
(7) Previous attempts: Old, D. W.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem.
Soc. 1998, 120, 9722-9723.
(8) Huang, X.; Anderson, K. W.; Zim, D.; Jiang, L.; Klapars, A.; Buchwald,
S. L. J. Am. Chem. Soc. 2003, 125, 6653-6655.
(9) (a) Nguyen, H. N.; Huang, X.; Buchwald, S. L. J. Am. Chem. Soc. 2003,
125, 11818-11819. (b) Gelman, D.; Buchwald, S. L. Angew. Chem., Int.
Ed. 2003, 42, 5993-5996.
9
13028
J. AM. CHEM. SOC. 2004, 126, 13028-13032
10.1021/ja0474493 CCC: $27.50 © 2004 American Chemical Society

A Catalyst for the Negishi Cross-Coupling Reaction
A R T I C L E S
Table 1. Negishi Cross-Coupling of Aryl Chlorides to Generate Biaryls
a Isolated yield; average of at least two runs. b Reaction run at room temperature. c Reaction run at 100 °C, in THF/NMP (1:1).
Results and Discussion
Initial studies indicated that phosphine 2 was an excellent
supporting ligand for the Pd-catalyzed coupling of aryl chlorides
with organozinc reagents. The arylzinc chlorides were typically
prepared in situ from the corresponding aryl bromides. Halogen-
lithium exchange followed by transmetalation with zinc(II)
chloride furnished the desired organozinc species. Zinc(II)
bromide could also be employed; however, the arylzinc chlorides
were more effective in the subsequent cross-coupling step. As
shown in Table 1, the use of a catalyst derived from 2 and Pd2-
dba3 in THF at 70 °C (standard conditions) demonstrated a
tolerance to a variety of common functional groups including
cyano, nitro, ester, alkoxy, and amino substituents (Table 1, 3a-
3f) and operated efficiently at low levels of catalyst. Previously
reported systems required higher temperatures and quantities
of catalyst for similar cross-couplings. For example, 2′-cyano-
2,3-dimethylbiphenyl (3a) (Table 1) was prepared using only
0.01% Pd, which, to our knowledge, is the lowest quantity of
Pd ever employed for the Negishi cross-coupling of an aryl
chloride. The reaction of 1-chloro-4-nitrobenzene with (2,3-
dimethylphenyl)zinc chloride in THF at 70 °C resulted in low
yields of the desired biaryl 3b. However, carrying out the
process at room temperature provided 3b in 94% yield using
as little as 0.1% Pd. In some cases, a directed ortho lithiation
approach10 could be used to access the required arylzinc
reagents, obviating the need to start with an aryl bromide. For
Figure 1. Ligand effects in the coupling of hindered substrates.
example, ortho lithiation of 1,3-dimethoxybenzene, followed by
Li f Zn exchange and cross-coupling of the resulting arylzinc
reagent with 4-chloroanisole, afforded the biaryl in quantitative
yield. Initial attempts to couple [2-(N,N-dimethylamino)-6-
methoxyphenyl]zinc chloride with 4-chloroanisole in THF at
70 °C led to poor conversions of aryl chloride. However,
employing a 1:1 mixture of THF/NMP,11 as described by Fu,3
as the solvent system and increasing the reaction temperature
to 100 °C furnished the desired biaryl 3f in 75% yield. The use
of electron-rich arylzinc reagents, as described above, demon-
strates how this protocol complements the Suzuki-Miyaura
reaction.
As part of our study, the effect of the structure of the
phosphine ligand on the challenging coupling reaction between
(2,4,6-triisopropylphenyl)zinc chloride and 2-chloroanisole was
examined (Figure 1). During this investigation we confirmed
previous work in our laboratory which had shown that use of
ligand 11 gave poor conversions to the desired product.7
However, increasing the size of the substituent on the lower
aromatic ring (ligands 6-9) enhanced the effectiveness of the
ligand. Moreover, increasing the electron density by introducing
methoxy groups on the lower ring (ligands 10 and especially
(10) (a) Sniekus, V. Chem. ReV. 1990, 90, 879-933. (b) Anctil, E. J.-G.; Sniekus,
V. J. Organomet. Chem. 2002, 653, 150-160.
(11) For other examples of the use of NMP in organozinc chemistry, see: Shibli,
A.; Varghese, J. P.; Knochel, P.; Marek, I. Synlett 2001, 818-820 and
references therein.
9
J. AM. CHEM. SOC. VOL. 126, NO. 40, 2004 13029

A R T I C L E S
Milne and Buchwald
Scheme 1. Preparation of Ligand 2 via an Efficient Lithium-Benzyne Procedure
1) markedly improved catalyst performance. A number of new
ligands with variations in the steric and electronic properties of
the lower (non-phosphine-containing) ring were prepared to aid
the investigation. A fine-tuning of the size of the o-alkoxy
substituents on the lower ring (ligands 1, 2, and 4) confirmed
2 as the superior ligand. Unfortunately, we have been unable
to prepare the ligand with two o-tert-butoxy groups on the lower
ring. However, a further increase in the size of the two ortho
O-substituents by installing OTIPS groups at this position (with
concomitant loss of electron-donating capability) provided a
ligand whose use gave a catalyst of diminished activity. These
results suggest that ligand 2 is suitably bulky to allow sufficient
quantities of LPd intermediates to be present in the reaction
mixture.12 Additionally, the level of bulk is small enough to do
this without slowing the transmetalation step. To ascertain the
effect of an electron-deficient lower ring, the difluoro ligand
14 was prepared. Comparing the catalysts derived from ligands
12 and 14 revealed that decreasing the electron density on the
lower ring may diminish catalyst activity. Interestingly, the use
of ligands 11 and 13, which possess N,N-dimethylamino groups,
gave poor conversion of the starting aryl chloride to biaryl
product. For these ligands it is possible that the organozinc
species coordinates to the amino group, slowing the desired
reaction.
Given its utility, it was necessary to develop a concise and
inexpensive route to ligand 2. We have previously demonstrated
that biaryldialkylphosphine ligands can be prepared by the
addition of an aryl Grignard reagent to benzyne (generated in
situ), followed by trapping of the newly formed organomag-
esium species, in the presence of a copper catalyst, with ClPR2.13
However, the presence of both the excess magnesium metal from
the Grignard formation in the first step and the copper salts
added to facilitate C-P bond formation can complicate the
workup of the reaction. Optimization of the synthesis of 2
revealed that a modified one-pot protocol employing a lithium-
benzyne step afforded the ligand cleanly in 53-71% yield
(Scheme 1). The route begins with the directed ortho lithiation
of 1,3-diisopropoxybenzene (1.1 equiv) achieved by heating at
80 °C in hexanes for 2 h in the presence of n-butyllithium (1.2
equiv). Maintaining the reaction at reflux, dropwise addition
of neat 2-bromochlorobenzene (1.0 equiv) over 50 min generated
the bromobiphenyl 15 via a tandem benzyne condensation-
bromine atom transfer sequence.14 Cooling of the reaction
mixture to -78 °C and halogen-lithium exchange with n-
butyllithium (1.1 equiv), followed by treatment with chlorodi-
cyclohexylphosphine (1.0 equiv), furnished ligand 2. This
method offered a number of advantages over previous methods
for the preparation of the ligands on a small scale: it was faster
and cleaner, eliminated the need for Cu(I) salts, and avoided
an additional step to prepare an aryl halide as the precursor.
Ligands 4, 5, and 14 were prepared using the same method.
A challenging problem in Negishi coupling processes is the
ability to combine sterically hindered substrates. This is
particularly difficult when both coupling partners have large
ortho substituents. Using the standard conditions, a range of
exceptionally hindered di- and tri-ortho-substituted biaryls were
prepared from aryl chlorides using only 0.1-1% Pd. Arylzinc
chlorides containing ortho substituents such as methyl, alkoxy,
isopropyl, and N,N-dimethylamino groups also underwent
smooth cross-coupling with complete conversion of the starting
aryl chlorides. We also found that it was possible to efficiently
couple (2,4,6-triisopropylphenyl)zinc chloride with a variety of
ortho-substituted aryl chlorides (16a-c). To our knowledge, this
is the most highly hindered arylzinc species ever employed in
a cross-coupling process. The reaction of [2-(N,N-dimethy-
lamino)-6-methoxyphenyl]zinc chloride with 2-chlorobenzoni-
trile proceeded smoothly in 66% yield; however, when 2-chlo-
roanisole was employed, no reaction was observed under the
standard conditions. Employing the procedure using THF/NMP
as the solvent mixture at 100 °C resulted in complete conversion
of the substrates and produced 16e in 76% yield.
Due to the high level of catalyst activity displayed by 2/Pd-
(0), we decided to investigate its use for the preparation of tetra-
ortho-substituted biaryls. To our delight, biaryls 16g and 16h
could be prepared in high yield from 9-chloroanthracene and
(12) (a) Alcazar-Roman, L. M.; Hartwig, J. F. J. Am. Chem. Soc. 2001, 123,
12905-12906. (b) Strieter, E. R.; Blackmond, D. G.; Buchwald, S. L. J.
Am. Chem. Soc. 2003, 125, 13978-13980.
(13) (a) Tomori, H.; Fox, J. M.; Buchwald, S. L. J. Org. Chem. 2000, 65, 5334-
5341. (b) Kaye, S.; Fox, J. M.; Hicks, F. A.; Buchwald, S. L. AdV. Synth.
Catal. 2001, 343, 789-794.
(14) Leroux, F.; Schlosser, M. Angew. Chem., Int. Ed. 2002, 41, 4272-4274.
9
13030 J. AM. CHEM. SOC. VOL. 126, NO. 40, 2004

A Catalyst for the Negishi Cross-Coupling Reaction
A R T I C L E S
Table 3. Negishi Cross-Coupling of Heteroaryl Halides to
Table 2. Negishi Cross-Coupling of Aryl Halides to Generate
Hindered Biaryls
Generate Biaryls
a Isolated yield; average of at least two runs. b Reaction run at room
temperature.
examined the catalyst derived from 2/Pd(0) for Negishi coupling
reactions involving heterocyclic substrates. As shown in Table
3, the cross-coupling of a variety of heterocyclic chlorides and
bromides was accomplished in good to excellent yields, includ-
ing a quinoxaline, benzothiazole, pyrimidine, pyrazine, quino-
line, tetrazole, and pyrazole as well as substituted pyridines.
Unfortunately, to date, our limited attempts to couple het-
eroarylzinc reagents have been unsuccessful. Current efforts are
focused in this area.
In summary, structural refinements to the ligand framework
of previously reported biarylphosphines gave a ligand whose
use provides a catalyst of unparalleled reactivity for Negishi
cross-coupling reactions. This system is the most effective of
any known for Negishi cross-couplings with respect to turnover
numbers for reactions of aryl chlorides, for the efficient
formation of hindered biaryls, and for substrate scope.
a Isolated yield; average of at least two runs. b Reaction run at 100 °C,
in THF/NMP (1:1).
1-chloro-2,6-dimethoxybenzene, respectively. To our knowl-
edge, the reaction of electron-rich 1-chloro-2,6-dimethoxyben-
zene with (2,6-dimethylphenyl)zinc chloride is the first example
of a Negishi reaction that forms a biaryl with four ortho
substituents employing a deactivated aryl chloride. Unfortu-
nately, the preparation of biaryls with four o-methyl groups was
plagued by homocoupling as a side reaction, along with the
desired cross-coupling, resulting in inseparable mixtures of
products.
The new catalyst system was also found to be quite useful
for coupling unactivated aryl bromides as shown in Table 2
(16i-l). First, we investigated the challenging formation of
triaryls 16i-k in THF at 70 °C; these reactions all proceeded
in >90% yield. However, the coupling of (2,4,6-triisopropy-
lphenyl)zinc chloride with 2-bromobiphenyl only proceeded to
41% conversion under the standard conditions using 1% Pd.
Increasing the catalyst loading to 2% Pd resulted in only a
moderate increase in efficiency (56% conversion). As before,
by using THF/NMP as the solvent system and increasing the
temperature to 100 °C, 16l could be isolated in 74% yield.
There are an increasing number of heteroaryl compounds
being discovered as lead structures for new pharmaceuticals.
Despite their importance, the cross-coupling reaction of het-
eroaryl halides remains challenging. With this in mind, we next
Experimental Section
The following are representative procedures for the Pd(0)-catalyzed
Negishi cross-coupling reaction. A full account of the reaction
conditions and characterization of products can be found in the
Supporting Information.
General Procedure: Pd-Catalyzed Negishi Couplings of Aryl
Halides and Arylzinc Chlorides. An oven-dried resealable Schlenk
tube containing a magnetic stir bar was capped with a rubber septum
and then evacuated and backfilled with argon (this sequence was
repeated three times). The Schlenk tube was charged with the aryl
bromide (0.75 mmol, 1.5 equiv) and dry THF (1.0 mL). The resulting
solution was cooled to -78 °C, then n-butyllithium (0.825 mmol, 1.65
equiv) was added dropwise via syringe through the septum, and the
resulting solution was stirred at -78 °C for 1 h. ZnCl2 (0.9 mmol, 1.8
equiv) was added in one solid portion by removal of the septum. After
30 min at -78 °C, the Schlenk tube was removed from the cooling
bath and the resulting solution stirred at room temperature for 1 h.
Pd2(dba)3 (2.3 mg, 0.5 mol %), 2 (4.7 mg, 2.0 mol %), and the aryl
chloride (0.5 mmol, 1.0 equiv) were added with the aid of THF (0.5
9
J. AM. CHEM. SOC. VOL. 126, NO. 40, 2004 13031

A R T I C L E S
Milne and Buchwald
mL), which was used to rinse the walls of the tube. The septum was
replaced with a Teflon screwcap, and the Schlenk tube was sealed.
The reaction mixture was placed in a preheated oil bath at 70 °C and
magnetically stirred until the aryl halide had been completely consumed
as judged by GC analysis. The reaction mixture was then cooled to
room temperature, diluted with water (1 mL), and extracted with diethyl
ether (4 × 10 mL). The combined organic phases were dried over Na2-
SO4 and concentrated under reduced pressure. The crude material
obtained was purified by flash chromatography on silica gel.
Modified Procedure for Pd-Catalyzed Negishi Couplings at Low
Catalyst Loading (e0.5% Pd). The general procedure was used with
the following changes: A separate vial was charged with Pd2(dba)3
(2.3 mg, 0.5 mol %) and 2 (4.7 mg, 2.0 mol %). The vial was sealed
with a Teflon-coated screwcap, a needle was inserted through the cap,
and the vial was purged with argon. Dry THF (2 mL) was added, and
the mixture was sonicated for ∼1 min to afford a homogeneous solution.
The required quantity of catalyst solution (e.g., 20 µL for reactions
carried out at 0.01% Pd) was added to the reaction mixture.
100 °C and magnetically stirred until the aryl halide had been
completely consumed as judged by GC analysis.
Modified Procedure for Pd-Catalyzed Negishi Couplings to
Generate Biaryls Combining a Directed Ortho Metalation Ap-
proach. The general procedure was used with the following changes
to the preparation of the arylzinc species: To a cold (0 °C), stirred
solution of 1,3-dimethoxybenzene (98 µL, 0.75 mmol) in dry THF (1
mL) was added n-butyllithium (0.33 mL, 2.5 M in hexanes, 0.825
mmol) dropwise via syringe over 10 min. The reaction mixture was
allowed to warm to room temperature and stirred for 2 h. The
aryllithium was cooled to -78 °C, and ZnCl2 (0.9 mmol, 1.8 equiv)
was added in one solid portion by removal of the septum. After 30
min at -78 °C, the reaction was removed from the cooling bath and
the resulting solution stirred at room temperature for 1 h.
Acknowledgment. We thank the National Institutes of Health
(Grant GM 46059) for funding this work. We are grateful to
Pfizer, Merck, Novartis, and Engelhard for additional support.
Modified Procedure for Pd-Catalyzed Negishi Couplings to
Generate Hindered Biaryls. The general procedure was used with
the following changes: After the addition of the catalyst and the aryl
halide, NMP (1.0 mL) was used to rinse the walls of the tube. The
septum was replaced with a Teflon screwcap, and the Schlenk tube
was sealed. The reaction mixture was placed in a preheated oil bath at
Supporting Information Available: Preparation and charac-
terization of ligands and products (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA0474493
9
13032 J. AM. CHEM. SOC. VOL. 126, NO. 40, 2004