Revealing a second transmetalation step in the Negishi coupling and its competition with reductive elimination: Improvement in the interpretation of the mechanism of biaryl syntheses

Article abstract of DOI:10.1021/ja903277d

This paper presents an experimental and theoretical investigation of the Pd-catalyzed Negishi coupling reaction and reveals a novel second transmetalation reaction between an Ar¹-Pd-Ar² species and the organozinc reagent Ar²-ZnX. Understanding of this second step reveals how homocoupling and dehalogenation products are formed. Thus, the second transmetalation generates Ar²PdAr² and Ar ¹ZnCl, which upon reductive elimination and hydrolysis, respectively, give the homocoupling product Ar²-Ar² and the dehalogenation product Ar¹H. The ratio of the cross-coupling product Ar¹-Ar² and the homocoupling product Ar ²-Ar² is determined by competition between the second transmetalation and reductive elimination steps. This mechanism is further supported by density functional theoretical calculations. Calculations on a series of reactions suggest a strategy in controlling the selectivity of cross-coupling and homocoupling pathways, which we have experimentally verified.

Full text of DOI:10.1021/ja903277d

Published on Web 07/02/2009
Revealing a Second Transmetalation Step in the Negishi
Coupling and Its Competition with Reductive Elimination:
Improvement in the Interpretation of the Mechanism of
Biaryl Syntheses
Qiang Liu,† Yu Lan,‡,§ Jing Liu,† Gang Li,† Yun-Dong Wu,*,‡,§ and Aiwen Lei*,†,⊥
College of Chemistry and Molecular Sciences, Wuhan UniVersity, Wuhan, Hubei 430072, China,
College of Chemistry, Peking UniVersity, Beijing, China, Department of Chemistry, The Hong
Kong UniVersity of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China,
and State Key Laboratory of Organometallic Chemistry, Shanghai Institute of Organic
Chemistry, Chinese Academy of Science, 354 Fenglin Lu Shanghai, 200032, China
Received March 26, 2009; E-mail: aiwenlei@whu.edu.cn; chydwu@ust.hk
Abstract: This paper presents an experimental and theoretical investigation of the Pd-catalyzed Negishi
coupling reaction and reveals a novel second transmetalation reaction between an Ar1-Pd-Ar2 species
and the organozinc reagent Ar2-ZnX. Understanding of this second step reveals how homocoupling and
dehalogenation products are formed. Thus, the second transmetalation generates Ar2PdAr2 and Ar1ZnCl,
which upon reductive elimination and hydrolysis, respectively, give the homocoupling product Ar2-Ar2 and
the dehalogenation product Ar1H. The ratio of the cross-coupling product Ar1-Ar2 and the homocoupling
product Ar2-Ar2 is determined by competition between the second transmetalation and reductive elimination
steps. This mechanism is further supported by density functional theoretical calculations. Calculations on
a series of reactions suggest a strategy in controlling the selectivity of cross-coupling and homocoupling
pathways, which we have experimentally verified.
Scheme 1. General Catalytic Cycle of Biaryl Syntheses
Introduction
Transition-metal-catalyzed cross-coupling reactions have
become extremely useful in organic syntheses.1 Because of the
recent emergence of a large number of drug candidates
containing biaryls,2 Pd- and Ni-catalyzed coupling of aryl
halides and aryl metal species has been extensively studied.1,3-6
Although transmetalation is an essential step in these catalytic
reactions, the mechanism of this step is not well studied
especially for Negishi coupling.7 Understanding this essential
step will enhance the development of this important class of
reactions and increase their synthetic utility. For example, it is
well known that the formation of homocoupling byproducts is
a serious limitation to synthetic chemists and an obstacle to
larger industrial-scale development.8 We show in this paper that
the selectivity of an important cross-coupling reaction, the
Negishi coupling, is sometimes strongly affected by the trans-
metalation step.
† Wuhan University.
‡ Peking University.
§ The Hong Kong University of Science and Technology.
⊥ Shanghai Institute of Organic Chemistry, Chinese Academy of Science.
(1) Anastasia, L.; Negishi, E.-i. In Handbook of Organopalladium
Chemistry for Organic Synthesis; 2002; Vol. 1, pp 311-334.
(2) Corbet, J.-P.; Mignani, G. Chem. ReV. 2006, 106, 2651–2710.
(3) Hassan, J.; Sevignon, M.; Gozzi, C.; Schulz, E.; Lemaire, M. Chem.
ReV. 2002, 102, 1359–1469.
A well-accepted catalytic cycle of cross-coupling reactions
is shown in Scheme 1, which includes three successive steps:
oxidative addition of Ar1X with the Pd catalyst, transmetalation
with Ar2M to furnish the key intermediate Ar1-Pd-Ar2, and
reductive elimination to yield the desired cross-coupling biaryls
and to regenerate the catalyst.4 However, the cross-coupling
reaction sometimes affords not only the cross-coupling product
but also the products of dehalogenation, Ar1-H, and homo-
coupling, Ar1-Ar1 and Ar2-Ar2,1,9-19 for which there are
(4) Meijere, A. d.; Diederich, F. Metal-Catalyzed Cross-Coupling Reac-
tions, 2nd ed.; Wiley-VCH: Weinheim, 2004.
(5) Negishi, E.-i. J. Organomet. Chem. 2002, 653, 34–40.
(6) Alberico, D.; Scott, M. E.; Lautens, M. Chem. ReV. 2007, 107, 174–
238.
(7) Casares, J. A.; Espinet, P.; Fuentes, B.; Salas, G. J. Am. Chem. Soc.
2007, 129, 3508–3509.
(8) Banno, T.; Hayakawa, Y.; Umeno, M. J. Organomet. Chem. 2002,
653, 288–291.
(9) Alimardanov, A.; Schmieder-van de Vondervoort, L.; de Vries,
A. H. M.; de Vries, J. G. AdV. Synth. Catal. 2004, 346, 1812–1817.
9
10.1021/ja903277d CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 10201–10210 10201

A R T I C L E S
Liu et al.
Table 1. Negishi Coupling of ArI with PhZnCla
almost no well-accepted mechanisms. Understanding these
mechanisms will help to optimize the catalytic cross-coupling
toward the desired reaction pathway.
While oxidative addition and reductive elimination of Pd-
catalyzed coupling reactions have been extensively studied both
experimentally and theoretically,7,20-46 the transmetalation
reaction involving Ar1-Pd-Ar2 has only been sparsely
studied.47,48 In this paper, we report an experimental and
yield %b
entry
R
4a
5
6
1
2
3
4
5
6
7
8
9
2-CO2Et
4-CO2Et
2-CONMe2
1a
1c
1d
1e
1f
1g
1h
1i
76
trace
39
70
0
38
60
19
0
20
trace
17
18
99c
61
(10) Bumagin, N. A.; Ponomarev, A. B.; Beletskaya, I. P. Zh. Org. Khim.
1987, 23, 1354–1364.
2-CONHBu
2-MeO
50
40
29
65c
86c
79
(11) Bumagin, N. A.; Ponomarev, A. B.; Beletskaya, I. P. Zh. Org. Khim.
1987, 23, 1345–1353.
4-MeO
13c,d
14
2-i Pr
(12) Organ, M. G.; Ghasemi, H.; Valente, C. Tetrahedron 2004, 60, 9453–
9461.
(13) Okamoto, Y.; Yoshioka, K.; Yamana, T.; Mori, H. J. Organomet.
Chem. 1989, 369, 285–290.
(14) Negishi, E.; Takahashi, T.; Akiyoshi, K. J. Organomet. Chem. 1987,
334, 181–194.
(15) Bumagin, N. A.; Ponomarev, A. B.; Beletskaya, I. P. J. Organomet.
Chem. 1985, 291, 129–132.
4-i Pr
trace
22
99
77
2,4,6-trimethyl
1j
a The reactions were conducted with 0.5 mmol of 1, 1 mmol of 2a,
and 3 mol % of PdCl2(dppf) in THF at 60 °C for 2 h. b The yield was
determined by GC. c Isolated yield. d 4,4′-Dimethoxybiphenyl (7) was
detected.
(16) Tsai, F.-Y.; Lin, B.-N.; Chen, M.-J.; Mou, C.-Y.; Liu, S.-T. Tetra-
hedron 2007, 63, 4304–4309.
(17) Lau, K. C. Y.; Chiu, P. Tetrahedron Lett. 2007, 48, 1813–1816.
(18) Coelho, A. V.; de Souza, A. L. F.; de Lima, P. G.; Wardell, J. L.;
Antunes, O. A. C. Tetrahedron Lett. 2007, 48, 7671–7674.
(19) Arcadi, A.; Cerichelli, G.; Chiarini, M.; Correa, M.; Zorzan, D. Eur.
J. Org. Chem. 2003, 4080–4086.
theoretical study of the Pd-catalyzed Negishi coupling reaction
of Ar1I with Ar2ZnCl. This study provides insight into the
transmetalation of Negishi coupling and leads to the develop-
ment of a strategy that maximizes the selectivity toward cross-
coupling products. Here we report these results and this new
strategy.
(20) Nova, A.; Ujaque, G.; Maseras, F.; Lledos, A.; Espinet, P. J. Am.
Chem. Soc. 2006, 128, 14571–14578.
(21) Nilsson, P.; Puxty, G.; Wendt, O. F. Organometallics 2006, 25, 1285–
1292.
Results and Discussion
(22) Ariafard, A.; Lin, Z.; Fairlamb, I. J. S. Organometallics 2006, 25,
5788–5794.
1. Observation of Formation of Dehalogenation and Homo-
coupling Side Products. During the course of studying the Negishi
coupling reaction using diene ligand 3,49 we examined the
reaction of ethyl o-iodobenzoate (1a) with phenylzinc chloride
(2a) as shown in eq 1. Although this is usually regarded as a
“standard” Negishi coupling for biaryl syntheses, surprisingly,
we found only trace amounts of the cross-coupling biaryl
product 6a was formed, and, instead, the major products were
biphenyl 4a and deiodoarylation product 5a. Furthermore, we
noticed that 4a and 5a were formed in almost equal amounts.
We screened 10 additional Pd catalysts and ligands (see Table
S1 in the Supporting Information) and generally found that this
reaction produced considerable levels of homocoupling and
dehalogenation products.
(23) Espinet, P.; Echavarren, A. M. Angew. Chem., Int. Ed. 2004, 43, 4704–
4734.
(24) Crociani, B.; Antonaroli, S.; Beghetto, V.; Matteoli, U.; Scrivanti, A.
Dalton Trans. 2003, 2194–2202.
(25) Ricci, A.; Angelucci, F.; Bassetti, M.; Lo Sterzo, C. J. Am. Chem.
Soc. 2002, 124, 1060–1071.
(26) Casares, J. A.; Espinet, P.; Salas, G. Chem.-Eur. J. 2002, 8, 4843–
4853.
(27) Franz, A. K.; Woerpel, K. A. J. Am. Chem. Soc. 1999, 121, 949–957.
(28) Casado, A. L.; Casares, J. A.; Espinet, P. Organometallics 1997, 16,
5730–5736.
(29) Louie, J.; Hartwig, J. F. J. Am. Chem. Soc. 1995, 117, 11598–11599.
(30) Perez-Temprano, M. H.; Nova, A.; Casares, J. A.; Espinet, P. J. Am.
Chem. Soc. 2008, 130, 10518–10520.
(31) Hills, I. D.; Netherton, M. R.; Fu, G. C. Angew. Chem., Int. Ed. 2003,
42, 5749–5752.
(32) Goossen, L. J.; Koley, D.; Hermann, H.; Thiel, W. Chem. Commun.
2004, 2141–2143.
(33) Ananikov, V. P.; Musaev, D. G.; Morokuma, K. Eur. J. Inorg. Chem.
2007, 5390–5399.
(34) Surawatanawong, P.; Fan, Y.; Hall, M. B. J. Organomet. Chem. 2008,
693, 1552–1563.
(35) Braga, A. A. C.; Morgon, N. H.; Ujaque, G.; Maseras, F. J. Am. Chem.
Soc. 2005, 127, 9298–9307.
(36) Henriksen, S. T.; Norrby, P.-O.; Kaukoranta, P.; Andersson, P. G.
J. Am. Chem. Soc. 2008, 130, 10414–10421.
(37) Kozuch, S.; Shaik, S. J. Am. Chem. Soc. 2006, 128, 3355–3365.
(38) Yan, X.-X.; Peng, Q.; Li, Q.; Zhang, K.; Yao, J.; Hou, X.-L.; Wu,
Y.-D. J. Am. Chem. Soc. 2008, 130, 14362–14363.
(39) Ahlquist, M.; Fristrup, P.; Tanner, D.; Norrby, P.-O. Organometallics
2006, 25, 2066–2073.
(40) Ahlquist, M.; Norrby, P.-O. Organometallics 2007, 26, 550–553.
(41) Ananikov, V. P.; Musaev, D. G.; Morokuma, K. Organometallics 2005,
24, 715–723.
(42) Braga, A. A. C.; Ujaque, G.; Maseras, F. Organometallics 2006, 25,
3647–3658.
To further clarify this trend, reactions exploring different aryl
iodide electrophiles with phenylzinc chloride 2a using
PdCl2(dppf) as the catalyst were examined, and the results were
presented in Table 1. Compared with ethyl o-iodobenzoate 1a,
(43) Goossen, L. J.; Koley, D.; Hermann, H. L.; Thiel, W. Organometallics
2005, 24, 2398–2410.
(44) Li, Z.; Fu, Y.; Guo, Q.-X.; Liu, L. Organometallics 2008, 27, 4043–
4049.
(45) Lucassen, A. C. B.; Shimon, L. J. W.; Van der Boom, M. E.
Organometallics 2006, 25, 3308–3310.
(48) Cardenas, D. J.; Martin-Matute, B.; Echavarren, A. M. J. Am. Chem.
Soc. 2006, 128, 5033–5040.
(46) Senn, H. M.; Ziegler, T. Organometallics 2004, 23, 2980–2988.
(47) Ozawa, F.; Fujimori, M.; Yamamoto, T.; Yamamoto, A. Organome-
tallics 1986, 5, 2144–2149.
(49) Liu, Q.; Duan, H.; Luo, X.; Tang, Y.; Li, G.; Huang, R.; Lei, A. AdV.
Synth. Catal. 2008, 350, 1349–1354.
9
10202 J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009

Interpretation of the Mechanism of Biaryl Syntheses
A R T I C L E S
Scheme 2. Possible Pathways for the Homocoupling and
Dehalogenation Side Products
Scheme 4. Deuteration Experiment
which gives only a small amount of cross-coupling product,
the reaction of ethyl p-iodobenzoate (1c) with phenylzinc
chloride 2a yielded exclusively the cross-coupling product
(Table 1, entry 2). However, other ortho-substituted aryl iodides
containing electron-withdrawing or electronic-rich groups showed
similar reactivity with phenylzinc chloride 2a as with ethyl
o-iodobenzoate 1a and generated significant amounts of homo-
coupling biphenyl and reduction products (Table 1, entries 3-5,
7, and 9). The reactions of para-substituted aryl iodides with
phenylzinc chloride 2a exhibited excellent selectivity favoring
the cross-coupling products (Table 1, entries 2, 6, and 8).
Four possible pathways for the formation of homocoupling
and dehalogenation side products are shown in Scheme 2.
Pathway A involves a transmetalation of the oxidative addition
intermediate Ar1-Pd-X with Ar2ZnX to form Ar2-Pd-X. This
species undergoes a second transmetalation to give Ar2-Pd-Ar2
that results in the formation of homocoupling product Ar2-Ar2.
This pathway was proposed by Elsevier et al. for the coupling
reaction of benzyl bromide and p-TolM (M ) MgBr or ZnBr)
in the presence of Pd catalyst using Ar-BIAN (dimethyl
fumarate) as the supporting ligand.50 Pathway B involves
disproportionation of Ar1-Pd-X, which leads to the formation
of Ar1-Pd-Ar1 and PdX2. Pathway C involves disproportion-
ation of Ar1-Pd-Ar2, which leads to the formation of
Ar1-Pd-Ar1 and Ar2-Pd-Ar2 and gives homocoupling side
products Ar1-Ar1 and Ar2-Ar2. Although the disproportionation
has been observed in stoichiometric reactions,28,51-58 pathways
B and C cannot explain the formation of dehalogenation side
product Ar1-H. Since we observe that homocoupling and
dehalogenation products are formed in equal amounts, this
suggests that the pathways leading to these two products are
related. Recently, Casares and Espinet et al. observed the
transmetalation between PdRfMe(PPh3)2 (Rf ) 3,5-dichloro-
2,4,6-trifluorophenyl) and Me2Zn.7 Thus, we propose pathway
D. This pathway involves a second transmetalation between
Ar2ZnX and Ar1-Pd-Ar2, which leads to the formation of
homocoupling and dehalogenation products.
the reaction mixture led to no reaction. However, when the same
reaction was carried out in the presence of 5 equiv of 2a, the
products 4a, 5b, and 6b were formed in 60%, 60%, and 40%
yields, respectively. This product distribution is consistent with
the proposed pathway D shown in Scheme 2. In this pathway,
the initial transmetalation reaction results in the formation of
Ar1-Pd-Ar2. Without excess Ar2ZnCl, its reductive elimination
yields the cross-coupling product. However, in the presence of
excess Ar2ZnCl, a second transmetalation reaction, which
competes with reductive elimination, occurs, leading to the
homocoupling product.
2. Identification of the Formation of Ar1ZnI from the
Reaction of Ar1I with Ar2ZnCl. For more information about
the reaction mechanism, we carried out experiments to identify
reaction intermediates and determine reaction kinetics. If
pathway D in Scheme 2 operates, Ar1ZnCl will be generated
and subsequent hydrolysis will form the dehalogenated side
product, Ar1H. Therefore, to verify the hypotheses, it became a
key issue to identify the formation of Ar1ZnCl (Scheme 2,
pathway D).
The first experiment involved deuteration (Scheme 4). The
reaction of ethyl o-iodobenzoate 1a with phenylzinc chloride
2a was carried out at 60 °C for 2 h and then treated by
CH3COOD. As a result, the deuterated product 10 was observed.
The molar ratio of 4a to 10 is 1:1. This provides indirect
evidence for the existence of Ar1ZnCl in the reaction solution
(Scheme 4).
The second attempt to identify 2-COOEtPhZnCl 9a involved
a trapping experiment shown in Scheme 5. In this experiment,
ethyl o-iodobenzoate 1a reacted with phenylzinc chloride 2a at
60 °C for 2 h, and to the resulting mixture was added
p-iodoanisole (1g). After an additional 2 h, the biaryl product
11 was isolated from this one-pot reaction in 56% yield. This
supports the formation of arylzinc chloride 9a, which then
reacted with 1g in the presence of palladium catalyst to afford
11. This experiment provides additional evidence for the
existence of 2-COOEtPhZnCl 9a from the reaction of ethyl
o-iodobenzoate 1a with phenylzinc chloride 2a.
To differentiate among these four pathways, stoichiometric
experiments were carried out as shown in Scheme 3. When the
oxidative addition intermediate 8 was reacted with only 1 equiv
of phenylzinc chloride (2a), the reaction generated the cross-
coupling product 6b in 95% yield. Further addition of 2a to
To obtain evidence for the formation of the Ar1ZnCl species
that results from the second transmetalation, we also performed
Scheme 3. Stoichiometric Experiments
9
J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009 10203

A R T I C L E S
Liu et al.
Scheme 5. Trapping Experiment
13C NMR studies. First, we prepared an authentic sample of
(o-MeOOCPh)ZnI from the reaction of methyl o-iodobenzoate
1b with activated Zn dust in DMSO. The 13C NMR of the
carbonyl region shows a resonance at 166.2 ppm (Figure 1 B).
In addition, the carbonyl region of 5b shows a resonace at 167.0
ppm (Figure 1D) and 6b shows a resonance at 169.1 ppm
(Figure 1C). Second, the reaction of methyl o-iodobenzoate 1b
with phenylzinc chloride 2a was performed in an NMR tube
for 2 h, and the 13C NMR of the reaction mixture was recorded.
The spectrum (Figure 1A) shows two resonances. The resonance
at 169.1 ppm is due to the product 6b, as it matches the spectrum
in Figure 1C. The resonance at 165.9 ppm nicely matches the
resonance of (o-MeOOCPh)ZnI (166.2 ppm, Figure 1B) and
confirms the identity of Ar1ZnCl.
We also monitored the reaction of methyl o-iodobenzoate
1b with phenylzinc chloride 2a using in situ IR. The data
are presented in Figure S1 and Figure S2 of the Supporting
Information. Compared with an authentic sample, 2-MeOOC-
PhZnCl 9b is assigned as a newly formed component from
the reaction of methyl o-iodobenzoate 1b and phenylzinc
chloride 2a. Furthermore, kinetic profiles of starting materials
of methyl o-iodobenzoate 1b and the 2-COOMePhZnCl 9b
are listed in Figure 2. By plotting the linear region the
observed reaction rate for the decline of 1b was 1.33 × 10-3
M · s-1 (Figure 2C), and the observed rate for the increase of
9b was 1.08 × 10-3 M · s-1 (Figure 2D). These kinetic data
support the formation of 9b from the reaction of 1b with
phenylzinc chloride 2a.
3. Proposed Mechanism. The above experiments show the
following: (1) The products of Ar2-Ar2 and Ar1-H are
formed in some of the reactions between Ar1-I and Ar2ZnCl,
and they are derived from reductive elimination of Ar2PdAr2
and hydrolysis of Ar1ZnCl, respectively. Both are derived
from a second transmetalation between Ar1PdAr2 and
Ar2ZnCl. (2) The two products are formed in about the same
amount. (3) The transmetalation between Ar1PdI and Ar2ZnCl
gives Ar2PdAr1. These observations allow us to exclude the
pathways shown in Scheme 2, eqs A-C, and to propose a
mechanism for this Negishi reaction as shown in Scheme 6
using 1b as an example. After the oxidative addition of 1b
to form I, first transmetalation results in the formation of
intermediate II. While reductive elimination of II leads to
the formation of the cross-coupling product 6b, a second
transmetalation can competitively occur to result in the
formation of 9b and intermediate III. Intermediate III
undergoes rapid reductive elimination to give the homocou-
pling product 4a, and the dehalogenation product comes from
the hydrolysis of 9b.
4. Theoretical Study. 4.1. The First Transmetalation Reac-
tion. To support our proposed mechanism and to provide more
insight into the nature of Negishi coupling, density functional
calculations were performed. We first studied the transmeta-
lation between I (in Scheme 6) and phenylzinc chloride. Our
calculations used Pd(Me2P-CH2-CH2-PMe2) as a model for
the catalyst and coupling substrates methyl o-iodobenzoate
1b and phenylzinc chloride 2a. Complex 12 (Figure 3) was
Figure 1. (A) 13C spectrum of the reaction mixture starting from 0.5 mmol of 1b and 1 mmol of 2a in THF; (B) 13C spectrum of authentic arylzinc reagent
9b in DMSO; (C) 13C spectrum of 6b in CDCl3; (D) 13C spectrum of 5a in CDCl3.
9
10204 J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009

Interpretation of the Mechanism of Biaryl Syntheses
A R T I C L E S
Scheme 6. Proposed Mechanism
Our calculations suggest that there are two possible pathways
for the transformation of 12. The first pathway involves the
replacement of the iodo ligand by the Ph group and results in
the formation of 14. The second pathway involves replacement
of the aryl group on the Pd in 12 by the Ph on Zn group to give
16.
The calculated activation free energy of the first pathway
occurring through transition state 13-ts is 18.6 kcal/mol, and
the relative free energy of the transmetalation product 14 is
5.6 kcal/mol. Although the generation of 14 is calculated to
be endergonic, aggregation of Zn(I)Cl is expected, which
should result in a significant stabilization to make the reaction
irreversible. In contrast, the calculated activation free energy
for the second pathway occurring via transition state 15-ts
is 26.7 kcal/mol. Thus, in terms of free energy of activation,
the first pathway is much more favorable than the second
pathway by 8.1 kcal/mol. Since the first pathway gives an
Ar1PdAr2 species, this result is in agreement with the
experimental observations (Scheme 3) and supports the
mechanism in Scheme 6.
4.2. Competition between Reductive Elimination of II
(Scheme 6) and Second Transmetalation. After confirming the
formation of intermediate II as shown in Scheme 6 as the
product of the first transmetalation reaction, we turned our
the starting structure. Because the complexation energy
between phenylzinc chloride and the Ar-Pd complex could
not be calculated accurately, the energy of this species was
set to zero.
Figure 2. (A) Kinetic profile of [1b] vs t. (B) Kinetic profile of [9b] Vs t. (C) Linear fit of part of the kinetic region of [1b] vs t. (D) Linear fit of part of
the kinetic region of [9b] vs t.
Figure 3. Free energy contrast of the two competitive transmetalation pathways.
9
J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009 10205

A R T I C L E S
Liu et al.
Figure 4. Free energy profile of the competition between reductive elimination (path 1) and transmetalation (path 2). The values are the calculated relative
Gibbs free energies with solvation of THF. The values in parentheses are the calculated relative enthalpies in THF.
attention to the competitive reaction between reductive
elimination and the second transmetalation with II. Again,
the coordination energy of phenylzinc chloride and II could
not be calculated exactly. Therefore, we set the II-PhZnCl
complex 17 as the reference of the free energy surface. This
approach makes the comparison of activation energy of
reductive elimination and second transmetalation steps more
reasonable because both are now unimolecular reactions, Vide
infra.
geometrical parameters are given in Table 2. The calculation
results can be summarized as follows:
(1) For the parent system (entry 1), where both Ar1 and
Ar2 are phenyl groups, the calculated ∆G1q and ∆G2q are 16.7
and 22.3 kcal/mol, respectively. The results show that without
a substituent on the two aryl groups the second transmeta-
lation hardly occurs due to the 5.6 kcal/mol activation energy
difference. This is in agreement with experiment.
(2) When Ar1 bears a para-substituent (entries 2, 3), ∆G1q is
The calculated potential free energy surface of the reductive
elimination and the second transmetalation reactions is shown
in Figure 4. The calculated free energy of activation of the
reductive elimination occurring through 18-ts to form the cross-
coupling biaryl product is 19.8 kcal/mol in THF solvent. We
also calculated the same reaction without the coordination of
phenylzinc chloride (Table 2) as 20.4 kcal/mol, indicating that
the coordination of phenylzinc chloride has little effect on the
reductive elimination. The reductive elimination reaction is
calculated to be exergonic by 12.4 kcal/mol (see 21 in Fig-
ure 4).
slightly increased. The most significant effect is the electron-
withdrawing ester group on ∆G2 , which is increased by about
3 kcal/mol. Thus, a para electron-withdrawing substituent on
Ar1 favors the reductive elimination with respect to the second
transmetalation, while a para electron-donating group has the
opposite effect, although both effects are mild.
q
(3) When Ar1 has an ortho substituent (entries 4-6), the
barrier to reductive elimination is increased by 2-4 kcal/mol.
The barrier to the second transmetalation is reduced by the ester
group (by 2.7 kcal/mol) and methoxyl group (by 0.7 kcal/mol),
probably due to the Zn-O coordination in the transition state.
This is increased by about 1 kcal/mol by an isopropyl group.
In the case of entry 7, where Ar1 has two ortho methyl groups,
the steric effect is increased, resulting in a further increase in
The second transmetalation occurs through transition state
19-ts. It has a calculated activation free energy of 18.3 kcal/
mol, which is 1.5 kcal/mol lower than that of the reductive
elimination reaction. In fact, the second transmetalation product
20 is calculated to be slightly more stable than 17. It undergoes
reductive elimination through transition state 22-ts with an
activation free energy of 14.7 kcal/mol. This result is in
qualitative agreement with the experimental observation that
products 4a (homocoupling) and 5a (dehalogenation) are
dominant when Ar1I is 1a (Table 1).
q
q
both ∆G1 and ∆G2 . Overall, the steric effect by the ortho
substituent(s) in Ar1 is more significant in the reductive
elimination transition state than in the transmetalation transition
state.
(4) The last four entries in Table 3 indicate that when the
Ar2 has an ortho substituent, the free energy of activation of
q
the second transmetalation reaction (∆G2 ) is significantly
increased. The transition state apparently is very sensitive to
4.3. Controlling Factors for the Competition between Reduc-
tive Elimination and Second Transmetalation. To help design
reactions that can effectively avoid homocoupling and dehalo-
genation reaction, it is important to understand how substitution
patterns can influence the competition between reductive elim-
ination and second transmetalation. For this purpose, a series
of reactions were calculated. The calculated free energies of
the steric effect of the ortho substituent in Ar2.
To understand the above results, we analyzed the steric effect
in transition states. Shown in Figure 5 are selected calculated
structures with some geometrical parameters. For example,
L2PdAr1Ar2 26 takes on a nearly square-planar geometry. The
Ar1 and Ar2 are both perpendicular to the L2Pd plane, so that
they have no repulsion with each other. In the reductive
elimination transition state, 27, the C1---C2 distance is shortened
q
activation for the reductive elimination (∆G1 ) and second
q
transmetalation reaction (∆G2 ) along with additional selected
9
10206 J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009

Interpretation of the Mechanism of Biaryl Syntheses
A R T I C L E S
q
q
Table 2. Calculated Activation Free Energies of Reductive Elimination (∆G1 ) and Second Transmetalation (∆G2 ) and Selected Geometric
Information of 24-ts with Different Substitutuens (free energies in kcal/mol, bond lengths in Å)
a
∆G2(Corr)q ) ∆G2 -2.8 kcal/mol, based on free volume theory for bimolecular reactions in solution.44
q
to about 1.95 Å, while both Pd-C1 and Pd-C2 distances are
only slightly lengthened compared to the distance in 26. In 27,
both phenyl groups maintain a perpendicular conformation and
the shortest H/H distance between the two phenyl groups is 2.2
Å, indicating that when an ortho substituent is introduced, a
significant steric interaction will occur.
Structure 30 is the transition state of the second transmeta-
lation for the parent system (entry 1 of Table 3). While the
spectrator phenyl group still coordinates with Pd in the L2Pd
plane with a perpendicular conformation, the two phenyl groups
that are involved in transmetalation are pointing up and down
the L2Pd plane, respectively. They are nearly coplanar with each
other but have to adopt a conformation so that they coordinate
Pd through their π faces. The Pd-C2 and Pd-C3 distances
are close to about 3.0 Å, significantly longer than a normal
Pd-C bond. On the other hand, the two Zn-C bonds are 2.02
Å. Thus, the transition state can be considered as a zwitterionic
species with the Pd moiety positively charged and the Zn moiety
negatively charged.
There is a short distance of 2.60 Å between the two
designated ortho hydrogens in 30b. This indicates that when
one is replaced by a bulky substituent, steric interaction will
be introduced. Structures 31 and 32 are the transition states
for the second transmetalation of entries 4 and 6, respectively.
9
J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009 10207

A R T I C L E S
Liu et al.
Table 3. Biaryl Syntheses through Negishi Coupling without
(entries 2 and 8 in Table 1), only the cross-coupling product
is observed, and the calculation results in Table 2 indicate
Homocoupling Productsa
q
that the reductive elimination has a much lower barrier (∆G1 )
q
than the second transmetalation (∆G2 ). However, when Ar1
has an ortho substituent (entries 1, 3-5, 7, and 9 in Table
1), a mixture of cross-coupling and homocoupling products
is experimentally observed. An explanation for this comes
from the theoretical calculations; as for the ortho-substituents
(entries 4-7 in Table 2), the free energies of activation of
reductive elimination and second transmetalation steps are
comparable.
yieldb
%
entry
R1
R2
4
5
6
1
2
3
4
5
H
4-MeO
(1k) 2-CO2Et
(9a) trace nd
(9a) trace
94
97
92
99
90
(1i)
2-CO2Et
2-CO2Et
0
4-CO2Et (1c)
(9a) trace nd
(2b) trace nd
H
4-MeO
(1k) 2-iPr
(1i)
2-iPr
2-iPr
(2b) trace
0
6
4-CO2Et (1c)
(2b) trace trace 88
7
8
9
10
H
(1k) 2,4,6-trimethyl (2c)
4-CO2Et (1c) 2,4,6-trimethyl (2c)
(1k) 2-MeO
(2d) 3c
(1i) 2-MeO
(2d) 6c
trace nd
90
75
95
93
trace
0
H
4-MeO
nd
nd
a The reactions were conducted with 0.5 mmol of ArI, 1 mmol of
ArZnCl, and 3 mol % of PdCl2(dppf) in THF at 60 °C for 2 h. “nd”
means “not determined”. b Isolated yield. c GC yield.
In both structures, the phenyl group with the substituent
(COOMe in 31 and i-Pr in 32) rotates by about 24-32° so
that the above-mentioned steric interaction is avoided. This
rotation of phenyl group only slightly affects the bonding in
Pd-C2 and Zn-C2. Thus, the steric interaction of the
substituent can be largely avoided. This explains the results
shown in entries 4-7 of Table 3, where an ortho substituent
in Ar1 only slightly increases the free energy of activation
The results in entries 8-11 of Table 2 are most interesting.
They suggest that if Ar2ZnCl bears an ortho substituent,
although the substituent increases the barrier of reductive
elimination, it has a greater effect on the barrier of second
transmetalation such that it effectively blocks the formation
of homocoupling and dehalogenation products (eq 4). Thus,
through theoretical insight, we can propose an experimental
strategy for a more efficient synthesis of ortho-substituted
biaryl compounds using the Negishi coupling reaction, which
is to use a less substituted Ar1-I and a more substituted
Ar2ZnCl.
q
of the second transmetalation (∆G2 ). In fact, in the cases of
q
COOMe and OMe substituents, ∆G2 is even reduced
compared to the parent system due to electrostatic interactions
of these substituents with Zn.
The situation for Ar2 with an ortho substituent is very
different. The spectrator aryl group in the second transmeta-
lation is Ar2. Inspection of structure 30 reveals that the
spectrator phenyl group orients in such a way that its two
ortho hydrogens point to the other two phenyl groups,
resulting in two short C---H distances of 3.06 and 2.82 Å,
respectively. When either of them is replaced by a bulky
group, significant steric interactions will be introduced. This
is indeed the case. Structure 33 is the transition state of
transmetalation for entry 10. Due to the isopropyl group in
the spectrator aryl (Ar2), the top aryl group (Ar1) suffers such
large interactions that it has to move away from the Pd. The
Pd-C3 distance changes from 3.0 Å in other structures to
3.7 Å in 33.
5. Experimental Verification. We have conducted a series
of experiments to test the above strategy (eq 4) for the formation
of biaryl compounds. These results are summarized in Table 3.
Cross-coupling products are produced in high yields for all the
tested reactions. Since the above theoretical calculations sug-
gested a strategy of matching less-substituted Ar1-I and more
subsituted Ar2-ZnCl, we are gratified to provide experimental
evidence that indeed this strategy works toward exclusively
generating the cross-coupled biaryls.
Conclusion
In summary, the Ar1-Pd-Ar2 is traditionally considered
as the key intermediate leading to the reductive elimination
step that provides the cross-coupling products in the Negishi
coupling reaction. In this paper, we report the observation
of a second transmetalation step that occurs between the key
intermediate Ar1-Pd-Ar2 and the organozinc reagent
Ar2-ZnX. This critical insight provides the first understand-
ing of how dehalogenated and homocoupling products are
formed in the Negishi reaction. Indeed, an initial transmeta-
lation step gives the Ar1PdAr2 species; however, we show
that a second transmetalation step between Ar2ZnX with
Ar1PdAr2 provides Ar2PdAr1 and Ar1ZnX. These species lead
to the homocoupling and dehalogenated products, respec-
tively. DFT calculations show that a critical competition
occurs in the second transmetalation and reductive elimination
steps. These calculations indicate that an ortho substituent
in Ar1I favors the second transmetalation reaction, while an
ortho substituent in Ar2ZnCl significantly disfavors a second
transmetalation. This suggests a strategy to avoid homocou-
pling and dehalogenation side product formation, i.e., to
q
q
To make the comparison between ∆G1 and ∆G2 more
meaningful, some adjustment to ∆G2q is necessary. The reduc-
tive elimination is a unimolecular reaction; thus, the calculated
entropy contribution to free energy of activation is reasonable.
However, since the second transmetalation reaction is a bimo-
lecular reaction, and it is difficult to accurately estimate the
activation entropy for this reaction because it is a solution-based
q
reaction, we corrected the activation free energy ∆G2 based
on the free volume theory.44 This provides a rate of a solution-
based bimolecular reaction that is about 80 times faster than
the corresponding reaction in the gas phase. This means that
the activation free energy for solution reaction should be about
2.8 kcal/mol smaller than the value calculated in the gas phase.
q
The adjusted activation free energies ∆G2 are given as
∆G2(Corr)q.
Our calculation results can easily explain the experimental
observations shown in Table 1. When Ar1 is para substituted
9
10208 J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009

Interpretation of the Mechanism of Biaryl Syntheses
A R T I C L E S
Figure 5. Calculated geometries of selected intermediates and transition states in Table 3: 26 and 27 are intermediate and reductive transition states of entry
1; 28 and 29 are reductive transition states of entries 4 and 6, respectively; 30, 31, 32, and 33 are transition states of transmetalation of entries 1, 4, 6, and
10, respectively.
choose a less sterically crowded Ar1I and use an ortho-
separated, and then extracted with ethyl acetate. The organic layer
was dried over sodium sulfate, filtered, and concentrated. Purifica-
substituted Ar2ZnCl. This strategy has been experimentally
tion of products was accomplished by silica gel chromatography.
Computation Methodology. All calculations were carried out
with the GAUSSIAN 03 package.59 The Becke proposed hybrid
proven to work for the simplest case (Table 3). Overall, this
paper provides considerable experimental and theoretical
insight into the mechanism of transmetalation for Negishi
coupling.
(55) Tsou, T. T.; Kochi, J. K. J. Am. Chem. Soc. 1979, 101, 7547–7560.
(56) Yagyu, T.; Hamada, M.; Osakada, K.; Yamamoto, T. Organometallics
2001, 20, 1087–1101.
Experimental Section
General Procedure for Negishi Coupling. Under nitrogen,
PdCl2(dppf) (0.015 mmol) and an iodoaryl (0.5 mmol) were
dissolved in THF (0.5 mL). Then, to this THF solution was added
the aryl zinc reagent (1 mmol). The mixture, which immediately
turned black, was stirred for 2 h at 60 °C. The reaction was
quenched with saturated aqueous ammonium chloride solution,
(57) Suzaki, Y.; Osakada, K. Organometallics 2003, 22, 2193–2195.
(58) Ozawa, F.; Hidaka, T.; Yamamoto, T.; Yamamoto, A. J. Organomet.
Chem. 1987, 330, 253–263.
(59) Frisch, M. J.; et al. Gaussian 03, revision C.02; Gaussian, Inc.:
Wallingford, CT, 2004.
(60) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(61) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B: Condens. Matter 1988,
37, 785–789.
(50) van Asselt, R.; Elsevier, C. J. Organometallics 1994, 13, 1972–1980.
(51) Ozawa, F.; Kurihara, K.; Fujimori, M.; Hidaka, T.; Toyoshima, T.;
Yamamoto, A. Organometallics 1989, 8, 180–188.
(62) Cances, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032–
3041.
(63) Cances, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1998, 109, 260–
266.
(52) Giovannini, R.; Stuedemann, T.; Devasagayaraj, A.; Dussin, G.;
Knochel, P. J. Org. Chem. 1999, 64, 3544–3553.
(64) Cossi, M.; Barone, V.; Mennucci, B.; Tomasi, J. Chem. Phys. Lett.
1998, 286, 253–260.
(53) Kraatz, H.-B.; Van der Boom, M. E.; Ben-David, Y.; Milstein, D.
Isr. J. Chem. 2001, 41, 163–171.
(65) McQuarrie, D. A. Statistical Mechanics; Harper and Row: New York,
1976.
(54) Casado, A. L.; Espinet, P. Organometallics 1998, 17, 3677–3683.
9
J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009 10209

A R T I C L E S
Liu et al.
(B3),60 together with the Lee-Yang-Parr (LYP)61 correlation
functionals, has been chosen for this work. The SDD basis set was
used for Pd, Zn, and I atoms. The standard all-electron 6-31G(d)
basis set was used for all the other atoms. Aqueous phase single-
point energies were carried out at the same level using the self-
consistent reaction field method (SCRF), by the integral equation
formalism polarizable continuum model62-64 (IEFPCM) with THF
solvent. Frequencies, thermal corrections, and entropies were
calculated at this level of theory using the standard statistical
mechanical method at 298 K.65
kind help in revising the manuscript. This work was supported by
National Natural Science Foundation of China (20832003, 20772093,
20502020, and 20225312), the Excellent Youth Foundation of
Hubei Scientific Committee, and specialized Research Fund for the
Doctoral Program of Higher Education (20060486005), and the
Research Grants Council of Hong Kong (N_HKUST 623/04).
Supporting Information Available: Spectroscopic data, ex-
periments details, and DFT calculation details are available free
of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We thank Dr. Shen Hong of Merck & Co.,
Inc. and Professor John Sowa of Seton Hall University for their
JA903277D
9
10210 J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009